
The current study investigated the probable impact from a tsunami to a populated area located along the northwest ecuadorian coast, specifically in the key oil-industrial city of esmeraldas. a numerical tsunami simulation was performed considering the seismological and tectonic aspects of the area. The damage probability was calculated using fragility functions (ffs). Briefly, 16 cases of source models with slightly different fault parameters were tested, where one was selected as the worst scenario of tsunami inundation. This scenario was a hypothetic earthquake case (mw 8.7) located in front of esmeraldas city, approximately 100 km offshore along the ecuador—colombia trench, with three shallow fault segments (top depth of 10 km), a strike aligned with the trench axis, a middle dip angle of 28°, assuming large slips of 5 to 15 m, and a rake angle of 90°. The results from the numerical simulation were comparable to a similar study previously conducted and with those of historically documented data. The tsunami damage estimation using FFs resulted in estimated damages of 50% and 44% in exposed buildings and population, respectively. Results also showed that the most impacted areas were located next to the coastal shoreline and river. tourism, oil exports, and port activities, in general, would be affected in this scenario; thus, compromising important industries that support the national budget. Results from this study would assist in designing or improving tsunami risk reduction strategies, disaster management, use of coastal zones, and planning better policies.
Citation: Teresa Vera San Martín, Leonardo Gutierrez, Mario Palacios, Erick Mas, Adriano Bruno, Shunichi Koshimura. Tsunami damage estimation in Esmeraldas, Ecuador using fragility functions[J]. AIMS Geosciences, 2021, 7(4): 669-694. doi: 10.3934/geosci.2021040
[1] | Yifeng Wang . Journal summary from Editor in Chief. AIMS Environmental Science, 2025, 12(2): 252-252. doi: 10.3934/environsci.2025011 |
[2] | Yifeng Wang . Journal summary from Editor in Chief. AIMS Environmental Science, 2021, 8(2): 100-100. doi: 10.3934/environsci.2021007 |
[3] | Yifeng Wang . Journal summary from Editor in Chief. AIMS Environmental Science, 2022, 9(2): 217-217. doi: 10.3934/environsci.2022014 |
[4] | Yifeng Wang . Journal summary from Editor in Chief. AIMS Environmental Science, 2023, 10(2): 245-245. doi: 10.3934/environsci.2023014 |
[5] | Yifeng Wang . Journal summary from Editor in Chief. AIMS Environmental Science, 2018, 5(1): 64-66. doi: 10.3934/environsci.2018.1.64 |
[6] | Yifeng Wang . Journal summary from Editor in Chief. AIMS Environmental Science, 2016, 3(1): 140-140. doi: 10.3934/environsci.2016.1.140 |
[7] | Yifeng Wang . Journal summary from Editor in Chief. AIMS Environmental Science, 2019, 6(4): 262-264. doi: 10.3934/environsci.2019.4.262 |
[8] | Yifeng Wang . Journal summary from Editor in Chief. AIMS Environmental Science, 2020, 7(2): 153-155. doi: 10.3934/environsci.2020009 |
[9] | Yifeng Wang . Journal summary from Editor in Chief. AIMS Environmental Science, 2017, 4(2): 287-288. doi: 10.3934/environsci.2017.2.287 |
[10] | Rukhsana Kokkadan, Resha Neznin, Praseeja Cheruparambath, Jerisa Cabilao, Salma Albouchi . A Study of Infaunal Abundance, Diversity and Distribution in Chettuva Mangrove, Kerala, India. AIMS Environmental Science, 2023, 10(1): 82-92. doi: 10.3934/environsci.2023005 |
The current study investigated the probable impact from a tsunami to a populated area located along the northwest ecuadorian coast, specifically in the key oil-industrial city of esmeraldas. a numerical tsunami simulation was performed considering the seismological and tectonic aspects of the area. The damage probability was calculated using fragility functions (ffs). Briefly, 16 cases of source models with slightly different fault parameters were tested, where one was selected as the worst scenario of tsunami inundation. This scenario was a hypothetic earthquake case (mw 8.7) located in front of esmeraldas city, approximately 100 km offshore along the ecuador—colombia trench, with three shallow fault segments (top depth of 10 km), a strike aligned with the trench axis, a middle dip angle of 28°, assuming large slips of 5 to 15 m, and a rake angle of 90°. The results from the numerical simulation were comparable to a similar study previously conducted and with those of historically documented data. The tsunami damage estimation using FFs resulted in estimated damages of 50% and 44% in exposed buildings and population, respectively. Results also showed that the most impacted areas were located next to the coastal shoreline and river. tourism, oil exports, and port activities, in general, would be affected in this scenario; thus, compromising important industries that support the national budget. Results from this study would assist in designing or improving tsunami risk reduction strategies, disaster management, use of coastal zones, and planning better policies.
Robot visual servo control is an advanced robotics technology widely used in agricultural [1], industrial [2] and medical [3] scenarios. Vision sensors enable agile access to rich environmental information, enabling robots to achieve precise and efficient operations in unstructured environments. The categories of eye-in-hand configuration and eye-to-hand configuration for visual servoing are based on the spatial relationship between the vision sensor and the robot. Distinguished by the information fed back from the vision sensor, visual servoing is classified into position-based visual servoing [4,5,6], image-based visual servoing [7,8,9,10,11] and hybrid visual servoing [12,13,14]. Among them, the research and application of IBVS are the most extensive. This study discusses the design of an image-based visual servo controller.
In visual servo tasks, the system constraints are an influential factor that must be considered. If any image feature is out of the field of the camera, it can be said to violate the visibility constraint. An actuator constraint violation occurs when the robot's maximum permissible torque is exceeded by the controller's input torque. All these violations of constraints can result in the failure of the visual servo task. Therefore, meeting the system constraints is a problem that must be solved in visual servo control. Model predictive control can naturally build system constraints into the optimization problem to ensure constraint satisfaction by constructing optimization problems to solve controller actions[15]. Thus, model predictive control methods are often used to execute constrained visual servo control [16,17,18,19,20,21]. A conjugate visual predictive control strategy was presented in [16] with internal and external constraints in uncertain environments for visual servoing. A predictive control method that simultaneously considers system constraints and uncertainties was proposed in [17], where the prediction model uses a depth-independent Jacobian matrix. However, it was only validated on a planar robot manipulator. A quasi-minimum-maximum MPC method applied to visual servoing was proposed in [18], where the depth values are fixed constants in the Jacobian matrix. A predictive control strategy based on a nonlinear state observer was proposed in [19] to achieve attitude control of the unmanned aerial vehicle (UAV) by using the IBVS method. A nonlinear model predictive control technique [20] based on Gaussian processes was suggested for limiting mobile robots while considering camera visibility limits and robot hardware constraints. An nonlinear model predictive model (NMPC) strategy[21] was proposed to solve the six-degree-of-freedom robotic constrained visual servo but with a heavy computational burden. It is necessary to select the weight matrix of the objective function in the model predictive controller, which reflects the relative importances of different system state variables to the objective function. Generally, the choice of the weight matrix is full of subjectivity and requires many attempts. However, even after a large number of tests, the best control quality cannot be obtained. The weight matrix tuning method of model predictive control has always been a concern. Shridhar and Cooper proposed a multivariable model predictive control adjustment strategy, but it is only applicable to unconstrained systems [22]. Particle swarm and genetic algorithms have also been used to rectify the model predictive controller parameters [23,24,25].
