Real-time and accurate network-wide traffic volume estimation/detection is an essential part of urban transport system planning and management. As it is impractical to install detectors on every road segment of the city network, methods on the network-wide flow estimation based on limited detector data are of considerable significance. However, when the plan of detector deployment is uncertain, existing methods are unsuitable to be directly used. In this study, a transfer component analysis (TCA)-based network-wide volume estimation model, considering the different traffic volume distributions of road segments and transforming traffic features into common data space, is proposed. Moreover, this study applied taxi GPS (global positioning system) data and cellular signaling data with the same spatio-temporal coverage to improve feature extraction. In numerical experiments, the robustness and stability of the proposed network-wide estimation method outperformed other baselines in the two subnetworks selected from the urban centers and suburbs.
Citation: Jiping Xing, Yunchi Wu, Di Huang, Xin Liu. Transfer learning for robust urban network-wide traffic volume estimation with uncertain detector deployment scheme[J]. Electronic Research Archive, 2023, 31(1): 207-228. doi: 10.3934/era.2023011
[1] | Kailash C. Misra, Sutida Patlertsin, Suchada Pongprasert, Thitarie Rungratgasame . On derivations of Leibniz algebras. Electronic Research Archive, 2024, 32(7): 4715-4722. doi: 10.3934/era.2024214 |
[2] | Yizheng Li, Dingguo Wang . Lie algebras with differential operators of any weights. Electronic Research Archive, 2023, 31(3): 1195-1211. doi: 10.3934/era.2023061 |
[3] | Shanshan Liu, Abdenacer Makhlouf, Lina Song . The full cohomology, abelian extensions and formal deformations of Hom-pre-Lie algebras. Electronic Research Archive, 2022, 30(8): 2748-2773. doi: 10.3934/era.2022141 |
[4] | Xueru Wu, Yao Ma, Liangyun Chen . Abelian extensions of Lie triple systems with derivations. Electronic Research Archive, 2022, 30(3): 1087-1103. doi: 10.3934/era.2022058 |
[5] | Pengliang Xu, Xiaomin Tang . Graded post-Lie algebra structures and homogeneous Rota-Baxter operators on the Schrödinger-Virasoro algebra. Electronic Research Archive, 2021, 29(4): 2771-2789. doi: 10.3934/era.2021013 |
[6] | Neşet Deniz Turgay . On the mod p Steenrod algebra and the Leibniz-Hopf algebra. Electronic Research Archive, 2020, 28(2): 951-959. doi: 10.3934/era.2020050 |
[7] | Hongliang Chang, Yin Chen, Runxuan Zhang . A generalization on derivations of Lie algebras. Electronic Research Archive, 2021, 29(3): 2457-2473. doi: 10.3934/era.2020124 |
[8] | Margarida Camarinha . A natural 4th-order generalization of the geodesic problem. Electronic Research Archive, 2024, 32(5): 3396-3412. doi: 10.3934/era.2024157 |
[9] | Baoling Guan, Xinxin Tian, Lijun Tian . Induced 3-Hom-Lie superalgebras. Electronic Research Archive, 2023, 31(8): 4637-4651. doi: 10.3934/era.2023237 |
[10] | Jinguo Jiang . Algebraic Schouten solitons associated to the Bott connection on three-dimensional Lorentzian Lie groups. Electronic Research Archive, 2025, 33(1): 327-352. doi: 10.3934/era.2025017 |
Real-time and accurate network-wide traffic volume estimation/detection is an essential part of urban transport system planning and management. As it is impractical to install detectors on every road segment of the city network, methods on the network-wide flow estimation based on limited detector data are of considerable significance. However, when the plan of detector deployment is uncertain, existing methods are unsuitable to be directly used. In this study, a transfer component analysis (TCA)-based network-wide volume estimation model, considering the different traffic volume distributions of road segments and transforming traffic features into common data space, is proposed. Moreover, this study applied taxi GPS (global positioning system) data and cellular signaling data with the same spatio-temporal coverage to improve feature extraction. In numerical experiments, the robustness and stability of the proposed network-wide estimation method outperformed other baselines in the two subnetworks selected from the urban centers and suburbs.
The concept of embedded tensors initially emerged in the research on gauged supergravity theory [1]. Using embedding tensors, the N=8 supersymmetric gauge theories as well as the Bagger-Lambert theory of multiple M2-branes were investigated in [2]. See [3,4,5] and the references therein for a great deal of literature on embedding tensors and related tensor hierarchies. In [6], the authors first observed the mathematical essence behind the embedding tensor and proved that the embedding tensor naturally produced Leibniz algebra. In the application of physics, they observed that in the construction of the corresponding gauge theory, they focused more on Leibniz algebra than on embedding tensor.
In [7], Sheng et al. considered cohomology, deformations, and homotopy theory for embedding tensors and Lie-Leibniz triples. Later on, the deformation and cohomology theory of embedding tensors on 3-Lie algebras were extensively elaborated in [8]. Tang and Sheng [9] first proposed the concept of a nonabelian embedding tensor on Lie algebras, which is a nonabelian generalization of the embedding tensors, and gave the algebraic structures behind the nonabelian embedding tensors as Leibniz-Lie algebras. This generalization for embedding tensors on associative algebras has been previously explored in [10,11], where they are referred to as average operators with any nonzero weights. Moreover, the nonabelian embedding tensor on Lie algebras has been extended to the Hom setting in [12].
On the other hand, Filippov [13] first introduced the concepts of 3-Lie algebras and, more generally, n-Lie algebras (also called Filippov algebras). Over recent years, the study and application of 3-Lie algebras have expanded significantly across the realms of mathematics and physics, including string theory, Nambu mechanics [14], and M2-branes [15,16]. Further research on 3-Lie algebras could be found in [17,18,19] and references cited therein.
Drawing inspiration from Tang and Sheng's [9] terminology of nonabelian embedding tensors and recognizing the significance of 3-Lie algebras, cohomology, and deformation theories, this paper primarily investigates the nonabelian embedding tensors on 3-Lie algebras, along with their fundamental algebraic structures, cohomology, and deformations.
This paper is organized as follows: Section 2 first recalls some basic notions of 3-Lie algebras and 3-Leibniz algebras. Then we introduce the coherent action of a 3-Lie algebra on another 3-Lie algebra and the notion of nonabelian embedding tensors on 3-Lie algebras with respect to a coherent action. In Section 3, the concept of 3-Leibniz-Lie algebra is presented as the fundamental algebraic structure for a nonabelian embedding tensor on the 3-Lie algebra. Naturally, a 3-Leibniz-Lie algebra induces a 3-Leibniz algebra. Subsequently, we study 3-Leibniz-Lie algebras induced by Leibniz-Lie algebras. In Section 4, the cohomology theory of nonabelian embedding tensors on 3-Lie algebras is introduced. As an application, we characterize the infinitesimal deformation using the first cohomology group.
All vector spaces and algebras considered in this paper are on the field K with the characteristic of 0.
This section recalls some basic notions of 3-Lie algebras and 3-Leibniz algebras. After that, we introduce the coherent action of a 3-Lie algebra on another 3-Lie algebra, and we introduce the concept of nonabelian embedding tensors on 3-Lie algebras by its coherent action as a nonabelian generalization of embedding tensors on 3-Lie algebras [8].
Definition 2.1. (see [13]) A 3-Lie algebra is a pair (L,[−,−,−]L) consisting of a vector space L and a skew-symmetric ternary operation [−,−,−]L:∧3L→L such that
[l1,l2,[l3,l4,l5]L]L=[[l1,l2,l3]L,l4,l5]L+[l3,[l1,l2,l4]L,l5]L+[l3,l4,[l1,l2,l5]L]L, | (2.1) |
for all li∈L,1≤i≤5.
A homomorphism between two 3-Lie algebras (L1,[−,−,−]L1) and (L2,[−,−,−]L2) is a linear map f:L1→L2 that satisfies f([l1,l2,l3]L1)=[f(l1),f(l2),f(l3)]L2, for all l1,l2,l3∈L1.
Definition 2.2. 1) (see [20]) A representation of a 3-Lie algebra (L,[−,−,−]L) on a vector space H is a skew-symmetric linear map ρ:∧2L→End(H), such that
ρ([l1,l2,l3]L,l4)=ρ(l2,l3)ρ(l1,l4)+ρ(l3,l1)ρ(l2,l4)+ρ(l1,l2)ρ(l3,l4), | (2.2) |
ρ(l1,l2)ρ(l3,l4)=ρ(l3,l4)ρ(l1,l2)+ρ([l1,l2,l3]L,l4)+ρ(l3,[l1,l2,l4]L), | (2.3) |
for all l1,l2,l3,l4∈L. We also denote a representation of L on H by (H;ρ).
2) A coherent action of a 3-Lie algebra (L,[−,−,−]L) on another 3-Lie algebra (H,[−,−,−]H) is defined by a skew-symmetric linear map ρ:∧2L→Der(H) that satisfies Eqs (2.2) and (2.3), along with the condition that
[ρ(l1,l2)h1,h2,h3]H=0, | (2.4) |
for all l1,l2∈L and h1,h2,h3∈H. We denote a coherent action of L on H by (H,[−,−,−]H;ρ†).
