generalized tasep on the edge of jamming. · generalized tasep on the edge of jamming....
TRANSCRIPT
![Page 1: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/1.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Generalized TASEP on the edge of jamming.
Alexander Povolotsky
Joint Institute for Nuclear Research, Dubna &
Higher School of Economics, Moscow
In collaboration with: A.E. Derbyshev, V.B. Priezzhev
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 2: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/2.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Outline
1 ModelsGeneralized TASEPZero-range chipping models with factorized steady stateIntegrability
2 Long-time behaviour of gTASEP.Stationary stateFluctuations of particle current
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 3: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/3.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Discrete time dynamics
Clusterwise update: At every time step each cluster is updatedindependently.First particle of a cluster jumps forward with probability p orstays with probability (1−p).If the first particle decided to jump, the next particle follows itwith probability µ and so do the second, third, e.t.c.Exclusion interaction (jumps to occupied sites are forbidden).
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 4: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/4.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Discrete time dynamics
Clusterwise update: At every time step each cluster is updatedindependently.First particle of a cluster jumps forward with probability p orstays with probability (1−p).If the first particle decided to jump, the next particle follows itwith probability µ and so do the second, third, e.t.c.Exclusion interaction (jumps to occupied sites are forbidden).
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 5: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/5.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Discrete time dynamics
Clusterwise update: At every time step each cluster is updatedindependently.First particle of a cluster jumps forward with probability p orstays with probability (1−p).If the first particle decided to jump, the next particle follows itwith probability µ and so do the second, third, e.t.c.Exclusion interaction (jumps to occupied sites are forbidden).
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 6: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/6.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Discrete time dynamics
Clusterwise update: At every time step each cluster is updatedindependently.First particle of a cluster jumps forward with probability p orstays with probability (1−p).If the first particle decided to jump, the next particle follows itwith probability µ and so do the second, third, e.t.c.Exclusion interaction (jumps to occupied sites are forbidden).
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 7: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/7.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Particular limits
µ = 0 — TASEP with parallel update (PU)µ = p — backward sequential update (BSU)µ → 1 — deterministic agregation (DA) limit
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 8: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/8.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Particular limits
µ = 0 — TASEP with parallel update (PU)µ = p — backward sequential update (BSU)µ → 1 — deterministic agregation (DA) limit
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 9: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/9.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Particular limits
µ = 0 — TASEP with parallel update (PU)µ = p — backward sequential update (BSU)µ → 1 — deterministic agregation (DA) limit
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 10: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/10.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
State space
M particles on the lattice
n = (0,1,1,0,1,1)x = (1,2,4,5)
,
Particle configurations
Occupation numbers: n = {ni}i∈L ,∑i∈L ni = M,ni ∈ {0,1}-ASEP like, ni ∈ Z≥0-ZRP likeParticle coordinates: x = (x1,< . . . ,< xM)⊂L -ASEP-like orx = (x1,≤ . . . ,≤ xM)-ZRP like
L =Z - infinite lattice orL = Z/LZ - periodic lattice with Lsites
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 11: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/11.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Markov chain.
Chapman-Kolmogorov equation:
Pt+1(n) = ∑{n′}
Mn,n′Pt(n′)
Stationary state
Pst(n) = ∑{n′}
Mn,n′Pst(n′)
Factorized stationary measure:
Pst(n) = ∏i∈L
f (ni )
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 12: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/12.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Markov chain.
Chapman-Kolmogorov equation:
Pt+1(n) = ∑{n′}
Mn,n′Pt(n′)
Stationary state
Pst(n) = ∑{n′}
Mn,n′Pst(n′)
Factorized stationary measure:
Pst(n) = ∏i∈L
f (ni )
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 13: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/13.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Markov chain.