The discipline of artificial intelligence greatly benefits from the interdisciplinary study of machine learning[26,27]. Among different approaches, RL, which has gained prominence recently, uses learning procedures to accomplish a certain goal to explain and address the issue that arises in an agent's interaction with the environment[28,29,30,31]. Many existing works use RL to assist intelligent control methods, to bring better control in various working scenarios. An adaptive PID control method [28] was proposed for wind energy conversion system control using actor-critic learning to adaptively adjust the controller parameters, and the robustness of the presented control strategy was confirmed by comparative simulation. A deep RL method[29] was used to learn an effective adaptive PID gain adjustment strategy for the control of a marine vessel. A double-Q-learning algorithm [30] was presented to adjust a multilevel PID controller, which was used for simulation research and experiments of a mobile robot. A maximum entropy depth RL method [31] was presented to adjust an AUV controller, to maintain the balance between performance and robustness. The Q-learning algorithm was adopted to adaptively adjust the visual servo gain in [32], which significantly improved the convergence speed and stability compared to the traditional IBVS method with a fixed servo gain. The effectiveness of the proposed method was verified by simulations and experiments. Several works combine model predictive control with RL to carry out research [33,34,35]. A model predictive control parameter rectification method based on RL was proposed in [33], and the ideal weight matrix was explored in a short time. The NPMC methods tuned by RL were applied to control a UAV in [34,35].
Although reinforcement learning has been applied to visual servoing, the most commonly used algorithm is still Q-learning. However, the performance of Q-learning in continuous action spaces needs to be improved[36]. DDPG is a reinforcement learning algorithm based on deep neural networks and policy gradient methods which can effectively solve the problem of continuous action spaces by outputting a probability distribution of continuous action values instead of hard-coding discrete actions [37]. Compared with Q-learning, DDPG uses neural networks to approximate the value function and policy, which can converge faster and apply a loss function that combines action value and policy gradient, thus better optimizing the policy and value function. In addition, DDPG also uses techniques such as experience replay buffer and batch training to improve the stability of the algorithm[38].
Inspired by the above works, this work presents a model predictive control strategy for robot manipulator visual servoing tuned by DDPG. The main contributions of this paper are as follows:
● To improve the servo efficiency and control accuracy of constrained IBVS systems, this paper proposes an MPC-based IBVS method tuned by DDPG. A depth-independent image interaction matrix is established as the predictive model, and more suitable predictive control weight matrix parameters are trained offline by the DDPG algorithm. Then, the control signal is given through the MPC controller.
● Compared with the traditional MPC-based IBVS method, the proposed method compensates for the disadvantages of the traditional trial-and-error method in tuning the weight matrix parameters and provides better weight matrix parameters, thereby improving visual servo efficiency and steady-state accuracy.
● Compared with the MPC-based IBVS method tuned by Q-learning, the proposed method obtains higher cumulative rewards in the continuous visual servo space and reduces the settling time.
The subsequent structure of this study is as follows. In Section 2, the model of the visual servo system is established to provide the prediction model needed by the model predictive controller. In Section 3, the model predictive controller of IBVS tuned by RL is designed. First, the visual servo model predictive control method is introduced, while later the RL and policy gradient algorithm is introduced, and finally the DDPG-based model prediction control weight matrix strategy is described. In Section 4, the DDPG learning process is introduced, and the effectiveness of the proposed method is verified by simulations. Finally, the research conclusion of this study is summarized in Section 5.
In robot manipulator visual servo tasks, visibility constraints and robot joint constraints are the system constraint terms that must be considered. The MPC method can solve the system input-output constraint problem, so this study is based on the MPC method to solve the constrained visual servo problem. The predictive model is a very important part of MPC, which can predict the future values of the process output according to the input of the designed controller and the past status of the system. In this study, the association between the rates of variation of image coordinates of feature points and the joint velocity is described by building a depth-independent Jacobian matrix. Figure 1 shows schematic diagrams of the robot manipulator visual servo eye-in-hand configuration and eye-to-hand configuration, respectively. In the two configurations, the image coordinates of feature points can be uniformly represented as
s=(uv)=1Zi(cT1cT2)P(α1) | (2.1) |
where s=(u,v)T denotes the 2-D coordinates of the feature point in the image framework, and u and v represent the pixel coordinates projected on the u- and v-axes, respectively. Zi denotes the depth of the feature point concerning the camera framework. C∈ℜ3×4 denotes the unknown perspective projection matrix, and cT1,cT2, and cT3 are the first, second and third rows of the matrix C, respectively. α denotes the 3-D coordinates of the feature point in the Cartesian coordinate system. P denotes the homogeneous transformation matrix, which is up to the forward kinematics of the robot manipulator. By differentiating both sides of the above equation simultaneously with respect to time, it can be obtained that
˙s=1Z2i((cT1cT2)P(˙α1)˙qZi−˙Zi(cT1cT2)P(α1))=1Zi(cT1cT2)P(˙α1)˙q−1Z2i˙Zi(cT1cT2)P(α1) | (2.2) |
The depth Zi can be linearly expressed as
Zi=c3TP(α1). | (2.3) |
Taking the derivative of both sides of Eq (2.3), it can be obtained that
˙Zi=cT3P(˙α1)˙q. | (2.4) |
Substituting Eq (2.4) into Eq (2.2), the time-related association between the joint velocity and the variation of the feature points can be given as
˙s=1Zi(cT1cT2)P(˙α1)˙q−1ZiscT3P(˙α1)˙q=1Zi(cT1−ucT3cT2−vcT3)P(˙α1)˙q=1ZiΩ˙q | (2.5) |
where q denotes the joint angle of the robot manipulator, ˙q denotes the joint velocity of the robot manipulator, and Ω denotes the image Jacobian matrix, which is independent of depth. It can be given as
Ω=(cT1−ucT3cT2−vcT3)P(˙α1) | (2.6) |
In the formulas above, the depth-independent image Jacobian matrix Ω and the depth of feature point Zi are nonlinear. However, they can be represented linearly with regressor matrices and unknown parameter vectors by the following property.
Property 1. For any vector χ, the products Ωχ and Ziχ can be parameterized in a linear form as
Ωχ=A(χ,q,s)θ, | (2.7) |
Ziχ=B(χ,s)θ, | (2.8) |
where A(χ,q,s) and B(χ,s) are the regressor matrices. θ is the corresponding parameter vector determined by the products of the camera parameters and the robot kinematic parameters.
A constrained control system is usually a kind of system that has constraints on control inputs, outputs and system states. In practical application scenarios, robot visual servo systems will be constrained by image visibility, physical constraints of joints, etc. Therefore, a robot visual servo system is a typical constraint control system. MPC is widely used in the control of constrained systems. In this study, the model predictive controller samples the image coordinates of feature points within a given sampling period and transforms the visual servo problem into a constrained optimization problem to generate the most suitable joint torque signal and minimize the incremental change in joint torque while minimizing the deviation of the predicted image. The purpose of visual servo control is to make the control process as stable as possible while completing visual servo control. The objective function in the optimization problem includes weight matrices, and an appropriate weight matrix will effectively improve the control performance of visual servoing. Previous weight matrices were obtained by researchers through repeated experiments, and it was challenging to achieve the best control effect. In this study, the DDPG-based RL strategy is adopted to modify the appropriate weight matrix of the objective function of MPC to optimize the visual servo performance of MPC-based IBVS. The design of the visual servo model predictive control is introduced in the last section, the RL and policy gradient algorithm is introduced in Section 2, and the reinforcement-learning-based objective function weight matrix rectification method is introduced in Section 3. The control scheme of the presented DDPG-based MPC (DDPG-MPC) IBVS algorithm is given in Figure 2.