Note that Eq (2.4) and ρ(l1,l2)∈Der(H) imply that
ρ(l1,l2)[h1,h2,h3]H=0. | (2.5) |
Example 2.3. Let (H,[−,−,−]H) be a 3-Lie algebra. Define ad:∧2H→Der(H) by
ad(h1,h2)h:=[h1,h2,h]H, for all h1,h2,h∈H. |
Then (H;ad) is a representation of (H,[−,−,−]H), which is called the adjoint representation. Furthermore, if the ad satisfies
[ad(h1,h2)h′1,h′2,h′3]H=0, for allh′1,h′2,h′3∈H, |
then (H,[−,−,−]H;ad†) is a coherent adjoint action of (H,[−,−,−]H).
Definition 2.4. (see [21]) A 3-Leibniz algebra is a vector space L together with a ternary operation [−,−,−]L:L⊗L⊗L→L such that
[l1,l2,[l3,l4,l5]L]L=[[l1,l2,l3]L,l4,l5]L+[l3,[l1,l2,l4]L,l5]L+[l3,l4,[l1,l2,l5]L]L, |
for all li∈L,1≤i≤5.
Proposition 2.5. Let (L,[−,−,−]L) and (H,[−,−,−]H) be two 3-Lie algebras, and let ρ be a coherent action of L on H. Then, L⊕H is a 3-Leibniz algebra under the following map:
[l1+h1,l2+h2,l3+h3]ρ:=[l1,l2,l3]L+ρ(l1,l2)h3+[h1,h2,h3]H, |
for all l1,l2,l3∈L and h1,h2,h3∈H. This 3-Leibniz algebra (L⊕H,[−,−,−]ρ) is called the nonabelian hemisemidirect product 3-Leibniz algebra, which is denoted by L⋉ρH.
Proof. For any l1,l2,l3,l4,l5∈L and h1,h2,h3,h4,h5∈H, by Eqs (2.1)–(2.5), we have
[l1+h1,l2+h2,[l3+h3,l4+h4,l5+h5]ρ]ρ−[[l1+h1,l2+h2,l3+h3]ρ,l4+h4,l5+h5]ρ−[l3+h3,[l1+h1,l2+h2,l4+h4]ρ,l5+h5]ρ−[l3+h3,l4+h4,[l1+h1,l2+h2,l5+h5]ρ]ρ=[l1,l2,[l3,l4,l5]L]L+ρ(l1,l2)ρ(l3,l4)h5+ρ(l1,l2)[h3,h4,h5]H+[h1,h2,ρ(l3,l4)h5]H+[h1,h2,[h3,h4,h5]H]H−[[l1,l2,l3]L,l4,l5]L−ρ([l1,l2,l3]L,l4)h5−[ρ(l1,l2)h3,h4,h5]H−[[h1,h2,h3]H,h4,h5]H−[l3,[l1,l2,l4]L,l5]L−ρ(l3,[l1,l2,l4]L)h5−[h3,ρ(l1,l2)h4,h5]H−[h3,[h1,h2,h4]H,h5]H−[l3,l4,[l1,l2,l5]L]L−ρ(l3,l4)ρ(l1,l2)h5−ρ(l3,l4)[h1,h2,h5]H−[h3,h4,ρ(l1,l2)h5]H−[h3,h4,[h1,h2,h5]H]H=[h1,h2,ρ(l3,l4)h5]H−ρ(l3,l4)[h1,h2,h5]H=0. |
Thus, (L⊕H,[−,−,−]ρ) is a 3-Leibniz algebra.
Definition 2.6. 1) A nonabelian embedding tensor on a 3-algebra (L,[−,−,−]L) with respect to a coherent action (H,[−,−,−]H;ρ†) is a linear map Λ:H→L that satisfies the following equation:
[Λh1,Λh2,Λh3]L=Λ(ρ(Λh1,Λh2)h3+[h1,h2,h3]H), | (2.6) |
for all h1,h2,h3∈H.
2) A nonabelian embedding tensor 3-Lie algebra is a triple (H,L,Λ) consisting of a 3-Lie algebra (L,[−,−,−]L), a coherent action (H,[−,−,−]H;ρ†) of L and a nonabelian embedding tensor Λ:H→L. We denote a nonabelian embedding tensor 3-Lie algebra (H,L,Λ) by the notation HΛ⟶L.
3) Let HΛ1⟶L and HΛ2⟶L be two nonabelian embedding tensor 3-Lie algebras. Then, a homomorphism from HΛ1⟶L to HΛ2⟶L consists of two 3-Lie algebras homomorphisms fL:L→L and fH:H→H, which satisfy the following equations:
Λ2∘fH=fL∘Λ1, | (2.7) |
fH(ρ(l1,l2)h)=ρ(fL(l1),fL(l2))fH(h), | (2.8) |
for all l1,l2∈L and h∈H. Furthermore, if fL and fH are nondegenerate, (fL,fH) is called an isomorphism from HΛ1⟶L to HΛ2⟶L.
Remark 2.7. If (H,[−,−,−]H) is an abelian 3-Lie algebra, then we can get that Λ is an embedding tensor on 3-Lie algebra (see [8]). In addition, If ρ=0, then Λ is a 3-Lie algebra homomorphism from H to L.
Example 2.8. Let H be a 4-dimensional linear space spanned by α1,α2,α3 and α4. We define a skew-symmetric ternary operation [−,−,−]H:∧3H→H by
[α1,α2,α3]H=α4. |
Then (H,[−,−,−]H) is a 3-Lie algebra. It is obvious that (H,[−,−,−]H;ad†) is a coherent adjoint action of (H,[−,−,−]H). Moreover,
Λ=(1000010000000000) |
is a nonabelian embedding tensor on (H,[−,−,−]H).
Next, we use graphs to describe nonabelian embedding tensors on 3-Lie algebras.
Theorem 2.9. A linear map Λ:H→L is a nonabelian embedding tensor on a 3-Lie algebra (L,[−,−,−]L) with respect to the coherent action (H,[−,−,−]H;ρ†) if and only if the graph Gr(Λ)={Λh+h|h∈H} forms a subalgebra of the nonabelian hemisemidirect product 3-Leibniz algebra L⋉ρH.
Proof. Let Λ:H→L be a linear map. Then, for any h1,h2,h3∈H, we have
[Λh1+h1,Λh2+h2,Λh3+h3]ρ=[Λh1,Λh2,Λh3]L+ρ(Λh1,Λh2)h3+[h1,h2,h3]H, |
Thus, the graph Gr(Λ)={Λh+h|h∈H} is a subalgebra of the nonabelian hemisemidirect product 3-Leibniz algebra L⋉ρH if and only if Λ satisfies Eq (2.6), which implies that Λ is a nonabelian embedding tensor on L with respect to the coherent action (H,[−,−,−]H;ρ†).
Because H and Gr(Λ) are isomorphic as linear spaces, there is an induced 3-Leibniz algebra structure on H.
Corollary 2.10. Let HΛ⟶L be a nonabelian embedding tensor 3-Lie algebra. If a linear map [−,−,−]Λ:∧3H→H is given by
[h1,h2,h3]Λ=ρ(Λh1,Λh2)h3+[h1,h2,h3]H, | (2.9) |
for all h1,h2,h3∈H, then (H,[−,−,−]Λ) is a 3-Leibniz algebra. Moreover, Λ is a homomorphism from the 3-Leibniz algebra (H,[−,−,−]Λ) to the 3-Lie algebra (L,[−,−,−]L). This 3-Leibniz algebra (H,[−,−,−]Λ) is called the descendent 3-Leibniz algebra.
Proposition 2.11. Let (fL,fH) be a homomorphism from HΛ1⟶L to HΛ2⟶L. Then fH is a homomorphism of descendent 3-Leibniz algebra from (H,[−,−,−]Λ1) to (H,[−,−,−]Λ2).
Proof. For any h1,h2,h3∈H, by Eqs (2.7)–(2.9), we have
fH([h1,h2,h3]Λ1)=fH(ρ(Λ1h1,Λ1h2)h3+[h1,h2,h3]H)=ρ(fL(Λ1h1),fL(Λ1h2))fH(h3)+fH([h1,h2,h3]H)=ρ(Λ2fL(h1),Λ2fL(h2))fH(h3)+[fH(h1),fH(h2),fH(h3)]H=[fH(h1),fH(h2),fH(h3)]Λ2. |
The proof is finished.
In this section, we present the concept of the 3-Leibniz-Lie algebra, which serves as the fundamental algebraic framework for the nonabelian embedding tensor 3-Lie algebra. Then we study 3-Leibniz-Lie algebras induced by Leibniz-Lie algebras.