Chapman-Kolmogorov equation:
Pt+1(n) = ∑{n′}
Mn,n′Pt(n′)
Stationary state
Pst(n) = ∑{n′}
Mn,n′Pst(n′)
Factorized stationary measure:
Pst(n) = ∏i∈L
f (ni )
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 14: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/14.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Chipping models
Dynamical rules:one-sided nearest neighbor hoppingzero-range interactionϕ(m|n)– probability for m particles to jump from a site withn ≥m particles
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 15: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/15.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Chipping models
Dynamical rules:one-sided nearest neighbor hoppingzero-range interactionϕ(m|n)– probability for m particles to jump from a site withn ≥m particles
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 16: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/16.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Chipping models
Dynamical rules:one-sided nearest neighbor hoppingzero-range interactionϕ(m|n)– probability for m particles to jump from a site withn ≥m particles
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 17: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/17.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Chipping models
Markov matrix:Mn,n′ = ∑{mk∈Z≥0}k∈L ∏i∈L Tmi−1,mi
ni ,n′i
Tmi−1,mini ,n′i
= δ(ni−n′i ),(mi−1−mi )ϕ(mi |n′i )
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 18: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/18.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Factorization of stationary measure
Theorem (Evans, Majumdar, Zia 2004)
The stationary measure of zero-range chipping models on a ring isthe product measure iff the chipping probability is of the form
ϕ(m|n) =v(m)w(n−m)
∑ni=0 v(i)w(n− i)
,
where w(k),v(m)≥ 0, in which case
Pst (n) =1
Z (M,N)
N
∏i=1
f (ni ) with f (n) =n
∑i=0
v(i)w(n− i)
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 19: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/19.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Factorization of stationary measure
Theorem (Evans, Majumdar, Zia 2004)
The stationary measure of zero-range chipping models on a ring isthe product measure iff the chipping probability is of the form
ϕ(m|n) =v(m)w(n−m)
∑ni=0 v(i)w(n− i)
,
where w(k),v(m)≥ 0, in which case
Pst (n) =1
Z (M,N)
N
∏i=1
f (ni ) with f (n) =n
∑i=0
v(i)w(n− i)
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 20: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/20.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Diagonalize that
Eigenvalue problem
MΨ = ΛΨ, ΨM = ΛΨ
Look for the eigenvector in the form:
Ψn = Ψ0nPst(n)
Inversion symmetry:
Ψn = ΠΨ0n
where Π(x1, . . . ,xN) = (−xN , . . . ,−x1)RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 21: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/21.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Diagonalize that
Eigenvalue problem
MΨ = ΛΨ, ΨM = ΛΨ
Look for the eigenvector in the form:
Ψn = Ψ0nPst(n)
Inversion symmetry:
Ψn = ΠΨ0n
where Π(x1, . . . ,xN) = (−xN , . . . ,−x1)RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 22: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/22.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Diagonalize that
Eigenvalue problem
MΨ = ΛΨ, ΨM = ΛΨ
Look for the eigenvector in the form:
Ψn = Ψ0nPst(n)
Inversion symmetry:
Ψn = ΠΨ0n
where Π(x1, . . . ,xN) = (−xN , . . . ,−x1)RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 23: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/23.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Diagonalize that
Eigenvalue problem
MΨ = ΛΨ, ΨM = ΛΨ
Look for the eigenvector in the form:
Ψn = Ψ0nPst(n)
Inversion symmetry:
Ψn = ΠΨ0n
where Π(x1, . . . ,xN) = (−xN , . . . ,−x1)RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 24: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/24.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Two-particle reducibility
One particle problem
ΛΨ0(x) = pΨ0(x−1) + (1−p)Ψ0(x)
Many particle problem (Interacting)
∑nk=0 ϕ(k |n)Ψ0(. . . ,(x−1)k ,xn−k , . . .)
Many particle problem (Free)
∑1k1=0 · · ·∑1
kn=0 pk1+···+kn(1−p)n−(k1+···+kn)Ψ0(. . . ,x−k1, . . . ,x−
kn, . . .)