To develop the MPC-based IBVS controller, the discrete-time system model is obtained according to Eq (2.5).
s(k+1)=s(k)+Td1ZiΩuτ(k) | (3.1) |
When the control sequence obtained from the prediction model is applied to the IBVS system, the predicted system output for the next Np time steps is
s(k+1∣k)=s(k)+Td1Zi(k)Ω(k)uτ(k)s(k+2∣k)=s(k)+Td1Zi(k)Ω(k)uτ(k)+Td1Zi(k+1)Ω(k+1)uτ(k+1)⋮s(k+Np∣k)=s(k)+Td1Zi(k)Ω(k)uτ(k)+⋯+Td1Zi(k+i−1)Ω(k+i−1)uτ(k+Nc−1) | (3.2) |
where Np is predictive time domain, Nc is control time domain, and Td is the sampling period. The following constrained optimization problem can be solved to calculate the optimal joint torque input.
Π:minΔUτ(k)(se(k),ΔUτ(k)) | (3.3) |
subject to
se(k+i)=ˉs(k+i)−s∗(k+i) | (3.4) |
ΔUτ(k)=[Δuτ(k)T,Δuτ(k+1)T⋯Δuτ(k+Np−2)T,Δuτ(k+Np−1)T] | (3.5) |
Δuτ(k)=uτ(k)−uτ(k−1) | (3.6) |
where s∗ denotes the desired image coordinates of feature points, ˉs denotes the predicted image coordinates of feature points, se denotes the image deviation within the prediction period, Δuτ(k) denotes the changing values of the control input, and ΔUτ(k) denotes the optimal sequence of the changing value of the control input within the prediction period. To minimize the predicted image coordinate deviation and the minimum control input variation, the quadratic cost function can be described as
G(se(k),ΔUτ(k))=Np∑i=1‖ˉs(k+i)−s∗(k+i)‖2Rs+Nc∑i=1‖Δuτ(k+i−1)‖2Ru | (3.7) |
where Np denotes the prediction horizon, Nc denotes the control horizon, and Nc≤Np normally. Rs≥0 and Ru≥0 denote the weight matrices of the image coordinate deviation and the sequence of changing values of the control input, respectively. The constraints in the limited visual servoing system can be presented by the following formulas:
smin≤s(k)≤smax | (3.8) |
Uτmin≤Uτ(k)≤Uτmax | (3.9) |
qmin≤q(k)≤qmax | (3.10) |
˙qmin≤˙q(k)≤˙qmax | (3.11) |
Equation (3.8) represents the visibility constraints, Eq (3.9) represents the torque constraints, Eq (3.10) represents the joint angle constraints, and Eq (3.11) represents the joint velocity constraints.
The weight matrices Rs and Ru in the minimization function affect the control effect of the model predictive control. The matrix Rs is used to describe the weights of the control deviations. When the value of Rs is too large, the image coordinates of feature points will converge to the target image coordinates during the control process. However, it ignores the control input variation, resulting in the appearance of an input jitter, which reduces the response quality of the control process. The matrix Ru is used to describe the weights of the change value of control inputs. When the value of the weight matrix Ru used to describe the control input variable is too large, it will cause the control process to pay too much attention to the slow change of the control input variables, and the control system's response time will lengthen such that the visual servo task cannot be completed quickly. It is time-consuming to adjust the weight matrix of the objective function manually, and even then, the optimal control state of the model predictive controller cannot be reached. In this work, an RL-based method is proposed to adjust the weight matrix of the objective function, and it is introduced in the next section.
RL has excellent ability to solve sequential problems. In addition, RL is a computational method, and robots can achieve their goals by interacting with the environment. RL is usually considered a Markov decision process (MDP). MDP is based on the tuple ⟨S,P,r,δ⟩ of the Markov reward process, denoted as ⟨S,A,P,r,δ⟩:
● S is the collection of states for the robot.
● A is the collection of actions that the robot performs according to its current state.
● P(s′∣s,a) is the state transition function of the current state s for action A to arrive at state s′
● r(s,a) is the reward function used to evaluate the reward generated by action a under the current state s.
● δ∈[0,1] is a discount factor, representing the importance of the future return series.
The robot makes a decision of action in a state of the environment and applies the action to the environment. The environment changes, and the reward is passed on to the robot in the next state. This kind of interaction is iterative, and the robot's goal is to maximize the accumulated reward expectation over the process of multiple rounds. When the robot is in the environment st to perform action at, there will be a reward rt+1, and the environment will be updated to st+1. When the time approaches infinity, the accumulated reward rt can be expressed as
rt=rt+1+δrt+2+δ2rt+3+⋯=∞∑k=0δkrt+k+1. | (3.12) |
In RL, the DDPG is a combination of the deterministic policy gradient and the deep neural network, which can effectively solve the problem of continuous action space. In this study, DDPG is adopted to adjust the appropriate IBVS model predictive control objective function weight matrix. The strategy for tuning weight matrix parameters can be expressed by ω.
The learning goal is to find the strategy with the highest expected cumulative return, which can be expressed as
ω∗(s)=argmaxωE[∞∑t=0δtr(st,ω(st))∣s0=s] | (3.13) |
The gradient under the highest accumulated reward is expressed as:
J(ω)=E[∞∑t=0δtr(st,ω(st))] | (3.14) |
According to the deterministic strategy gradient theory, the parameterized vector η of the optimal strategy is expressed as
∇ηJ(ω)=Est∼ρ[∇aQ(s,a)∣a=ω(s∣η)∇ηω(s∣η)] | (3.15) |
where ρ is a random probabilistic strategy, and Q(s,a) denotes the action value function.
To optimize the visual servo control performance by tuning the MPC weight matrices based on DDPG, the element of the MDP solved by DDPG can be expressed as follows:
● As shown by Eq (3.1), the variation in feature points is jointly influenced by Td1ZiΩ and ˙q. The parameters of the weight matrices determine the output joint velocity ˙q, which depends on Td1ZiΩ. There, S is defined as Td1ZiΩ.
● A is defined as the MPC weight matrix (Rs,Ru), where 0I8×8≤Rs≤30I8×8, 0I2×2≤Ru≤5I6×6.
● r(s,a) represents the effect on the MPC-based IBVS task with action a under state s. If the deviation between the state of the feature points and the desired state of the u- and v- axes is within the threshold value, a positive reward will be given. If the visual servo task fails, nonpositive reward will be given. Every control action that affects the image coordinate deviation and deviation rate will be punished accordingly, so that the visual servo task is successful, and the control efficiency is improved.
rt={1−1−sTeWese−sT˙eW˙es˙efor|se(u)|<ζ,|se(v)|<ζvisual servo task failedelse |
The process of the DDPG-based MPC IBVS algorithm is shown in Algorithm 1. Four networks are required in the DDPG algorithm: The actor network is denoted as ωQ, the critic network is denoted as ωτ, the actor target network is denoted as ωQ′, and the critic target network is denoted as ωτ′. The robot performs action at at state st, updates to the next state st+1 and obtains the reward rt. The Q-value of the critic target network Qc−t can be obtained by the following formula:
Qc−t=ri+δQ′(st+1,τ′(st+1|ωτ′)|ωQ′) | (3.16) |
Algorithm 1: DDPG-based MPC IBVS Algorithm |
1 Initialize soft updating rate σ and discount factor δ; 2 Initialize the parameterized actor network ωQ and parameterized critic network ωτ; 3 Initialize the parameterized actor target network ωQ′ and parameterized critic target network ωτ′; 4 Initialize Gaussian noise κ; 5 Initialize memory pool Mp; 6 for episode = 1, 2, ⋯, N do ![]() |
The minimal loss function of the critical network is calculated by the gradient descent algorithm
Lωτ=1NN∑i=1Qc−t−Q(st,at|ωτ), | (3.17) |
∇Lωτ=1N[Qc−t−Q(st,at|ωτ)]∇ωτQ(st,at|ωQ) | (3.18) |
The actor network parameters are updated with the following formula:
∇ωQJ≈1N∑i∇aQ(s,a|ωτ)|a=Q(s|ωQ)∇ωQQ(s|ωQ). | (3.19) |
The actor target network and the critic target network are updated by the exponential smoothing method.