Definition 3.1. A 3-Leibniz-Lie algebra (H,[−,−,−]H,{−,−,−}H) encompasses a 3-Lie algebra (H,[−,−,−]H) and a ternary operation {−,−,−}H:∧3H→H, which satisfies the following equations:
{h1,h2,h3}H=−{h2,h1,h3}H, | (3.1) |
{h1,h2,{h3,h4,h5}H}H={{h1,h2,h3}H,h4,h5}H+{h3,{h1,h2,h4}H,h5}H+{h3,h4,{h1,h2,h5}H}H+{[h1,h2,h3]H,h4,h5}H+{h3,[h1,h2,h4]H,h5}H, | (3.2) |
{h1,h2,[h3,h4,h5]H}H=[{h1,h2,h3}H,h4,h5]H=0, | (3.3) |
for all h1,h2,h3,h4,h5∈H.
A homomorphism between two 3-Leibniz-Lie algebras (H1,[−,−,−]H1,{−,−,−}H1) and (H2,[−,−,−]H2,{−,−,−}H2) is a 3-Lie algebra homomorphism f:(H1,[−,−,−]H1)→(H2,[−,−,−]H2) such that f({h1,h2,h3}H1)={f(h1),f(h2),f(h3)}H2, for all h1,h2,h3∈H1.
Remark 3.2. A 3-Lie algebra (H,[−,−,−]H) naturally constitutes a 3-Leibniz-Lie algebra provided that the underlying ternary operation {h1,h2,h3}H=0, for all h1,h2,h3∈H.
Example 3.3. Let (H,[−,−,−]H) be a 4-dimensional 3-Lie algebra given in Example 2.8. We define a nonzero operation {−,−,−}H:∧3H→H by
{α1,α2,α3}H=−{α2,α1,α3}H=α4. |
Then (H,[−,−,−]H,{−,−,−}H) is a 3-Leibniz-Lie algebra.
The subsequent theorem demonstrates that a 3-Leibniz-Lie algebra inherently gives rise to a 3-Leibniz algebra.
Theorem 3.4. Let (H,[−,−,−]H,{−,−,−}H) be a 3-Leibniz-Lie algebra. Then the ternary operation ⟨−,−,−⟩H:∧3H→H, defined as
⟨h1,h2,h3⟩H:=[h1,h2,h3]H+{h1,h2,h3}H, | (3.4) |
for all h1,h2,h3∈H, establishes a 3-Leibniz algebra structure on H. This structure is denoted by (H,⟨−,−,−⟩H) and is referred to as the subadjacent 3-Leibniz algebra.
Proof. For any h1,h2,h3,h4,h5∈H, according to (H,[−,−,−]H) is a 3-Lie algebra and Eqs (3.2)–(3.4), we have
⟨h1,h2,⟨h3,h4,h5⟩H⟩H−⟨⟨h1,h2,h3⟩H,h4,h5⟩H−⟨h3,⟨h1,h2,h4⟩H,h5⟩H−⟨h3,h4,⟨h1,h2,h5⟩H⟩H=[h1,h2,[h3,h4,h5]H]H+[h1,h2,{h3,h4,h5}H]H+{h1,h2,[h3,h4,h5]H}H+{h1,h2,{h3,h4,h5}H}H−[[h1,h2,h3]H,h4,h5]H−[{h1,h2,h3}H,h4,h5]H−{[h1,h2,h3]H,h4,h5}H−{{h1,h2,h3}H,h4,h5}H−[h3,[h1,h2,h4]H,h5]H−[h3,{h1,h2,h4}H,h5]H−{h3,[h1,h2,h4]H,h5}H−{h3,{h1,h2,h4}H,h5}H−[h3,h4,[h1,h2,h5]H]H−[h3,h4,{h1,h2,h5}H]H−{h3,h4,[h1,h2,h5]H}H−{h3,h4,{h1,h2,h5}H}H={h1,h2,{h3,h4,h5}H}H−{[h1,h2,h3]H,h4,h5}H−{{h1,h2,h3}H,h4,h5}H−{h3,[h1,h2,h4]H,h5}H−{h3,{h1,h2,h4}H,h5}H−{h3,h4,{h1,h2,h5}H}H=0. |
Hence, (H,⟨−,−,−⟩H) is a 3-Leibniz algebra.
The following theorem shows that a nonabelian embedding tensor 3-Lie algebra induces a 3-Leibniz-Lie algebra.
Theorem 3.5. Let HΛ⟶L be a nonabelian embedding tensor 3-Lie algebra. Then (H,[−,−,−]H,{−,−,−}Λ) is a 3-Leibniz-Lie algebra, where
{h1,h2,h3}Λ:=ρ(Λh1,Λh2)h3, | (3.5) |
for all h1,h2,h3∈H.
Proof. For any h1,h2,h3,h4,h5∈H, by Eqs (2.3), (2.6), and (3.5), we have
{h1,h2,h3}Λ=ρ(Λh1,Λh2)h3=−ρ(Λh2,Λh1)h3=−{h2,h1,h3}Λ,{{h1,h2,h3}Λ,h4,h5}Λ+{h3,{h1,h2,h4}Λ,h5}Λ+{h3,h4,{h1,h2,h5}Λ}Λ+{[h1,h2,h3]H,h4,h5}Λ+{h3,[h1,h2,h4]H,h5}Λ−{h1,h2,{h3,h4,h5}Λ}Λ=ρ(Λρ(Λh1,Λh2)h3,Λh4)h5+ρ(Λh3,Λρ(Λh1,Λh2)h4)h5+ρ(Λh3,Λh4)ρ(Λh1,Λh2)h5+ρ(Λ[h1,h2,h3]H,Λh4)h5+ρ(Λh3,Λ[h1,h2,h4]H)h5−ρ(Λh1,Λh2)ρ(Λh3,Λh4)h5=ρ(Λρ(Λh1,Λh2)h3,Λh4)h5+ρ(Λh3,Λρ(Λh1,Λh2)h4)h5+ρ(Λh3,Λh4)ρ(Λh1,Λh2)h5+ρ([Λh1,Λh2,Λh3]L−Λρ(Λh1,Λh2)h3,Λh4)h5+ρ(Λh3,[Λh1,Λh2,Λh4]L−Λρ(Λh1,Λh2)h4)h5−ρ(Λh1,Λh2)ρ(Λh3,Λh4)h5=ρ(Λh3,Λh4)ρ(Λh1,Λh2)h5+ρ([Λh1,Λh2,Λh3]L,Λh4)h5+ρ(Λh3,[Λh1,Λh2,Λh4]L)h5−ρ(Λh1,Λh2)ρ(Λh3,Λh4)h5=0. |
Furthermore, by Eqs (2.4), (2.5), and (3.5), we have
[{h1,h2,h3}Λ,h4,h5]H=[ρ(Λh1,Λh2)h3,h4,h5]H=0,{h1,h2,[h3,h4,h5]H}Λ=ρ(Λh1,Λh2)[h3,h4,h5]H=0. |
Thus, (H,[−,−,−]H,{−,−,−}Λ) is a 3-Leibniz-Lie algebra.
Proposition 3.6. Let (fL,fH) be a homomorphism from HΛ1⟶L to HΛ2⟶L. Then fH is a homomorphism of 3-Leibniz-Lie algebras from (H,[−,−,−]H,{−,−,−}Λ1) to (H,[−,−,−]H,{−,−,−}Λ2).
Proof. For any h1,h2,h3∈H, by Eqs (2.7), (2.8), and (3.5), we have
fH({h1,h2,h3}Λ1)=fH(ρ(Λ1h1,Λ1h2)h3)=ρ(fL(Λ1h1),fL(Λ1h2))fH(h3)=ρ(Λ2fH(h1),Λ2fH(h2))fH(h3)={fH(h1),fH(h2),fH(h3)}Λ2. |
The proof is finished.
Motivated by the construction of 3-Lie algebras from Lie algebras [17], at the end of this section, we investigate 3-Leibniz-Lie algebras induced by Leibniz-Lie algebras.
Definition 3.7. (see [9]) A Leibniz-Lie algebra (H,[−,−]H,⊳) encompasses a Lie algebra (H,[−,−]H) and a binary operation ⊳:H⊗H→H, ensuring that
h1⊳(h2⊳h3)=(h1⊳h2)⊳h3+h2⊳(h1⊳h3)+[h1,h2]H⊳h3,h1⊳[h2,h3]H=[h1⊳h2,h3]H=0, |
for all h1,h2,h3∈H.
Theorem 3.8. Let (H,[−,−]H,⊳) be a Leibniz-Lie algebra, and let ς∈H∗ be a trace map, which is a linear map that satisfies the following conditions:
ς([h1,h2]H)=0,ς(h1⊳h2)=0,for allh1,h2∈H. |
Define two ternary operations by
[h1,h2,h3]Hς=ς(h1)[h2,h3]H+ς(h2)[h3,h1]H+ς(h3)[h1,h2]H,{h1,h2,h3}Hς=ς(h1)h2⊳h3−ς(h2)h1⊳h3,for allh1,h2,h3∈H. |
Then (H,[−,−,−]Hς,{−,−,−}Hς) is a 3-Leibniz-Lie algebra.