Boundary conditions
Ψ0(x ,x−1) = αΨ0(x−1,x−1) + β Ψ0(x−1,x) + γΨ0(x ,x)
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 25: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/25.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Two-particle reducibility
One particle problem
ΛΨ0(x) = pΨ0(x−1) + (1−p)Ψ0(x)
Many particle problem (Interacting)
∑nk=0 ϕ(k |n)Ψ0(. . . ,(x−1)k ,xn−k , . . .)
Many particle problem (Free)
∑1k1=0 · · ·∑1
kn=0 pk1+···+kn(1−p)n−(k1+···+kn)Ψ0(. . . ,x−k1, . . . ,x−
kn, . . .)
Boundary conditions
Ψ0(x ,x−1) = αΨ0(x−1,x−1) + β Ψ0(x−1,x) + γΨ0(x ,x)
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 26: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/26.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Two-particle reducibility
One particle problem
ΛΨ0(x) = pΨ0(x−1) + (1−p)Ψ0(x)
Many particle problem (Interacting)
∑nk=0 ϕ(k |n)Ψ0(. . . ,(x−1)k ,xn−k , . . .)
Many particle problem (Free)
∑1k1=0 · · ·∑1
kn=0 pk1+···+kn(1−p)n−(k1+···+kn)Ψ0(. . . ,x−k1, . . . ,x−
kn, . . .)
Boundary conditions
Ψ0(x ,x−1) = αΨ0(x−1,x−1) + β Ψ0(x−1,x) + γΨ0(x ,x)
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 27: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/27.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Two-particle reducibility
One particle problem
ΛΨ0(x) = pΨ0(x−1) + (1−p)Ψ0(x)
Many particle problem (Interacting)
∑nk=0 ϕ(k |n)Ψ0(. . . ,(x−1)k ,xn−k , . . .)
Many particle problem (Free)
∑1k1=0 · · ·∑1
kn=0 pk1+···+kn(1−p)n−(k1+···+kn)Ψ0(. . . ,x−k1, . . . ,x−
kn, . . .)
Boundary conditions
Ψ0(x ,x−1) = αΨ0(x−1,x−1) + β Ψ0(x−1,x) + γΨ0(x ,x)
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 28: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/28.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Generalized quantum binomial
Problem reformulationConsider an associative algebra with two generators A,B satisfyinggeneral homogeneous quadratic relation
BA = αAA+ βAB + γBB,
where α,β ,γ ∈ C such that α + β + γ = 1. Find the coefficients forthe generalized quantum binomial
(pA+ (1−p)B)n =n
∑m=0
ϕ(m|n)AmBn−m.
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 29: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/29.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Generalized quantum binomial
Theorem (P.)
ϕ(m|n) = µm (ν/µ;q)m(µ;q)n−m
(ν ;q)n
(q;q)n
(q;q)m(q;q)n−m,
where α = ν(1−q)1−qν
, β = q−ν
1−qν, γ = 1−q
1−qν,µ = p+ ν(1−p) and
ν 6= q−k , for k ∈ N. (For ν = q−k see Corwin, Petrov, 2015 )
In particular the functions v(k),w(k) and f (k) are
v(k) = µk (ν/µ;q)k
(q;q)k, w(k) =
(µ;q)k
(q;q)k, f (n) =
(ν ;q)n
(q,q)n
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 30: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/30.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
ZRP-ASEP mapping and particle hole transformation
φ
φ
φ φ
φφ
(a)
(b)
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 31: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/31.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
q-Hahn process. Particular cases.
q = 1;ϕ(m|n) = pm(1−p)n−mCmn : Independent particles
µ = qν ;ϕ(m|n) = [n]q q-boson (Sasamoto,Wadati 98) andq-TASEP (Borodin Corwin, 2011)ν → µ = q, ϕ(m|n)' dt/[n]1/q MADM and and long rangehopping models, (Sasamoto Wadati, 1998; Alimohammadi,Karimipour, Khorrami, 1998)ν = 0, Geometric q-TASEP (Borodin Corwin, 2013)q→ 1,µ = qα ,ν = qα+β , ϕ(m|n) — Beta-Binomialdistribution (Barraquand, Corwin, 2015)
Generalized TASEP (M. Woelki, 2005, Derbyshev, Poghosyan, P.,Priezzhev, 2012)
q = 0,ϕ(m|n) =
(1−p), m = 0;pµm−1 (1−µ) , 0 < m < n;pµn−1 m = n,
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 32: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/32.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
q-Hahn process. Particular cases.