ωQ′←σωQ+(1−σ)ωQ′ | (3.20) |
ωτ′←σωτ+(1−σ)ωτ′ | (3.21) |
To confirm the effectiveness of the proposed method, this section gives the simulation comparison experiments of different IBVS methods acting on the same visual servo task. The differences in performance between different control methods will be described in detail.
The simulation object of this study is a visual servo system of a 6-DOF robot manipulator[39]. The camera is placed on the last joint of the 6-DOF robot manipulator. The focal length of the camera f0 is 0.0005 m. The scaling factors along the u- and v- axes are 269,167 pixels/m and 267,778 pixels/m, respectively. The coordinates of the image feature points are fed back from the camera and passed to the 6-DOF robot manipulator, which generates a change in pose so that the image feature points respond from the initial position to the desired position. Simulation training and experiments were performed using MATLAB/Simulink on a laptop computer with a 2.3 GHz Intel Core i7. In RL, the process of an intelligence executing a certain strategy to reach the termination state from the start state is usually referred to as an episode. In this study, the following rules need to be followed each time during the learning process:
1) During the learning process, if the squared difference between the current coordinates of the image feature point and the desired coordinates is lower than the set threshold, it is considered that the visual servo task is successful, and this round of learning ends.
2) During the learning process, if the squared difference between the current coordinates of the image feature point and the desired coordinates is higher than the set threshold, it is considered that the visual servo task has failed, and the current round of learning is ended.
3) During the learning process, if the current coordinates of the image feature point break the image constraint, the visual servo task is considered to have failed, and the current round of learning ends.
In the actor and actor target network, the actor and actor target are the input quantities, and action a is the output quantity. In the critic and critic target networks, the action pair (s,a) is the input, and the action value function Q(s,a) is the output. The activation function gives the neural network a nonlinear modeling capability. In this study, the activation function is chosen as the ReLU function described by Eq (4.1).
fr=max(0,t) | (4.1) |
The reward weight matrices are set as follows:
We=[0.5001], | (4.2) |
W˙e=[0.1000.05] | (4.3) |
A total of 1000 experiments were carried out, with 20 episodes in each round. The reward value for each round of experiments is the average value of 20 experiments. The hyperparameters of the DDPG algorithm are shown in Table 1.
Parameter | Value |
Discount Factor (δ) | 0.95 |
Network soft update parameters (σ) | 0.0005 |
Experience replay pool size | 105 |
Numbers of hidden layers | 2 |
First hidden layer size | 40 |
Second hidden layer size | 30 |
In the simulation training and experiments, the visual servo has four image feature points. The 3-D Cartesian coordinates of the image feature points are mapped to the two-dimensional image framework, arranged counterclockwise from O1→O2→O3→O4. We set the desired position of the visual servo feature points in the image plane as (400,525)Tpixels, (720,525)Tpixels, (720,320)Tpixels, (400,320)Tpixels.
Remark 1. The visibility constraint of the camera u-axis and v-axis can be expressed as umin=0,umax=1292,vmin=0,vmax=964. If the initial coordinates of randomly generated image feature points are outside the camera vision constraint, a set of initial image feature point coordinates will be randomly generated again until the initial coordinates of all image feature points are within the camera vision constraint.
The purpose of the comparative simulation experiments is to demonstrate the accuracy and stability of the DDPG-MPC IBVS method proposed in this paper. The simulation is divided into two parts. First, the control effect of this method is compared with the traditional model predictive control visual servo method (MPC-IBVS)[21]. Then, the control effect of the proposed method in this study is compared with the model predictive control IBVS method tuned by Q-learning (Q-learning-MPC IBVS).
In this section, the control effects of the proposed DDPG-MPC IBVS method and traditional MPC-IBVS are analyzed. In addition, to prove that the proposed method has a more stable and accurate control effect, this section selects two groups of different weight matrices for the MPC-IBVS method, which are called MPC-IBVS-A and MPC-IBVS-B. The weight matrix of MPC-IBVS-A is set as Rs=2I8×8,Ru=I6×6, the predictive time domain is set as Np=5, and the control time domain is set as Nc=2. The weight matrix of MPC-IBVS-B is set as Rs=25I8×8,Ru=5I6×6, the predictive time domain is set as Np=5, and the control time domain is set as Nc=2. The sampling period of the controller is 40 ms.
The weight matrix after training and rectification by the proposed method in the last part is
Rs=[10000000008000000007000000005000000005000000006000000008000000009],Ru=[500000030000002000000300000020000004] | (4.4) |
The predictive time domain is set as Np=5, and the control time domain is set as Nc=2. To test the stability of various control methods, the desired feature point image coordinates are set to (400,525)T pixels, (720,525)T pixels, (720,320)T pixels, (400,320)T pixels. The first set of initial feature point image coordinates p1 are set to (233.5,740.7)T pixels, (240,600)T pixels, (115.2,612.7)T pixels, (107.3,760.1)T pixels, and the second set of initial feature point image coordinates p2 are set to (917.2,350.1)T pixels, (919,200.3)T pixels, (801.9,226.1)T pixels, (800.2,374.8)T pixels. The robot manipulator joint velocity constraint is limited to 0.5 rad/s.
First, Figure 3 gives the comparative simulation results of the DDPG-MPC IBVS, MPC-IBVS-A and MPC-IBVS-B methods under the initial coordinates p1. From Figure 3(a)–(c), it can be observed that under all three IBVS methods, the image feature points can successfully respond from the initial coordinates to the desired coordinates and finally stabilize in the desired state. However, Figure 3(d)–(f) shows that the image deviation converges at the fastest rate under the proposed method compared to the other methods. The settling time of the DDPG-MPC IBVS method is approximately 5 s. The settling time of MPC-IBVS-A is approximately 10 s. The settling time of MPC-IBVS-B is approximately 8 s. The steady-state error of the proposed method is less than those of MPC-IBVS-A and MPC-IBVS-B. The above experiments show that the proposed method has better control performance when the initial position is p1.
Then, the results of the comparative simulation at the initial coordinates p2 are given in Figure 4. From Figure 4(a)–(c), it can be observed that the image feature points can also respond from the initial coordinates to the desired coordinates and stabilize in the desired state under all three MPC-based IBVS methods. Similar to the above simulation results, the proposed method can make the image feature points respond and stabilize at the desired coordinates as fast as possible (approximately 5 s). Different from the above simulation results, in the initial state p2, the settling time of MPC-IBVS-A (approximately 8 s) is longer than that of MPC-IBVS-B (approximately 10 s). This indicates that the different parameters of the weight matrix of the objective function set considered cannot remain stable in giving the optimal control effect in different visual servo tasks. This is because the weight matrix contains multiple parameters, and the selection of the optimal weight matrix parameters by humans is also not achievable in practical applications. With the increase in accumulated rewards, the intensive training method can gradually find the optimal weight matrix parameters, thus achieving better visual servo control performance.