Proof. First, we know from [17] that (H,[−,−,−]Hς) is a 3-Lie algebra. Next, for any h1,h2,h3,h4,h5∈H, we have
{h1,h2,h3}Hς=ς(h1)h2⊳h3−ς(h2)h1⊳h3=−(ς(h2)h1⊳h3−ς(h1)h2⊳h3)=−{h2,h1,h3}Hς |
and
{{h1,h2,h3}Hς,h4,h5}Hς+{h3,{h1,h2,h4}Hς,h5}Hς+{h3,h4,{h1,h2,h5}Hς}Hς+{[h1,h2,h3]Hς,h4,h5}Hς+{h3,[h1,h2,h4]Hς,h5}Hς−{h1,h2,{h3,h4,h5}Hς}Hς=ς(h1)ς(h2⊳h3)h4⊳h5−ς(h4)ς(h1)(h2⊳h3)⊳h5−ς(h2)ς(h1⊳h3)h4⊳h5+ς(h4)ς(h2)(h1⊳h3)⊳h5+ς(h3)ς(h1)(h2⊳h4)⊳h5−ς(h1)ς(h2⊳h4)h3⊳h5−ς(h3)ς(h2)(h1⊳h4)⊳h5+ς(h2)ς(h1⊳h4)h3⊳h5+ς(h1)ς(h3)h4⊳(h2⊳h5)−ς(h1)ς(h4)h3⊳(h2⊳h5)−ς(h2)ς(h3)h4⊳(h1⊳h5)+ς(h2)ς(h4)h3⊳(h1⊳h5)+ς(h1)ς([h2,h3]H)h4⊳h5−ς(h4)ς(h1)[h2,h3]H⊳h5+ς(h2)ς([h3,h1]H)h4⊳h5−ς(h4)ς(h2)[h3,h1]H⊳h5+ς(h3)ς([h1,h2]H)h4⊳h5−ς(h4)ς(h3)[h1,h2]H⊳h5+ς(h3)ς(h1)[h2,h4]H⊳h5−ς(h1)ς([h2,h4]H)h3⊳h5+ς(h3)ς(h2)[h4,h1]H⊳h5−ς(h2)ς([h4,h1]H)h3⊳h5+ς(h3)ς(h4)[h1,h2]H⊳h5−ς(h4)ς([h1,h2]H)h3⊳h5−ς(h1)ς(h3)h2⊳(h4⊳h5)+ς(h2)ς(h3)h1⊳(h4⊳h5)+ς(h1)ς(h4)h2⊳(h3⊳h5)−ς(h2)ς(h4)h1⊳(h3⊳h5)=−ς(h4)ς(h1)(h2⊳h3)⊳h5+ς(h4)ς(h2)(h1⊳h3)⊳h5+ς(h3)ς(h1)(h2⊳h4)⊳h5−ς(h3)ς(h2)(h1⊳h4)⊳h5+ς(h1)ς(h3)h4⊳(h2⊳h5)−ς(h1)ς(h4)h3⊳(h2⊳h5)−ς(h2)ς(h3)h4⊳(h1⊳h5)+ς(h2)ς(h4)h3⊳(h1⊳h5)−ς(h4)ς(h1)[h2,h3]H⊳h5−ς(h4)ς(h2)[h3,h1]H⊳h5+ς(h3)ς(h1)[h2,h4]H⊳h5+ς(h3)ς(h2)[h4,h1]H⊳h5−ς(h1)ς(h3)h2⊳(h4⊳h5)+ς(h2)ς(h3)h1⊳(h4⊳h5)+ς(h1)ς(h4)h2⊳(h3⊳h5)−ς(h2)ς(h4)h1⊳(h3⊳h5)=0. |
Similarly, we obtain
{h1,h2,[h3,h4,h5]Hς}Hς=ς(h1)ς(h3)h2⊳[h4,h5]H−ς(h2)ς(h3)h1⊳[h4,h5]H+ς(h1)ς(h4)h2⊳[h5,h3]H−ς(h2)ς(h4)h1⊳[h5,h3]H+ς(h1)ς(h5)h2⊳[h3,h4]H−ς(h2)ς(h5)h1⊳[h3,h4]H=0 |
and
[{h1,h2,h3}Hς,h4,h5]Hς=ς(h1)ς(h2⊳h3)[h4,h5]H+ς(h4)ς(h1)[h5,h2⊳h3]H+ς(h5)ς(h1)[h2⊳h3,h4]H−ς(h2)ς(h1⊳h3)[h4,h5]H−ς(h4)ς(h2)[h5,h1⊳h3]H−ς(h5)ς(h2)[h1⊳h3,h4]H=0. |
Hence Eqs (3.1)–(3.3) hold and we complete the proof.
In this section, we revisit fundamental results pertaining to the representations and cohomologies of 3-Leibniz algebras. We construct a representation of the descendent 3-Leibniz algebra (H,[−,−,−]Λ) on the vector space L and define the cohomologies of a nonabelian embedding tensor on 3-Lie algebras. As an application, we characterize the infinitesimal deformation using the first cohomology group.
Definition 4.1. (see [22]) A representation of the 3-Leibniz algebra (H,[−,−,−]H) is a vector space V equipped with 3 actions
l:H⊗H⊗V→V,m:H⊗V⊗H→V,r:V⊗H⊗H→V, |
satisfying for any a1,a2,a3,a4,a5∈H and u∈V
l(a1,a2,l(a3,a4,u))=l([a1,a2,a3]H,a4,u)+l(a3,[a1,a2,a4]H,u)+l(a3,a4,l(a1,a2,u)), | (4.1) |
l(a1,a2,m(a3,u,a5))=m([a1,a2,a3]H,u,a5)+m(a3,l(a1,a2,u),a5)+m(a3,u,[a1,a2,a5]H), | (4.2) |
l(a1,a2,r(u,a4,a5))=r(l(a1,a2,u),a4,a5)+r(u,[a1,a2,a4]H,a5)+r(u,a4,[a1,a2,a5]H), | (4.3) |
m(a1,u,[a3,a4,a5]H)=r(m(a1,u,a3),a4,a5)+m(a3,m(a1,u,a4),a5)+l(a3,a4,m(a1,u,a5)), | (4.4) |
r(u,a2,[a3,a4,a5]H)=r(r(u,a2,a3),a4,a5)+m(a3,r(u,a2,a4),a5)+l(a3,a4,r(u,a2,a5)). | (4.5) |
For n≥1, denote the n-cochains of 3-Leibniz algebra (H,[−,−,−]H) with coefficients in a representation (V;l,m,r) by
Cn3Leib(H,V)=Hom(n−1⏞∧2H⊗⋯⊗∧2H⊗H,V). |
The coboundary map δ:Cn3Leib(H,V)→Cn+13Leib(H,V), for Ai=ai∧bi∈∧2H,1≤i≤n and c∈H, as
(δφ)(A1,A2,…,An,c)=∑1≤j<k≤n(−1)jφ(A1,…,^Aj,…,Ak−1,ak∧[aj,bj,bk]H+[aj,bj,ak]H∧bk,…,An,c)+n∑j=1(−1)jφ(A1,…,^Aj,…,An,[aj,bj,c]H)+n∑j=1(−1)j+1l(Aj,φ(A1,…,^Aj,…,An,c))+(−1)n+1(m(an,φ(A1,…,An−1,bn),c)+r(φ(A1,…,An−1,an),bn,c)). |
It was proved in [23,24] that δ2=0. Therefore, (⊕+∞n=1Cn3Leib(H,V),δ) is a cochain complex.
Let HΛ⟶L be a nonabelian embedding tensor 3-Lie algebra. By Corollary 2.10, (H,[−,−,−]Λ) is a 3-Leibniz algebra. Next we give a representation of (H,[−,−,−]Λ) on L.
Lemma 4.2. With the above notations. Define 3 actions
lΛ:H⊗H⊗L→L,mΛ:H⊗L⊗H→L,rΛ:L⊗H⊗H→L, |
by
lΛ(h1,h2,l)=[Λh1,Λh2,l]L,mΛ(h1,l,h2)=[Λh1,l,Λh2]L−Λρ(Λh1,l)h2,rΛ(l,h1,h2)=[l,Λh1,Λh2]L−Λρ(l,Λh1)h2, |
for all h1,h2∈H,l∈L. Then (L;lΛ,mΛ,rΛ) is a representation of the descendent 3-Leibniz algebra (H,[−,−,−]Λ).