q = 1;ϕ(m|n) = pm(1−p)n−mCmn : Independent particles
µ = qν ;ϕ(m|n) = [n]q q-boson (Sasamoto,Wadati 98) andq-TASEP (Borodin Corwin, 2011)ν → µ = q, ϕ(m|n)' dt/[n]1/q MADM and and long rangehopping models, (Sasamoto Wadati, 1998; Alimohammadi,Karimipour, Khorrami, 1998)ν = 0, Geometric q-TASEP (Borodin Corwin, 2013)q→ 1,µ = qα ,ν = qα+β , ϕ(m|n) — Beta-Binomialdistribution (Barraquand, Corwin, 2015)
Generalized TASEP (M. Woelki, 2005, Derbyshev, Poghosyan, P.,Priezzhev, 2012)
q = 0,ϕ(m|n) =
(1−p), m = 0;pµm−1 (1−µ) , 0 < m < n;pµn−1 m = n,
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 33: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/33.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
q-Hahn process. Particular cases.
q = 1;ϕ(m|n) = pm(1−p)n−mCmn : Independent particles
µ = qν ;ϕ(m|n) = [n]q q-boson (Sasamoto,Wadati 98) andq-TASEP (Borodin Corwin, 2011)ν → µ = q, ϕ(m|n)' dt/[n]1/q MADM and and long rangehopping models, (Sasamoto Wadati, 1998; Alimohammadi,Karimipour, Khorrami, 1998)ν = 0, Geometric q-TASEP (Borodin Corwin, 2013)q→ 1,µ = qα ,ν = qα+β , ϕ(m|n) — Beta-Binomialdistribution (Barraquand, Corwin, 2015)
Generalized TASEP (M. Woelki, 2005, Derbyshev, Poghosyan, P.,Priezzhev, 2012)
q = 0,ϕ(m|n) =
(1−p), m = 0;pµm−1 (1−µ) , 0 < m < n;pµn−1 m = n,
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 34: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/34.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
q-Hahn process. Particular cases.
q = 1;ϕ(m|n) = pm(1−p)n−mCmn : Independent particles
µ = qν ;ϕ(m|n) = [n]q q-boson (Sasamoto,Wadati 98) andq-TASEP (Borodin Corwin, 2011)ν → µ = q, ϕ(m|n)' dt/[n]1/q MADM and and long rangehopping models, (Sasamoto Wadati, 1998; Alimohammadi,Karimipour, Khorrami, 1998)ν = 0, Geometric q-TASEP (Borodin Corwin, 2013)q→ 1,µ = qα ,ν = qα+β , ϕ(m|n) — Beta-Binomialdistribution (Barraquand, Corwin, 2015)
Generalized TASEP (M. Woelki, 2005, Derbyshev, Poghosyan, P.,Priezzhev, 2012)
q = 0,ϕ(m|n) =
(1−p), m = 0;pµm−1 (1−µ) , 0 < m < n;pµn−1 m = n,
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 35: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/35.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
q-Hahn process. Particular cases.
q = 1;ϕ(m|n) = pm(1−p)n−mCmn : Independent particles
µ = qν ;ϕ(m|n) = [n]q q-boson (Sasamoto,Wadati 98) andq-TASEP (Borodin Corwin, 2011)ν → µ = q, ϕ(m|n)' dt/[n]1/q MADM and and long rangehopping models, (Sasamoto Wadati, 1998; Alimohammadi,Karimipour, Khorrami, 1998)ν = 0, Geometric q-TASEP (Borodin Corwin, 2013)q→ 1,µ = qα ,ν = qα+β , ϕ(m|n) — Beta-Binomialdistribution (Barraquand, Corwin, 2015)
Generalized TASEP (M. Woelki, 2005, Derbyshev, Poghosyan, P.,Priezzhev, 2012)
q = 0,ϕ(m|n) =
(1−p), m = 0;pµm−1 (1−µ) , 0 < m < n;pµn−1 m = n,
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 36: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/36.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
q-Hahn process. Particular cases.