In the IBVS tasks, the relationship between the target object and the visual servo system is usually represented by four or more feature points. As shown in Eq (4.4), the four feature points in this study correspond to the eight parameters in the image deviation weight matrix Rs. At the same time, the 6-DOF manipulator corresponds to six parameters in the weight matrix Ru. Then, there are fourteen parameters of the weight matrices that need to be set in the MPC-based IBVS method. Although these fourteen parameters can be given artificially through trial and error, it is possible to complete the visual servo task, but finding the optimal weight matrix parameters requires a lot of work, which is impossible in practical applications. The proposed method can get better weight matrix parameters with a certain number of training sets in less than one hour, improve the servo efficiency and accuracy and effectively improve the control quality.
The evaluation results of the above simulation experiments are shown in Table 2, which demonstrates that compared with the traditional MPC-based IBVS method, the DDPG-MPC IBVS method can complete the constrained visual servo task with higher control quality.
Initial state | Control method | DDPG-MPC-IBVS | MPC-IBVS-A | MPC-IBVS-B |
p1 | Settling time (s) | 5.24 | 9.98 | 8.13 |
Image deviation (pixels) | 4.98 | 10.21 | 10.5 | |
Control overshoot | 12% | 20% | 19% | |
p2 | Settling time (s) | 4.89 | 8.07 | 10.1 |
Image deviation (pixels) | 5.1 | 11.26 | 11.18 | |
Control overshoot | 12% | 18% | 20% |
In this part, the results of the comparative simulation between the DDPG-MPC IBVS method and the Q-Learning-MPC IBVS method are analyzed. Q-learning is an extremely important RL algorithm, and Q(s,a) is the benefit obtained by taking action a in a certain state s. The environment will provide feedback on the corresponding reward according to the action reward, so the main idea of the Q-learning algorithm is to build a Q-table with states and actions to store Q-values and then select the action that can obtain the maximum gain according to the Q-values. In the simulation of this study, the Q-learning algorithm is composed of a low-latitude Q-table. The objective function weight matrix is discretely tuned by the Q-learning algorithm.
In the comparison experiments, the desired coordinates of the feature points are the same as in the previous section, and the initial coordinates of the feature points are p1. From Figure 5(a), (b) we find that both the DDPG-MPC IBVS method and the Q-learning-MPC IBVS method can cope well with the system constraints. From Figure 5(d), (e), it can be observed that DDPG-MPC IBVS has a faster servo deviation convergence speed than the Q-Learning-MPC IBVS method, at 5 s versus 7 s, respectively. Furthermore, the deviation response curve based on DDPG-MPC IBVS is smoother. This is because the action set of Q-learning is discrete. In contrast, DDPG trains and corrects continuous visual servo actions through continuous network functions. Therefore, the DDPG algorithm tunes the more appropriate weight matrix of the objective function, so as to achieve a better visual servo control. As shown in Figure 5(c), the cumulative rewards of the two RL methods increase with the increase of rounds. However, the final cumulative rewards of the DDPG algorithm are higher than that of the Q-learning algorithm, which indicates that the DDPG algorithm has achieved better training results. This means that the DDPG-IBVS method can adjust more suitable weight matrix parameters, which leads to better visual servo control performance. The changes of the camera velocity V=[vx,vy,vz,wx,wy,wz]T∈R6×1 of the two methods are shown in Figure 5(f). The camera velocities do not fluctuate too much under the two methods, which means that the response processes are both smooth.
To describe the effectiveness of the proposed method more vividly, we quantitatively analyze the visual servo control effect for the model prediction control rectified by two different reinforcement learning algorithms. As seen from Table 3, after 30 random experiments, both methods can complete the constrained visual servo task with a high success rate and similar average image deviation. However, the DDPG-MPC IBVS method always has a faster convergence speed.
Control method | DDPG-MPC IBVS | Q-learning-MPC IBVS |
Success rate | 100% | 100% |
Average overshoot | 12% | 14% |
Average settling time (s) | 5.12 | 7.39 |
Average image deviation (pixels) | 5.06 | 5.2 |
In conclusion, the proposed method has faster servo efficiency and better control quality compared to the existing IBVS methods. The proposed method can effectively perform visual servo tasks.
In visual servoing, system constraints contain visibility constraints, and actuator constraints must be considered. To solve the constrained visual servo problem, a model predictive control IBVS method tuned by RL is proposed in this study. First, a depth-independent Jacobian matrix is established as the predictive model, and the optimal control input is found by minimizing the cost function of the predictive error. Different from traditional model predictive control methods, the weight matrix of the objective function is adjusted offline by the DDPG algorithm. Appropriate states, rewards and actions are defined in the training progress. Then, the accumulated rewards converge to specific values, which means that the DDPG algorithm has successfully learned the appropriate weight matrix parameters. Finally, in simulation experiments of a 6-DOF manipulator, the control effect of the proposed method is compared and analyzed with other visual servo control methods, and we find that the proposed method has better performance. In future work, we plan to explore the delayed visual servo control strategies caused by low-quality visual signals. We will design a predictive control IBVS method based on a time-delay predictive model, to improve the control stability of visual servo systems with time delay.
This research was funded by the Natural Science Foundation of Heilongjiang Province, grant number KY10400210217, and the Fundamental Strengthening Program Technical Field Fund, grant number 2021-JCJQ-JJ-0026.
The authors declare there is no conflict of interest.