Proof. For any h1,h2,h3,h4,h5∈H and l∈L, by Eqs (2.1), (2.3)–(2.6), and (2.9), we have
lΛ(h1,h2,lΛ(h3,h4,l))−lΛ([h1,h2,h3]Λ,h4,l)−lΛ(h3,[h1,h2,h4]Λ,l)−lΛ(h3,h4,lΛ(h1,h2,l))=[Λh1,Λh2,[Λh3,Λh4,l]L]L−[[Λh1,Λh2,Λh3]L,Λh4,l]L−[Λh3,[Λh1,Λh2,Λh4]L,l]L−[Λh3,Λh4,[Λh1,Λh2,l]L]L=0 |
and
lΛ(h1,h2,mΛ(h3,l,h5))−mΛ([h1,h2,h3]Λ,l,h5)−mΛ(h3,lΛ(h1,h2,l),h5)−mΛ(h3,l,[h1,h2,h5]Λ)=[Λh1,Λh2,[Λh3,l,Λh5]L]L−[Λh1,Λh2,Λρ(Λh3,l)h5]L−[[Λh1,Λh2,Λh3]L,l,Λh5]L+Λρ([Λh1,Λh2,Λh3]L,l)h5−[Λh3,[Λh1,Λh2,l]L,Λh5]L+Λρ(Λh3,[Λh1,Λh2,l]L)h5−[Λh3,l,[Λh1,Λh2,Λh5]L]L+Λρ(Λh3,l)ρ(Λh1,Λh2)h5+Λρ(Λh3,l)[h1,h2,h5]H=−[Λh1,Λh2,Λρ(Λh3,l)h5]L+Λρ([Λh1,Λh2,Λh3]L,l)h5+Λρ(Λh3,[Λh1,Λh2,l]L)h5+Λρ(Λh3,l)ρ(Λh1,Λh2)h5+Λρ(Λh3,l)[h1,h2,h5]H=−Λ(ρ(Λh1,Λh2)ρ(Λh3,l)h5+[h1,h2,ρ(Λh3,l)h5]H)+Λρ(Λh1,Λh2)ρ(Λh3,l)h5+Λρ(Λh3,l)[h1,h2,h5]H=−Λ[h1,h2,ρ(Λh3,l)h5]H+Λρ(Λh3,l)[h1,h2,h5]H=0, |
which imply that Eqs (4.1) and (4.2) hold. Similarly, we can prove that Eqs (4.3)–(4.5) are true. The proof is finished.
Proposition 4.3. Let HΛ1⟶L and HΛ2⟶L be two nonabelian embedding tensor 3-Lie algebras and (fL,fH) a homomorphism from HΛ1⟶L to HΛ2⟶L. Then the induced representation (L;lΛ1,mΛ1,rΛ1) of the descendent 3-Leibniz algebra (H,[−,−,−]Λ1) and the induced representation (L;lΛ2,mΛ2,rΛ2) of the descendent 3-Leibniz algebra (H,[−,−,−]Λ2) satisfying the following equations:
fL(lΛ1(h1,h2,l))=lΛ2(fH(h1),fH(h2),fL(l)), | (4.6) |
fL(mΛ1(h1,l,h2))=mΛ2(fH(h1),fL(l),fH(h2)), | (4.7) |
fL(rΛ1(l,h1,h2))=rΛ2(fL(l),fH(h1),fH(h2)), | (4.8) |
for all h1,h2∈H,l∈L. In other words, the following diagrams commute:
![]() |
Proof. For any h1,h2∈H,l∈L, by Eqs (2.7) and (2.8), we have
fL(lΛ1(h1,h2,l))=fL([Λ1h1,Λ1h2,l]L)=[fL(Λ1h1),fL(Λ1h2),fL(l)]L=[Λ2fH(h1),Λ2fH(h2),fL(l)]L=lΛ2(fH(h1),fH(h2),fL(l)),fL(mΛ1(h1,l,h2))=fL([Λ1h1,l,Λ1h2]L−Λ1ρ(Λ1h1,l)h2)=[fL(Λ1h1),fL(l),fL(Λ1h2)]L−fL(Λ1ρ(Λ1h1,l)h2)=[Λ2fH(h1),fL(l),Λ2fH(h2)]L−Λ2fH(ρ(Λ1h1,l)h2)=[Λ2fH(h1),fL(l),Λ2fH(h2)]L−Λ2ρ(Λ2fH(h1),fL(l))fH(h2)=mΛ2(fH(h1),fL(l),fH(h2)). |
And the other equation is similar to provable.
For n≥1, let δΛ:Cn3Leib(H,L)→Cn+13Leib(H,L) be the coboundary operator of the 3-Leibniz algebra (H,[−,−,−]Λ) with coefficients in the representation (L;lΛ,mΛ,rΛ). More precisely, for all ϕ∈Cn3Leib(H,L),Hi=ui∧vi∈∧2H,1≤i≤n and w∈H, we have
(δΛϕ)(H1,H2,…,Hn,w)=∑1≤j<k≤n(−1)jϕ(H1,…,^Hj,…,Hk−1,uk∧[uj,vj,vk]Λ+[uj,vj,uk]Λ∧vk,…,Hn,w)+n∑j=1(−1)jϕ(H1,…,^Hj,…,Hn,[uj,vj,w]Λ)+n∑j=1(−1)j+1lΛ(Hj,ϕ(H1,…,^Hj,…,Hn,w))+(−1)n+1(mΛ(un,ϕ(H1,…,Hn−1,vn),w)+rΛ(ϕ(H1,…,Hn−1,un),vn,w)). |
In particular, for ϕ∈C13Leib(H,L):=Hom(H,L) and u1,v1,w∈H, we have
(δΛϕ)(u1,v1,w)=−ϕ([u1,v1,w]Λ)+lΛ(u1,v1,ϕ(w))+mΛ(u1,ϕ(v1),w)+rΛ(ϕ(u1),v1,w)=−ϕ([u1,v1,w]Λ)+[Λu1,Λv1,ϕ(w)]L+[Λu1,ϕ(v1),Λw]L−Λρ(Λu1,ϕ(v1))w+[ϕ(u1),Λv1,Λw]L−Λρ(ϕ(u1),Λv1)w. |
For any (a1,a2)∈C03Leib(H,L):=∧2L, we define δΛ:C03Leib(H,L)→C13Leib(H,L),(a1,a2)↦δΛ(a1,a2) by
δΛ(a1,a2)u=Λρ(a1,a2)u−[a1,a2,Λu]L,∀u∈H. |
Proposition 4.4. Let HΛ⟶L be a nonabelian embedding tensor 3-Lie algebra. Then δΛ(δΛ(a1,a2))=0, that is, the composition C03Leib(H,L)δΛ⟶C13Leib(H,L)δΛ⟶C23Leib(H,L) is the zero map.
Proof. For any u1,v1,w∈V, by Eqs (2.1)–(2.6) and (2.9) we have
δΛ(δΛ(a1,a2))(u1,v1,w)=−δΛ(a1,a2)([u1,v1,w]Λ)+[Λu1,Λv1,δΛ(a1,a2)(w)]L+[Λu1,δΛ(a1,a2)(v1),Λw]L−Λρ(Λu1,δΛ(a1,a2)(v1))w+[δΛ(a1,a2)(u1),Λv1,Λw]L−Λρ(δΛ(a1,a2)(u1),Λv1)w=−Λρ(a1,a2)[u1,v1,w]Λ+[a1,a2,[Λu1,Λv1,Λw]L]L+[Λu1,Λv1,Λρ(a1,a2)w]L−[Λu1,Λv1,[a1,a2,Λw]L]L+[Λu1,Λρ(a1,a2)v1,Λw]L−[Λu1,[a1,a2,Λv1]L,Λw]L−Λρ(Λu1,Λρ(a1,a2)v1)w+Λρ(Λu1,[a1,a2,Λv1]L)w+[Λρ(a1,a2)u1,Λv1,Λw]L−[[a1,a2,Λu1]L,Λv1,Λw]L−Λρ(Λρ(a1,a2)u1,Λv1)w+Λρ([a1,a2,Λu1]L,Λv1)w=−Λρ(a1,a2)ρ(Λu1,Λv1)w−Λρ(a1,a2)[u1,v1,w]H+Λρ(Λu1,Λv1)ρ(a1,a2)w+Λ[u1,v1,ρ(a1,a2)w]H+Λρ(Λu1,Λρ(a1,a2)v1)w+Λ[u1,ρ(a1,a2)v1,w]H−Λρ(Λu1,Λρ(a1,a2)v1)w+Λρ(Λu1,[a1,a2,Λv1]L)w+Λ(Λρ(a1,a2)u1,Λv1)w+Λ[ρ(a1,a2)u1,v1,w]H−Λρ(Λρ(a1,a2)u1,Λv1)w+Λρ([a1,a2,Λu1]L,Λv1)w=−Λρ(a1,a2)ρ(Λu1,Λv1)w+Λρ(Λu1,Λv1)ρ(a1,a2)w+Λρ(Λu1,Λρ(a1,a2)v1)w−Λρ(Λu1,Λρ(a1,a2)v1)w+Λρ(Λu1,[a1,a2,Λv1]L)w+Λ(Λρ(a1,a2)u1,Λv1)w−Λρ(Λρ(a1,a2)u1,Λv1)w+Λρ([a1,a2,Λu1]L,Λv1)w=−Λρ(a1,a2)ρ(Λu1,Λv1)w+Λρ(Λu1,Λv1)ρ(a1,a2)w+Λρ(Λu1,[a1,a2,Λv1]L)w+Λρ([a1,a2,Λu1]L,Λv1)w=0. |
Therefore, we deduce that δΛ(δΛ(a1,a2))=0.
Now we develop the cohomology theory of a nonabelian embedding tensor Λ on the 3-Lie algebra (L,[−,−,−]L) with respect to the coherent action (H,[−,−,−]H;ρ†).
For n≥0, define the set of n-cochains of Λ by CnΛ(H,L):=Cn3Leib(H,L). Then (⊕∞n=0CnΛ(H,L),δΛ) is a cochain complex.