q = 1;ϕ(m|n) = pm(1−p)n−mCmn : Independent particles
µ = qν ;ϕ(m|n) = [n]q q-boson (Sasamoto,Wadati 98) andq-TASEP (Borodin Corwin, 2011)ν → µ = q, ϕ(m|n)' dt/[n]1/q MADM and and long rangehopping models, (Sasamoto Wadati, 1998; Alimohammadi,Karimipour, Khorrami, 1998)ν = 0, Geometric q-TASEP (Borodin Corwin, 2013)q→ 1,µ = qα ,ν = qα+β , ϕ(m|n) — Beta-Binomialdistribution (Barraquand, Corwin, 2015)
Generalized TASEP (M. Woelki, 2005, Derbyshev, Poghosyan, P.,Priezzhev, 2012)
q = 0,ϕ(m|n) =
(1−p), m = 0;pµm−1 (1−µ) , 0 < m < n;pµn−1 m = n,
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 37: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/37.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Bethe ansatz
Eigenvector
Ψ0(x|z) = ∑σ∈SN sgn(σ)∏Mi=1 ∏ j > i
uσi−quσjui−quj
(1−νuσi )xi
(1−uσi )xi
Eigenvalue
ΛN = ∏Ni=1
(1−µui1−νui
)Periodic boundary conditions
Ψ(x1, . . . ,xN |z) = Ψ(x2, . . . ,xN ,x1 +L|z).
Bethe equations(1−νui
1−ui
)L
= (−1)N−1N
∏j=1
ui −quj
uj −qui, i = 1, . . . ,N.
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 38: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/38.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Bethe ansatz
Eigenvector
Ψ0(x|z) = ∑σ∈SN sgn(σ)∏Mi=1 ∏ j > i
uσi−quσjui−quj
(1−νuσi )xi
(1−uσi )xi
Eigenvalue
ΛN = ∏Ni=1
(1−µui1−νui
)Periodic boundary conditions
Ψ(x1, . . . ,xN |z) = Ψ(x2, . . . ,xN ,x1 +L|z).
Bethe equations(1−νui
1−ui
)L
= (−1)N−1N
∏j=1
ui −quj
uj −qui, i = 1, . . . ,N.
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 39: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/39.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Bethe ansatz
Eigenvector
Ψ0(x|z) = ∑σ∈SN sgn(σ)∏Mi=1 ∏ j > i
uσi−quσjui−quj
(1−νuσi )xi
(1−uσi )xi
Eigenvalue
ΛN = ∏Ni=1
(1−µui1−νui
)Periodic boundary conditions
Ψ(x1, . . . ,xN |z) = Ψ(x2, . . . ,xN ,x1 +L|z).
Bethe equations(1−νui
1−ui
)L
= (−1)N−1N
∏j=1
ui −quj
uj −qui, i = 1, . . . ,N.
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 40: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/40.jpg)
ModelsLong-time behaviour of gTASEP.
SummaryZero-range chipping models with factorized steady stateIntegrability
Bethe ansatz
Eigenvector
Ψ0(x|z) = ∑σ∈SN sgn(σ)∏Mi=1 ∏ j > i
uσi−quσjui−quj
(1−νuσi )xi
(1−uσi )xi
Eigenvalue
ΛN = ∏Ni=1
(1−µui1−νui
)Periodic boundary conditions
Ψ(x1, . . . ,xN |z) = Ψ(x2, . . . ,xN ,x1 +L|z).
Bethe equations(1−νui
1−ui
)L
= (−1)N−1N
∏j=1
ui −quj
uj −qui, i = 1, . . . ,N.