[1] | Gupta HK, Gahalaut VK (2013) Three Great Tsunamis: Lisbon(1755) Sumatra-Andaman (2004) and Japan(2011), Springer Briefs in Earth Science, Springer. |
[2] | Levin B, Nosov M (2016) Physics of tsunamis, Springer. 399. |
[3] | IOC, Tsunami Glossary, 2013, Unesco: Paris, 45. |
[4] | Levin B, Nosov M (2009) Physics of Tsunamis, Springer. 327. |
[5] | Bryan E (2014) Tsunami The Underrated Hazard, Springer-Praxis. |
[6] |
Geist EL (2002) Complex earthquake rupture and local tsunamis. J Geophys Res 107: 2086. doi: 10.1029/2000JB000139
![]() |
[7] | Rooney A (2012) Tsunami! (NATURE'S FURY), Encyclopedia Britannica, Inc. |
[8] | Tolkova E (2018) Tsunami Propagation In Tidal Rivers. Sringer Briefs in Earth Sciences, Springer, 123. |
[9] | USGS, Search Earthquake Catalog, 2019. Available from: https://earthquake.usgs.gov/earthquakes/search/. |
[10] |
Otero LJ, Restrepo JC, Gonzalez M (2014) Tsunami hazard assessment in the southern Colombian Pacific basin and a proposal to regenerate a previous barrier island as protection. Nat Hazards Earth Sys Sci 14: 1155-1168. doi: 10.5194/nhess-14-1155-2014
![]() |
[11] |
Chlieh M, Mothes PA, Nocquet JM, et al. (2014) Distribution of discrete seismic asperities and aseismic slip along the Ecuadorian megathrust. Earth Planet Sci Lett 400: 292-301. doi: 10.1016/j.epsl.2014.05.027
![]() |
[12] |
Ioualalen M, Monfret T, Béthoux N, et al. (2014) Tsunami mapping in the Gulf of Guayaqui, Ecuador, due to local seismicity. Mar Geophys Res 35: 361-378. doi: 10.1007/s11001-014-9225-9
![]() |
[13] |
Ye L, Kanamori H, Avouac JP, et al. (2016) The 16 April 2016, M W 7.8 (M S 7.5) Ecuador earthquake: A quasi-repeat of the 1942 M S 7.5 earthquake and partial re-rupture of the 1906 M S 8.6 Colombia-Ecuador earthquake. Earth Planet Sci Lett 454: 248-258. doi: 10.1016/j.epsl.2016.09.006
![]() |
[14] |
Bethoux N, Segovia M, Alvarez V, et al. (2011) Seismological study of the central Ecuadorian margin: Evidence of upper plate deformation. J South Am Earth Sci 31: 139-152. doi: 10.1016/j.jsames.2010.08.001
![]() |
[15] |
Yi L, Xu CJ, Wen YM, et al. (2018) Rupture process of the 2016 Mw 7.8 Ecuador earthquake from joint inversion of InSAR data and teleseismic P waveforms. Tectonophysics 722: 163-174. doi: 10.1016/j.tecto.2017.10.028
![]() |
[16] |
Collot JY, Ribodetti A, Agudelo W, et al. (2011) The South Ecuador subduction channel: Evidence for a dynamic mega-shear zone from 2D fine-scale seismic reflection imaging and implications for material transfer. J Geophys Res Solid Earth 116: 1-20. doi: 10.1029/2011JB008429
![]() |
[17] |
Gutscher MA, Malavieille J, Lallemand S, et al. (1999) Tectonic segmentation of the North Andean margin: impact of the Carnegie Ridge collision. Earth Planet Scie Lett 168: 255-270. doi: 10.1016/S0012-821X(99)00060-6
![]() |
[18] | Kanamori H, McNally K (1982) Variable Rupture Mode of the Subduction Zone along the Ecuador-Colombia Coast. Bull Seismol Soc Am 72: 1241-1253. |
[19] | Arreaga P, Ortiz M, Farreras S (2006) Mapping the possible tsunami hazard as the first step towards a tsunami resistant community in Esmeraldas, Ecuador, Tsunamis Case Studies and Recent Developments, Springer: Netherlands, 203-215. |
[20] | Chunga K, Toulkeridis T, Vear-Grunauer X, et al. (2017) Review of Earthquakes and Tsunami Records and Characterization of Capable Faults on the Northwestern Coast Of Ecuador. J Tsunami Soc Int 36: 100-127. |
[21] | Contreras M (2014) Riesgo de tsunami en Ecuador. Ingenius 12: 68-75. |
[22] | Espinoza J (1992) Terremotos tsunamigénicos en el Ecuador. Acta Oceanográfica del Pacífico 7: 21-28. |
[23] |
Ioualalen M, Ratzov G, Collot J, et al. (2011) The tsunami signature on a submerged promontory: the case study of the Atacames Promontory, Ecuador. Geophys J Int 184: 680-688. doi: 10.1111/j.1365-246X.2010.04878.x
![]() |
[24] | Pararas C (2012) Potential of Tsunami Generation along the Colombia/Ecuador Subduction Margin and the Dolores-Guayaquil Mega-Thrust. J Tsunami Soc Int 31: 209-230. |
[25] |
De Risi R, Goda K (2016) Probabilistic Earthquake-Tsunami Multi-Hazard Analysis: Application to the Tohoku Region, Japan. Front Built Environ 2: 19. doi: 10.3389/fbuil.2016.00025
![]() |
[26] |
Koshimura S, Namegaya Y, Yanagisawa H (2009) Tsunami Fragility—A New Measure to Identify Tsunami Damage. J Disaster Res 4: 479-488. doi: 10.20965/jdr.2009.p0479
![]() |
[27] | Macabuag J (2017) Tsunami Damage Prediction for Buildings: Development of Methods for Empirical and Analytical Fragility Function Derivation, Department of Civil, Environmental and Geomatic Engineering, University College London. |
[28] |
Mas E, Koshimura S, Suppasri A, et al. (2012) Developing Tsunami fragility curves using remote sensing and survey data of the 2010 Chilean Tsunami in Dichato. Nat Hazards Earth Sys Sci 8: 2689-2697. doi: 10.5194/nhess-12-2689-2012
![]() |
[29] | Vera T (2015) Tsunami Damage Estimation in Esmeraldas, Ecuador Using Tsunami Fragility Functions, Tsukuba, Japan, 59. |
[30] |
Charvet I, Suppasri A, Imamura F (2014) Empirical fragility analysis of building damage caused by the 2011 Great East Japan tsunami in Ishinomaki city using ordinal regression, and influence of key geographical features. Stoch Environ Res Risk Assess 28: 1853-1867. doi: 10.1007/s00477-014-0850-2
![]() |
[31] |
Reese S, Bradley B, Bind J, et al. (2011) Empirical building fragilities from observed damage in the 2009 South Pacific tsunami. Earth Sci Rev 107: 156-173. doi: 10.1016/j.earscirev.2011.01.009
![]() |
[32] |
Alam MS, Barbosa AR, Scott MH, et al. (2018) Development of Physics-Based Tsunami Fragility Functions Considering Structural Member Failures. J Struct Eng 144: 04017221. doi: 10.1061/(ASCE)ST.1943-541X.0001953
![]() |
[33] |
Attary N, van de Lindt JW, Unnikrishnan VU, et al. (2017) Methodology for Development of Physics-Based Tsunami Fragilities. J Struct Eng 143: 12. doi: 10.1061/(ASCE)ST.1943-541X.0001715
![]() |
[34] |
Dias WPS, Yapa HD, Peiris LMN (2009) Tsunami vulnerability functions from field surveys and Monte Carlo simulation. Civ Eng Environ Syst 26: 181-194. doi: 10.1080/10286600802435918
![]() |
[35] | Karafagka S, Fotopoulou S, Pitilakis K (2016) Tsunami fragility curves for seaport buildings and warehouses. Nat Hazards Infrastruct, 28-30. |
[36] | Kircher CA, Bouabid J (2014) New Building Damage and Loss Functions for Tsunami. Tenth U S Nat Conf Earthquake Eng. |