For n≥1, we denote the set of n-cocycles by ZnΛ(H,L), the set of n-coboundaries by BnΛ(H,L), and the n-th cohomology group of the nonabelian embedding tensor Λ by
HHnΛ(H,L)=ZnΛ(H,L)BnΛ(H,L). |
Proposition 4.5. Let HΛ1⟶L and HΛ2⟶L be two nonabelian embedding tensor 3-Lie algebras and let (fL,fH) be a homomorphism from HΛ1⟶L to HΛ2⟶L in which fH is invertible. We define a map Ψ:CnΛ1(H,L)→CnΛ2(H,L) by
Ψ(ϕ)(H1,H2,…,Hn−1,w)=fL(ϕ(f−1H(u1)∧f−1H(v1),…,f−1H(un−1)∧f−1H(vn−1),f−1H(w))), |
for all ϕ∈CnΛ1(H,L),Hi=ui∧vi∈∧2H,1≤i≤n−1, and w∈H. Then Ψ:(Cn+1Λ1(H,L),δΛ1)→(Cn+1Λ2(H,L),δΛ2) is a cochain map.
That is, the following diagram commutes:
![]() |
Consequently, it induces a homomorphism Ψ∗ from the cohomology group HHn+1Λ1(H,L) to HHn+1Λ2(H,L).
Proof. For any ϕ∈CnΛ1(H,L),Hi=ui∧vi∈∧2H,1≤i≤n, and w∈H, by Eqs (4.6)–(4.8) and Proposition 2.11, we have
(δΛ2Ψ(ϕ))(H1,H2,…,Hn,w)=∑1≤j<k≤n(−1)jΨ(ϕ)(H1,…,^Hj,…,Hk−1,uk∧[uj,vj,vk]Λ2+[uj,vj,uk]Λ2∧vk,…,Hn,w)+n∑j=1(−1)jΨ(ϕ)(H1,…,^Hj,…,Hn,[uj,vj,w]Λ2)+n∑j=1(−1)j+1lΛ2(Hj,Ψ(ϕ)(H1,…,^Hj,…,Hn,w))+(−1)n+1mΛ2(un,Ψ(ϕ)(H1,…,Hn−1,vn),w)+(−1)n+1rΛ2(Ψ(ϕ)(H1,…,Hn−1,un),vn,w)=∑1≤j<k≤n(−1)jfL(ϕ(f−1H(u1)∧f−1H(v1),…,^Hj,…,f−1H(uk−1)∧f−1H(vk−1),f−1H(uk)∧f−1H([uj,vj,vk]Λ2)+f−1H([uj,vj,uk]Λ2)∧f−1H(vk),…,f−1H(un)∧f−1H(vn),f−1H(w)))+n∑j=1(−1)jfL(ϕ(f−1H(u1)∧f−1H(v1),…,^Hj,…,f−1H(un)∧f−1H(vn),f−1H([uj,vj,w]Λ2)))+n∑j=1(−1)j+1lΛ2(Hj,fL(ϕ(f−1H(u1)∧f−1H(v1),…,^Hj,…,f−1H(un)∧f−1H(vn),f−1H(w))))+(−1)n+1mΛ2(un,fL(ϕ(f−1H(u1)∧f−1H(v1),…,f−1H(un−1)∧f−1H(vn−1),f−1H(vn))),w)+(−1)n+1rΛ2(fL(ϕ(f−1H(u1)∧f−1H(v1),…,f−1H(un−1)∧f−1H(vn−1),f−1H(un))),vn,w)=fL(∑1≤j<k≤n(−1)jϕ(f−1H(u1)∧f−1H(v1),…,^Hj,…,f−1H(uk−1)∧f−1H(vk−1),f−1H(uk)∧[f−1H(uj),f−1H(vj),f−1H(vk)]Λ1+[f−1H(uj),f−1H(vj),f−1H(uk)]Λ1∧f−1H(vk),…,f−1H(un)∧f−1H(vn),f−1H(w))+n∑j=1(−1)jϕ(f−1H(u1)∧f−1H(v1),…,^Hj,…,f−1H(un)∧f−1H(vn),[f−1H(uj),f−1H(vj),f−1H(w)]Λ1)+n∑j=1(−1)j+1lΛ1(f−1H(uj),f−1H(vj),ϕ(f−1H(u1)∧f−1H(v1),…,^Hj,…,f−1H(un)∧f−1H(vn),f−1H(w)))+(−1)n+1mΛ1(f−1H(un),ϕ(f−1H(u1),f−1H(v1),…,f−1H(un−1)∧f−1H(vn−1),f−1H(vn)),f−1H(w))+(−1)n+1rΛ1(ϕ(f−1H(u1)∧f−1H(v1),…,f−1H(un−1)∧f−1H(vn−1),f−1H(un)),f−1H(vn),f−1H(w)))=fL(δΛ1ϕ)(f−1H(u1)∧f−1H(v1),…,f−1H(un)∧f−1H(vn),f−1H(w))=Ψ(δΛ1ϕ)(H1,H2,…,Hn,w). |
Hence, Ψ is a cochain map and induces a cohomology group homomorphism Ψ∗:HHn+1Λ1(H,L) →HHn+1Λ2(H,L).
At the conclusion of this section, we employ the well-established cohomology theory to describe the infinitesimal deformations of nonabelian embedding tensors on 3-Lie algebras.
Definition 4.6. Let Λ:H→L be a nonabelian embedding tensor on a 3-Lie algebra (L,[−,−,−]L) with respect to a coherent action (H,[−,−,−]H;ρ†). An infinitesimal deformation of Λ is a nonabelian embedding tensor of the form Λt=Λ+tΛ1, where t is a parameter with t2=0.
Let Λt=Λ+tΛ1 be an infinitesimal deformation of Λ, then we have
[Λtu1,Λtu2,Λtu3]L=Λtρ(Λtu1,Λtu2)u3+Λt[u1,u2,u3]H, |
for all u1,u2,u3∈H. Therefore, we obtain the following equation:
[Λ1u1,Λu2,Λu3]L+[Λu1,Λ1u2,Λu3]L+[Λu1,Λu2,Λ1u3]L=Λ1ρ(Λu1,Λu2)u3+Λρ(Λ1u1,Λu2)u3+Λρ(Λu1,Λ1u2)u3+Λ1[u1,u2,u3]H. | (4.9) |
It follows from Eq (4.9) that Λ1∈C1Λ(H,L) is a 1-cocycle in the cohomology complex of Λ. Thus the cohomology class of Λ1 defines an element in HH1Λ(H,L).
Let Λt=Λ+tΛ1 and Λ′t=Λ+tΛ′1 be two infinitesimal deformations of Λ. They are said to be equivalent if there exists a1∧a2∈∧2L such that the pair (idL+tad(a1,a2),idH+tρ(a1,a2)) is a homomorphism from HΛt⟶L to HΛ′t⟶L. That is, the following conditions must hold:
1) The maps idL+tad(a1,a2):L→L and idH+tρ(a1,a2):H→H are two 3-Lie algebra homomorphisms,
2) The pair (idL+tad(a1,a2),idH+tρ(a1,a2)) satisfies:
(idH+tρ(a1,a2))(ρ(a,b)u)=ρ((idL+tad(a1,a2))a,(idL+tad(a1,a2))b)(idH+tρ(a1,a2))(u),(Λ+tΛ′1)(idH+tρ(a1,a2))(u)=(idL+tad(a1,a2))((Λ+tΛ1)u), | (4.10) |
for all a,b∈L,u∈H. It is easy to see that Eq (4.10) gives rise to
Λ1u−Λ′1u=Λρ(a1,a2)u−[a1,a2,Λu]=δΛ(a1,a2)u∈C1Λ(H,L). |
This shows that Λ1 and Λ′1 are cohomologous. Thus, their cohomology classes are the same in HH1Λ(H,L).
Conversely, any 1-cocycle Λ1 gives rise to the infinitesimal deformation Λ+tΛ1. Furthermore, we have arrived at the following result.
Theorem 4.7. Let Λ:H→L be a nonabelian embedding tensor on (L,[−,−,−]L) with respect to (H,[−,−,−]H;ρ†). Then, there exists a bijection between the set of all equivalence classes of infinitesimal deformations of Λ and the first cohomology group HH1Λ(H,L).
The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.
This research is supported by the National Natural Science Foundation of China (Grant No. 12361005) and the Universities Key Laboratory of System Modeling and Data Mining in Guizhou Province (Grant No. 2023013).
The authors declare there is no conflicts of interest.