[37] |
Nanayakkara KIU, Dias WPS (2015) Fragility curves for structures under tsunami loading. Nat Hazards 80: 471-486. doi: 10.1007/s11069-015-1978-1
![]() |
[38] |
Park S, van de Lindt JW, Cox D, et al. (2012) Successive Earthquake-Tsunami Analysis to Develop Collapse Fragilities. J Earthquake Eng 16: 851-863. doi: 10.1080/13632469.2012.685209
![]() |
[39] |
Petrone C, Rossetto T, Goda K (2017) Fragility assessment of a RC structure under tsunami actions via nonlinear static and dynamic analyses. Eng Struct 136: 36-53. doi: 10.1016/j.engstruct.2017.01.013
![]() |
[40] |
Medina S, Lizarazo-Marriaga J, Estrada M, et al. (2019) Tsunami analytical fragility curves for the Colombian Pacific Coast: A reinforced concrete building example. Eng Struct 196: 109309. doi: 10.1016/j.engstruct.2019.109309
![]() |
[41] |
Belliazzi S, Lignola GP, Di Ludovico M, et al. (2021) Preliminary tsunami analytical fragility functions proposal for Italian coastal residential masonry buildings. Structures 31: 68-79. doi: 10.1016/j.istruc.2021.01.059
![]() |
[42] | Ferrotto MF, Cavaleri L (2021) Masonry structures: A proposal of analytical generation of fragility functions for tsunami impact—Application to the Mediterranean coasts. Eng Struct 242. |
[43] | INEC, INEC presenta sus proyecciones poblacionales cantonales 2010-2020. 2010. Available from: https://www.ecuadorencifras.gob.ec/inec-presenta-sus-proyecciones-poblacionales-cantonales/. |
[44] |
Prezzi C, Silberleigt V (2015) Seismic hazards along Ecuador, Perú and northern Chile (South America). Nat Hazards 79: 1159-1175. doi: 10.1007/s11069-015-1900-x
![]() |
[45] |
Yoshimoto M, Kumagai H, Acero W, et al. (2017) Depth-dependent rupture mode along the Ecuador-Colombia subduction zone. Geophys Res Lett 44: 2203-2210. doi: 10.1002/2016GL071929
![]() |
[46] | Social, Sistema de Indicadores Sociales del Ecuador. 2019. Available from: http://www.siise.gob.ec/siiseweb/. |
[47] | Poulard S (2012) REDATAM +SP, CEPAL. Available from: https://www.cepal.org/es/temas/redatam/acerca-redatam. |
[48] |
Luque A, Edwards GAS, Lalande C (2012) The local governance of climate change: new tools to respond to old limitations in Esmeraldas, Ecuador. Local Environ 18: 738-751. doi: 10.1080/13549839.2012.716414
![]() |
[49] | MINTUR, Boletín de Estadísticas-Turísticas 2012-2016. 2017. Ministerio de Turismo del Ecuador. |
[50] | INEC, Presentacion Esmeraldas Censo económico 2011. 2011. Available from: https://anda.inec.gob.ec/anda/index.php/catalog/586/related_materials. |
[51] | Valdivia G (2018) "Wagering Life" in the Petro-City: Embodied Ecologies of Oil Flow, Capitalism, and Justice in Esmeraldas, Ecuador. Ann Am Assoc Geogr 108: 549-557. |
[52] | Bank EC (2018) Reporte de Sector Petrolero. Ⅱ trimestre de 2018, Publicaciones económicas, Ecuadorian Central Bank: Ecuador. |
[53] | Bank EC (2017) Reporte de Sector Petrolero. IV trimestre de 2017, Publicaciones económicas, Ecuadorian Central Bank: Ecuador. |
[54] | MINTUR (2018) Perfiles de Turismo Internacional 2017,152. |
[55] | INEC (2010) CPV Interactivo para Investigadores y Académicos: Glosario de Términos Censales*, INEC, 15. |
[56] |
Dominey-Howes D, Papathoma M (2007) Validating a Tsunami Vulnerability Assessment Model (the PTVA Model) Using Field Data from the 2004 Indian Ocean Tsunami. Nat Hazards 40: 113-136. doi: 10.1007/s11069-006-0007-9
![]() |
[57] | Turismo MD (2017) Boletín de Estadísticas Turisticas, 2012-2016. 179. |
[58] |
Kendrick E, Bevis M, Smalley Jr R, et al. (2003) The Nazca-South America Euler vector and its rate of change. J South Am Earth Sci 16: 125-131. doi: 10.1016/S0895-9811(03)00028-2
![]() |
[59] |
Gailler A, Charvis P, Flueh ER (2007) Segmentation of the Nazca and South American plates along the Ecuador subduction zone from wide angle seismic profiles. Earth Planet Sci Lett 260: 444-464. doi: 10.1016/j.epsl.2007.05.045
![]() |
[60] |
DeMets C, Gordon RG, Argus DF, et al. (1990) Current plate motions. Geophys J Int 101: 425-478. doi: 10.1111/j.1365-246X.1990.tb06579.x
![]() |
[61] |
Ruff LJ, Kanamori H (1980) Seismicity and the subduction process. Phys Earth Planet Inter 23: 240-252. doi: 10.1016/0031-9201(80)90117-X
![]() |
[62] |
Collot JY, Charvis P, Gutscher MA, et al. (2002) Exploring the Ecuador-Colombia Active Margin and Interplate Seismogenic Zone. EOS 23: 185-190. doi: 10.1029/2002EO000120
![]() |
[63] |
Proust JN, Martillo C, Michaud F, et al. (2016) Subduction of seafloor asperities revealed by a detailed stratigraphic analysis of the active margin shelf sediments of Central Ecuador. Mar Geol 380: 345-362. doi: 10.1016/j.margeo.2016.03.014
![]() |
[64] | Collot JY, Michaud F, Alvarado A, et al. (2009) Visión general de la morfologia submarina del margen convergente de Ecuador-Sur de Colombia: Implicaciones sobre la transferencia de masa y la edad de la subducción de la Cordillera de Carnegie. 47-74. |
[65] |
Bilek SL (2010) Invited review paper: Seismicity along the South American subduction zone: Review of large earthquakes, tsunamis, and subduction zone complexity. Tectonophysics 495: 2-14. doi: 10.1016/j.tecto.2009.02.037
![]() |
[66] |
Parra H, Benito MB, Gaspar-Escribano JM (2016) Seismic hazard assessment in continental Ecuador Parra. Bull Earthquake Eng 14: 2129-2159. doi: 10.1007/s10518-016-9906-7
![]() |
[67] | Swenson J, Beck S (1996) Historical 1942 Ecuador and 1942 Peru Subduction Earthquakes, and Earthquake cycles along Colombia-Ecuador and Peru Subduction segments. Pageoph 146: 35. |
[68] | Arreaga P (2004) Análisis de riesgo por tsunamis en la ciudad de Esmeraldas, Escuela Superior Politécnica del Litoral: Guayaquil-Ecuador, 320. |
[69] | Segovia M (2001) El sismo de Bahía del 4 de agosto de 1998: Caracterización del mecanismo de ruptura y análisis de la sismicidad en la zona costera. Titulo de Ingeniera Geologia tesis: Escuela Politecnica Nacional, Quito, Ecuador. Available from: https://bibdigital.epn.edu.ec/. |
[70] |
Nocquet JM, Jarrin P, Vallee M, et al. (2016) Supercycle at the Ecuadorian subduction zone revealed after the 2016 Pedernales earthquake. Nat Geosci 10: 145-149. doi: 10.1038/ngeo2864
![]() |
[71] |
Vaca S, Vallee M, Nocquet JM, et al. (2018) Recurrent slow slip events as a barrier to the northward rupture propagation of the 2016 Pedernales earthquake (Central Ecuador). Tectonophysics 724-725: 80-92. doi: 10.1016/j.tecto.2017.12.012
![]() |
[72] | Contreras M (2013) Chronology of tsunamis in Ecuador from 1586 to 2012. Revista La Técnica 11: 50-59. |
[73] |
Papazachos BC, Scordilis ME, Panagiotopoulos DG, et al. (2001) Global Relations Between Seismic Fault Parameteres and Moment Magnitude of Earthquakes. Bull Geol Soc Greece 36: 1482-1489. doi: 10.12681/bgsg.16538
![]() |
[74] |
Jimenez C, Moggiano N, Mas E, et al. (2013) Seismic Source of 1746 Callao Earthquake from Tsunami Numerical Modeling. J Disaster Res 8: 266-273. doi: 10.20965/jdr.2013.p0266
![]() |
[75] | Imamura F, Yalciner CA, Ozyurt G (1995) Tsunami Modelling Manual. IOC Int Train Course Tsunami Numer Modell, 58. |
[76] |
Okada Y (1985) Surface deformation due to shear and tensile faults in a half-space. Bull Seismol Soc Am 75: 1135-1154. doi: 10.1785/BSSA0750041135
![]() |
[77] |
Tanioka Y, Satake K (1996) Tsunami generation by horizontal displacement of ocean bottom. Geophys Res Lett 15: 861-864. doi: 10.1029/96GL00736
![]() |
[78] | Arreaga P (2015) Tsunami Inundation Modeling and Hazard Mapping of the South Coast of Ecuador. National Graduate Institute for Policy Studies. Building Reserch Institute, 73. |
[79] | Charvet I, Macabuag J, Rossetto T (2017) Estimating Tsunami-Induced Building Damage through Fragility Functions: Critical Review and Research Needs. Front Built Environ 3. |
[80] |
Suppasri A, Mas E, Charvet I, et al. (2013) Building damage characteristics based on surveyed data and fragility curves of the 2011 Great East Japan tsunami. Nat Hazards 66: 319-341. doi: 10.1007/s11069-012-0487-8
![]() |
[81] |
Koshimura S, Oie T, Yanagisawa H, et al. (2009) Developing Fragility Functions for Tsunami Damage Estimation Using Numerical Model and Post-Tsunami Data from Banda Aceh, Indonesia. Coastal Eng J 51: 243-273. doi: 10.1142/S0578563409002004
![]() |
[82] | UNDRR. Terminology|PreventionWeb.net. 2020. Available from: https://digitallibrary.un.org/record/857333. |
[83] | Shi P, Kasperson R (2015) World Atlas of Natural Disaster Risk, BNUP-Springer, 381. |
[84] |
Rehman, K, Cho YS (2016) Building Damage Assessment Using Scenario Based Tsunami Numerical Analysis and Fragility Curves. Water 8: 109. doi: 10.3390/w8030109
![]() |
[85] | Suppasri A, Mas E, Koshimura S, et al. (2012) Developing Tsunami Fragility Curves from the Surveyed Data of the 2011 Great East Japan Tsunami in Sendai and Ishinomaki Plains. Coastal Eng J 54: 1250008. |
[86] | SGR (2012) Tsunami Inundation Charts. Available from: http://www.gestionderiesgos.gob.ec/wp-content/uploads/downloads/2013/12/Muisne.pdf. |
[87] | Cruz M, Acosta M, Vasquez N (2005) Riesgos por Tsunami en la Costa ecuatoriana, Nacional del Ecuador del Instituto Panamericano de Geografia e Historia, Organizacion de Estados Americanos, OEA. Publicaciones de geofísica. |
[88] | INEC (2010) Censo de Población y Vivienda. Available from: https://www.ecuadorencifras.gob.ec/base-de-datos-censo-de-poblacion-y-vivienda-2010/. |
[89] |
Vera San Martín T, Rodriguez Rosado G, Arreaga Vargas P, et al. (2018) Population and building vulnerability assessment by possible worst-case tsunami scenarios in Salinas, Ecuador. Nat Hazards 93: 275-297. doi: 10.1007/s11069-018-3300-5
![]() |
1. | Kottakkaran Sooppy Nisar, Muhammad Farman, Mahmoud Abdel-Aty, Chokalingam Ravichandran, A review of fractional order epidemic models for life sciences problems: Past, present and future, 2024, 95, 11100168, 283, 10.1016/j.aej.2024.03.059 | |
2. | Kaushik Dehingia, Salah Boulaaras, Evren Hinçal, Kamyar Hosseini, Thabet Abdeljawad, M.S. Osman, On the dynamics of a financial system with the effect financial information, 2024, 106, 11100168, 438, 10.1016/j.aej.2024.08.049 | |
3. | Rituparna Bhattacharyya, Brajesh Kumar Jha, Computational Fuzzy Modelling Approach to Analyze Neuronal Calcium Dynamics With Intracellular Fluxes, 2024, 1085-9195, 10.1007/s12013-024-01541-0 | |
4. | Parvaiz Ahmad Naik, Anum Zehra, Muhammad Farman, Aamir Shehzad, Sundas Shahzeen, Zhengxin Huang, Forecasting and dynamical modeling of reversible enzymatic reactions with a hybrid proportional fractional derivative, 2024, 11, 2296-424X, 10.3389/fphy.2023.1307307 | |
5. | Kottakkaran Sooppy Nisar, Muhammad Farman, Anum Zehra, Evren Hincal, Numerical and analytical study of fractional order tumor model through modeling with treatment of chemotherapy, 2024, 0228-6203, 1, 10.1080/02286203.2024.2327659 | |
6. | Muhammad Farman, Rabia Sarwar, Evren Hincal, Dumitru Baleanu, Kottakkaran Sooppy Nisar, Aceng Sambas, Muhammad Manan Akram, Dynamical Evaluation of the Mittag-Leffler Properties-Based Marine Ecosystem Model, 2025, 2509-9426, 10.1007/s41748-025-00617-y | |
7. | Md. Saifur Rahman, Rehena Nasrin, Md. Haider Ali Biswas, Mathematical Modeling and Analysis of Human-to-Human Transmitted Viral Encephalitis, 2025, 13, 2227-7390, 1809, 10.3390/math13111809 |
Parameter | Value |
Discount Factor (δ) | 0.95 |
Network soft update parameters (σ) | 0.0005 |
Experience replay pool size | 105 |
Numbers of hidden layers | 2 |
First hidden layer size | 40 |
Second hidden layer size | 30 |
Initial state | Control method | DDPG-MPC-IBVS | MPC-IBVS-A | MPC-IBVS-B |
p1 | Settling time (s) | 5.24 | 9.98 | 8.13 |
Image deviation (pixels) | 4.98 | 10.21 | 10.5 | |
Control overshoot | 12% | 20% | 19% | |
p2 | Settling time (s) | 4.89 | 8.07 | 10.1 |
Image deviation (pixels) | 5.1 | 11.26 | 11.18 | |
Control overshoot | 12% | 18% | 20% |
Control method | DDPG-MPC IBVS | Q-learning-MPC IBVS |
Success rate | 100% | 100% |
Average overshoot | 12% | 14% |
Average settling time (s) | 5.12 | 7.39 |
Average image deviation (pixels) | 5.06 | 5.2 |
Parameter | Value |
Discount Factor (δ) | 0.95 |
Network soft update parameters (σ) | 0.0005 |
Experience replay pool size | 105 |
Numbers of hidden layers | 2 |
First hidden layer size | 40 |
Second hidden layer size | 30 |
Initial state | Control method | DDPG-MPC-IBVS | MPC-IBVS-A | MPC-IBVS-B |
p1 | Settling time (s) | 5.24 | 9.98 | 8.13 |
Image deviation (pixels) | 4.98 | 10.21 | 10.5 | |
Control overshoot | 12% | 20% | 19% | |
p2 | Settling time (s) | 4.89 | 8.07 | 10.1 |
Image deviation (pixels) | 5.1 | 11.26 | 11.18 | |
Control overshoot | 12% | 18% | 20% |
Control method | DDPG-MPC IBVS | Q-learning-MPC IBVS |
Success rate | 100% | 100% |
Average overshoot | 12% | 14% |
Average settling time (s) | 5.12 | 7.39 |
Average image deviation (pixels) | 5.06 | 5.2 |