[1] |
Z. He, W. Zhang, N. Jia, Estimating carbon dioxide emissions of freeway traffic: a spatiotemporal cell-based model, IEEE Trans. Intell. Transp. Syst., 21 (2020), 1976–1986. https://doi.org/10.1109/tits.2019.2909316 doi: 10.1109/tits.2019.2909316
![]() |
[2] |
Q. Cheng, Z. Liu, Y. Lin, X. Zhou, An s-shaped three-parameter (S3) traffic stream model with consistent car following relationship, Transp. Res. Part B Methodol., 153 (2021), 246–271. https://doi.org/10.1016/j.trb.2021.09.004 doi: 10.1016/j.trb.2021.09.004
![]() |
[3] |
Q. Cheng, Z. Liu, J. Guo, X. Wu, R. Pendyala, B. Belezamo, et al., Estimating key traffic state parameters through parsimonious spatial queue models, Transp. Res. Part C Emerging Technol., 137 (2022), 103596. https://doi.org/10.1016/j.trc.2022.103596 doi: 10.1016/j.trc.2022.103596
![]() |
[4] | Z. Shan, D. Zhao, Y. Xia, Urban road traffic speed estimation for missing probe vehicle data based on multiple linear regression model, in 16th International IEEE Conference on Intelligent Transportation Systems, 1 (2013), 118–123. https://doi.org/10.1109/ITSC.2013.6728220 |
[5] |
Z. Liu, Z. Li, M. Li, W. Xing, D. Lu, Mining road network correlation for traffic estimation via compressive sensing, IEEE Trans. Intell. Transp. Syst., 17 (2016), 1880–1893. https://doi.org/10.1109/tits.2016.2514519 doi: 10.1109/tits.2016.2514519
![]() |
[6] |
Z. He, G. Qi, L. Lu, Y. Chen, Network-wide identification of turn-level intersection congestion using only low-frequency probe vehicle data, Transp. Res. Part C Emerging Technol., 108 (2019), 320–339. https://doi.org/10.1016/j.trc.2019.10.001 doi: 10.1016/j.trc.2019.10.001
![]() |
[7] | J. Aslam, S. Lim, X. Pan, D. Rus, City-scale traffic estimation from a roving sensor network, in Proceedings of the 10th ACM Conference on Embedded Network Sensor Systems, 2012 (2012), 141–154. https://doi.org/10.1145/2426656.2426671 |
[8] |
Y. Song, X. Wang, G. Wright, D. Thatcher, P. Wu, P. Felix, Traffic volume prediction with segment-based regression kriging and its implementation in assessing the impact of heavy vehicles, IEEE Trans. Intell. Transp. Syst., 20 (2019), 232–243. https://doi.org/10.1109/tits.2018.2805817 doi: 10.1109/tits.2018.2805817
![]() |
[9] |
Z. Liu, Z. Wang, Q. Cheng, R. Yin, M. Wang, Estimation of urban network capacity with second-best constraints for multimodal transport systems, Transp. Res. Part B Methodol., 152 (2021), 276–294. https://doi.org/10.1016/j.trb.2021.08.011 doi: 10.1016/j.trb.2021.08.011
![]() |
[10] |
Z. Cheng, J. Lu, H. Zhou, Y. Zhang, L. Zhang, Short-Term traffic flow prediction: an integrated method of econometrics and hybrid deep learning, IEEE Trans. Intell. Transp. Syst., 23 (2022), 5231–5244. https://doi.org/10.1109/TITS.2021.3052796 doi: 10.1109/TITS.2021.3052796
![]() |
[11] |
L. Li, J. Zhang, Y. Wang, B. Ran, Missing value imputation for traffic-related time series data based on a multi-view learning method, IEEE Trans. Intell. Transp. Syst., 20 (2019), 2933–2943. https://doi.org/10.1109/tits.2018.2869768 doi: 10.1109/tits.2018.2869768
![]() |
[12] | C. Meng, X. Yi, L. Su, J. Gao, Y. Zheng, City-wide traffic volume inference with loop detector data and taxi trajectories, in Proceedings of the 25th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, 2 (2017), 1–10. https://doi.org/10.1145/3139958.3139984 |
[13] |
Z. Yi, X. C. Liu, N. Markovic, J. Phillips, Inferencing hourly traffic volume using data-driven machine learning and graph theory, Comput. Environ. Urban Syst., 85 (2021), 101548. https://doi.org/10.1016/j.compenvurbsys.2020.101548 doi: 10.1016/j.compenvurbsys.2020.101548
![]() |
[14] | Z. Pan, Y. Liang, W. Wang, Y. Yu, Y. Zheng, J. Zhang, Urban traffic prediction from spatio-temporal data using deep meta learning, in Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2019 (2019), 1720–1730. https://doi.org/10.1145/3292500.3330884 |
[15] |
J. Li, N. Xie, K. Zhang, F. Guo, S. Hu, X. Chen, Network-scale traffic prediction via knowledge transfer and regional MFD analysis, Transp. Res. Part C Emerging Technol., 141 (2022), 12–34. https://doi.org/10.1016/j.trc.2022.103719 doi: 10.1016/j.trc.2022.103719
![]() |
[16] |
S. Luan, R. Ke, Z. Huang, X. Ma, Traffic congestion propagation inference using dynamic Bayesian graph convolution network, Transp. Res. Part C Emerging Technol., 135 (2022), 103526. https://doi.org/10.1016/j.trc.2021.103526 doi: 10.1016/j.trc.2021.103526
![]() |
[17] |
Z. Liu, Y. Liu, Q. Meng, Q. Cheng, A tailored machine learning approach for urban transport network flow estimation, Transp. Res. Part C Emerging Technol., 108 (2019), 130–150. https://doi.org/10.1016/j.trc.2019.09.006 doi: 10.1016/j.trc.2019.09.006
![]() |
[18] |
Q. Cheng, Z. Liu, W. Y. Szeto, A cell-based dynamic congestion pricing scheme considering travel distance and time delay, Transportmetrica B Transport Dyn., 7 (2019), 1286–1304 https://doi.org/10.1080/21680566.2019.1602487 doi: 10.1080/21680566.2019.1602487
![]() |
[19] |
Q. Cheng, S. Wang, Z. Liu, Y. Yuan, Surrogate-based simulation optimization approach for day-to-day dynamics model calibration with real data, Transp. Res. Part C Emerging Technol., 105 (2019), 422–438. https://doi.org/10.1016/j.trc.2019.06.009 doi: 10.1016/j.trc.2019.06.009
![]() |
[20] |
D. Huang, J. Xing, Z. Liu, Q. An, A multi-stage stochastic optimization approach to the stop-skipping and bus lane reservation schemes, Transportmetrica A Transport Sci., 17 (2021), 1272–1304. https://doi.org/10.1080/23249935.2020.1858206 doi: 10.1080/23249935.2020.1858206
![]() |
[21] |
D. Huang, Y. Wang, S. Jia, Z. Liu, S. Wang, A Lagrangian relaxation approach for the electric bus charging scheduling optimisation problem, Transportmetrica A Transport Sci., 202232 (2022). https://doi.org/10.1080/23249935.2021.2023690 doi: 10.1080/23249935.2021.2023690
![]() |
[22] |
Z. Liu, X. Chen, Q. Meng, I. Kim, Remote park-and-ride network equilibrium model and its applications, Transp. Res. Part B Methodol., 117 (2018), 37–62. https://doi.org/https://doi.org/10.1016/j.trb.2018.08.004 doi: 10.1016/j.trb.2018.08.004
![]() |
[23] |
J. Xing, W. Wu, Q. Cheng, R. Liu, Traffic state estimation of urban road networks by multi-source data fusion: review and new insights, Physica A Stat. Mech. Appl., 595 (2022), 127079. https://doi.org/https://doi.org/10.1016/j.physa.2022.127079 doi: 10.1016/j.physa.2022.127079
![]() |
[24] |
Z. Zhang, X. Lin, M. Li, Y. Wang, A customized deep learning approach to integrate network-scale online traffic data imputation and prediction, Transp. Res. Part C Emerging Technol., 132 (2021), 103372. https://doi.org/10.1016/j.trc.2021.103372 doi: 10.1016/j.trc.2021.103372
![]() |
[25] |
L. Li, R. Jiang, Z. He, X. Chen, X. Zhou, Trajectory data-based traffic flow studies: a revisit, Transp. Res. Part C Emerging Technol., 114 (2020), 225–240. https://doi.org/10.1016/j.trc.2020.02.016 doi: 10.1016/j.trc.2020.02.016
![]() |
[26] |
Z. Liu, P. Zhou, Z. Li, M. Li, Think like a graph: real-time traffic estimation at city-scale, IEEE Trans. Mobile Comput., 18 (2019), 2446–2459. https://doi.org/10.1109/tmc.2018.2873642 doi: 10.1109/tmc.2018.2873642
![]() |
[27] |
Q. Cao, G. Ren, D. Li, J. Ma, H. Li, Semi-supervised route choice modeling with sparse automatic vehicle identification data, Transp. Res. Part C Emerging Technol., 121 (2020). https://doi.org/10.1016/j.trc.2020.102857 doi: 10.1016/j.trc.2020.102857
![]() |
[28] |
Y. Yu, X. Tang, H. Yao, X. Yi, Z. Li, Citywide traffic volume inference with surveillance camera records, IEEE Trans. Big Data, 7 (2021), 900–912. https://doi.org/10.1109/tbdata.2019.2935057 doi: 10.1109/tbdata.2019.2935057
![]() |
[29] |
X. Zhan, Z. Yu, X. Yi, S. V. Ukkusuri, Citywide traffic volume estimation using trajectory data, IEEE Trans. Knowl. Data Eng., 29 (2017), 272–285. https://doi.org/10.1109/TKDE.2016.2621104 doi: 10.1109/TKDE.2016.2621104
![]() |
[30] |
P. Wang, Z. Huang, J. Lai, Z. Zheng, Y. Liu, T. Lin, Traffic speed estimation based on multi-source GPS data and mixture model, IEEE Trans. Intell. Transp. Syst., 23 (2021), 10708–10720. https://doi.org/10.1109/tits.2021.3095408 doi: 10.1109/tits.2021.3095408
![]() |
[31] |
M. Seppecher, L. Leclercq, A. Furno, D. Lejri, T. V. da Rocha, Estimation of urban zonal speed dynamics from user-activity-dependent positioning data and regional paths, Transp. Res. Part C Emerging Technol., 129 (2021), 103183. https://doi.org/10.1016/j.trc.2021.103183 doi: 10.1016/j.trc.2021.103183
![]() |
[32] |
E. Saffari, M. Yildirimoglu, M. Hickman, Data fusion for estimating Macroscopic Fundamental Diagram in large-scale urban networks, Transp. Res. Part C Emerging Technol., 137 (2022), 103555. https://doi.org/10.1016/j.trc.2022.103555 doi: 10.1016/j.trc.2022.103555
![]() |
[33] |
M. Rodriguez-Vega, C. Canudas-de-Wit, H. Fourati, Dynamic density and flow reconstruction in large-scale urban networks using heterogeneous data sources, Transp. Res. Part C Emerging Technol., 137 (2022), 103569. https://doi.org/10.1016/j.trc.2022.103569 doi: 10.1016/j.trc.2022.103569
![]() |
[34] |
M. Yun, W. Qin, Minimum sampling size of floating cars for urban link travel time distribution estimation, Transp. Res. Rec. J. Transp. Res. Board, 2673 (2019), 24–43. https://doi.org/10.1177/0361198119834297 doi: 10.1177/0361198119834297
![]() |
[35] |
Z. Huang, X. Ling, P. Wang, F. Zhang, Y. Mao, T. Lin, et al., Modeling real-time human mobility based on mobile phone and transportation data fusion, Transp. Res. Part C Emerging Technol., 96 (2018), 251–269. https://doi.org/10.1016/j.trc.2018.09.016 doi: 10.1016/j.trc.2018.09.016
![]() |
[36] |
C. Wu, I. Kim, H. Chung, The effects of built environment spatial variation on bike-sharing usage: a case study of Suzhou, China, Cities, 110 (2021), 103063. https://doi.org/10.1016/j.cities.2020.103063 doi: 10.1016/j.cities.2020.103063
![]() |
[37] |
C. Wu, H. Chung, Z. Liu, I. Kim, Examining the effects of the built environment on topological properties of the bike-sharing network in Suzhou, China, Int. J. Sustainable Transp., 15 (2021), 338–350. https://doi.org/10.1080/15568318.2020.1780652 doi: 10.1080/15568318.2020.1780652
![]() |
[38] |
S. J. Pan, I. W. Tsang, J. T. Kwok, Q. Yang, Domain adaptation via transfer component analysis, IEEE Trans. Neural Networks, 22 (2011), 199–210. https://doi.org/10.1109/TNN.2010.2091281 doi: 10.1109/TNN.2010.2091281
![]() |
[39] | S. J. Pan, J. T. Kwok, Q. Yang, Transfer learning via dimensionality reduction, in Proceedings of the Twenty-Third AAAI Conference on Artificial Intelligence, 2 (2008), 677–682. https://dl.acm.org/doi/abs/10.5555/1620163.1620177 |
[40] |
J. Wang, Y. Chen, W. Feng, H. Yu, Q. Yang, Transfer learning with dynamic distribution adaptation, ACM Trans. Intell. Syst. Technol., 11 (2020), 1–25. https://doi.org/10.1145/3360309 doi: 10.1145/3360309
![]() |
[41] |
Z. Cheng, L. Zhang, Y. Zhang, S. Wang, W. Huang, A systematic approach for evaluating spatiotemporal characteristics of traffic violations and crashes at road intersections: an empirical study, Transportmetrica A Transp. Sci., 2022 (2022), 1–23. https://doi.org/10.1080/23249935.2022.2060368 doi: 10.1080/23249935.2022.2060368
![]() |
[42] |
Y. Gu, A. Chen, S. Kitthamkesorn, Accessibility-based vulnerability analysis of multi-modal transportation networks with weibit choice models, Multimodal Transp., 1 (2022), 100029. https://doi.org/10.1016/j.multra.2022.100029 doi: 10.1016/j.multra.2022.100029
![]() |
[43] |
Y. Zheng, W. Li, F. Qiu, A slack arrival strategy to promote flex-route transit services, Transp. Res. Part C Emerging Technol., 92 (2018), 442–455. https://doi.org/10.1016/j.trc.2018.05.015 doi: 10.1016/j.trc.2018.05.015
![]() |
[44] |
S. Wang, D. Yu, M. P. Kwan, L. Zheng, H. Miao, Y. Li, The impacts of road network density on motor vehicle travel: an empirical study of Chinese cities based on network theory, Transp. Res. Part A Policy Pract., 132 (2020), 144–156. https://doi.org/10.1016/j.tra.2019.11.012 doi: 10.1016/j.tra.2019.11.012
![]() |
[45] |
Z. Zhang, M. Li, X. Lin, Y. Wang, Network-wide traffic flow estimation with insufficient volume detection and crowdsourcing data, Transp. Res. Part C Emerging Technol., 121 (2020), 102870. https://doi.org/10.1016/j.trc.2020.102870 doi: 10.1016/j.trc.2020.102870
![]() |
[46] |
D. Huang, S. Wang, A two-stage stochastic programming model of coordinated electric bus charging scheduling for a hybrid charging scheme, Multimodal Transp., 1 (2022), 100006. https://doi.org/10.1016/j.multra.2022.100006 doi: 10.1016/j.multra.2022.100006
![]() |
[47] |
D. Xiao, I. Kim, N. Zheng, Recent advances in understanding the impact of built environment on traffic performance, Multimodal Transp., 1 (2022), 100034. https://doi.org/10.1016/j.multra.2022.100034 doi: 10.1016/j.multra.2022.100034
![]() |
[48] |
J. Xing, Z. Liu, C. Wu, S. Chen, Traffic volume estimation in multimodal urban networks using cellphone location data, IEEE Intell. Transp. Syst. Mag., 11 (2019), 93–104. https://doi.org/10.1109/mits.2019.2919593 doi: 10.1109/mits.2019.2919593
![]() |
[49] |
A. H. F. Chow, Z. C. Su, E. M. Liang, R. X. Zhong, Adaptive signal control for bus service reliability with connected vehicle technology via reinforcement learning, Transp. Res. Part C Emerging Technol., 129 (2021), 103264. https://doi.org/10.1016/j.trc.2021.103264 doi: 10.1016/j.trc.2021.103264
![]() |
[50] |
S. Wang, D. Yu, X. Ma, X. Xing, Analyzing urban traffic demand distribution and the correlation between traffic flow and the built environment based on detector data and POIs, Eur. Transport Res. Rev., 10 (2018). https://doi.org/10.1186/s12544-018-0325-5 doi: 10.1186/s12544-018-0325-5
![]() |
[51] |
R. Zhong, J. Luo, H. Cai, A. Sumalee, F. Yuan, A. Chow, Forecasting journey time distribution with consideration to abnormal traffic conditions, Transp. Res. Part C Emerging Technol., 85 (2017), 292–311. https://doi.org/https://doi.org/10.1016/j.trc.2017.08.021 doi: 10.1016/j.trc.2017.08.021
![]() |
[52] |
W. Qin, X. Ji, F. Liang, Estimation of urban arterial travel time distribution considering link correlations, Transportmetrica A Transport Sci., 16 (2020), 1429–1458. https://doi.org/10.1080/23249935.2020.1751341 doi: 10.1080/23249935.2020.1751341
![]() |
[53] |
S. Kullback, R. A. Leibler, On information and sufficiency, Ann. Math. Statist., 22 (1951), 79–86. https://doi.org/10.1214/aoms/1177729694 doi: 10.1214/aoms/1177729694
![]() |
[54] |
Y. Jiang, O. A. Nielsen, Urban multimodal traffic assignment, Multimodal Transp., 1 (2022), 100027. https://doi.org/10.1016/j.multra.2022.100027 doi: 10.1016/j.multra.2022.100027
![]() |
[55] |
R. Yan, S. Wang, Integrating prediction with optimization: models and applications in transportation management, Multimodal Transp., 1 (2022), 100018. https://doi.org/10.1016/j.multra.2022.100018 doi: 10.1016/j.multra.2022.100018
![]() |
[56] |
Y. Zheng, W. Li, F. Qiu, H. Wei, The benefits of introducing meeting points into flex-route transit services, Transp Res. Part C Emerging Technol., 106 (2019), 98–112. https://doi.org/10.1016/j.trc.2019.07.012 doi: 10.1016/j.trc.2019.07.012
![]() |