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 41: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/41.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Two regimes
p = 0.1,µ = 0 p = 0.1,µ = 0,995
We expect change of behaviour the limit µ → 1:1−µ > 0 - KPZ-like behaviour, ∆∼ L−1/2
µ → 1 - DA limit (all particles stick together into a singlecluster, which moves diffusively)∆ = const
What is in between?
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 42: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/42.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Two regimes
p = 0.1,µ = 0 p = 0.1,µ = 0,995
We expect change of behaviour the limit µ → 1:1−µ > 0 - KPZ-like behaviour, ∆∼ L−1/2
µ → 1 - DA limit (all particles stick together into a singlecluster, which moves diffusively)∆ = const
What is in between?
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 43: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/43.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Stationary state
Consider a limit L→ ∞,M → ∞,M/L = c
Questions to answerCluster distribution.Particle current.Correlation length.
When µ = 1, there is a single cluster moving diffusively with thevelocity p. How this regime is approached?
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 44: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/44.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Partition function for ZRP-like model
Partition function (M paricles, N = L−M sites):
Z (M,N) = ∑n1,...,nN≥0
δ‖n‖,M
N
∏i=1
f (ni ) =∮
Γ0
[F (z)]N
zM+1 ,
where F (z) = ∑∞n=0 znf (n) .
Cluster size distribution:
P(n) = f (n)Z (M−n,N−1)
Z (M,N).
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 45: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/45.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Partition function for ZRP-like model
Partition function (M paricles, N = L−M sites):
Z (M,N) = ∑n1,...,nN≥0
δ‖n‖,M
N
∏i=1
f (ni ) =∮
Γ0
[F (z)]N
zM+1 ,
where F (z) = ∑∞n=0 znf (n) .
Cluster size distribution:
P(n) = f (n)Z (M−n,N−1)
Z (M,N).
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 46: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/46.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Particle current
Mean number of particles jumping per time step
J =N
Z (M,N)
∮Γ0
[F (z)]N
zMV ′(z)
V (z)
dz2π i
,
and
V (t) =1−νt1−µt
, W (t) =1−µt1− t
, F (t) = V (t)W (t) =1−νt1− t
are the generating functions of sequences w(k),v(k) and f (k)
v(k) = µk(δk,0 + (1−δk,0)(1−ν/µ)), (1)
w(k) = (δk,0 + (1−δk,0)(1−µ)), (2)f (n) = (δn,0 + (1−δn,0)(1−ν)) (3)
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 47: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/47.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Exact formuals
Z (M,N) =
(L−1M
)2F1 (−M,−N;1−L;ν),
J =(µ−ν)NM
(L−1)
F1(1−M;1−N,1;2−L;ν ,µ)
2F1 (−M,−N;1−L;ν).
Gauss hypergeometri function 2F1(a,b;c ;x) = ∑∞n=0
(a)n(b)n(c)nn! xn;
Appell hypergeometric functionF1(α;β ,β ′;γ;x ,y) = ∑
∞n,m=0
(α)m+n(β)m(β ′)n(γ)m+nm!n! xmyn
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 48: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/48.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Exact formuals
Z (M,N) =
(L−1M
)2F1 (−M,−N;1−L;ν),
J =(µ−ν)NM
(L−1)
F1(1−M;1−N,1;2−L;ν ,µ)
2F1 (−M,−N;1−L;ν).
Gauss hypergeometri function 2F1(a,b;c ;x) = ∑∞n=0
(a)n(b)n(c)nn! xn;
Appell hypergeometric functionF1(α;β ,β ′;γ;x ,y) = ∑
∞n,m=0
(α)m+n(β)m(β ′)n(γ)m+nm!n! xmyn
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 49: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/49.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Saddle point approximation
for integral of the form:
IN (h(z),g(z)) =∮
Γ0
eNh(z)g(z)dz2π iz
,
where h(z) = ln(1−νz)− ln(1− z)−ρ lnz .Critical point, h′(z−) = 0:
z− = 1+(1−ν)
2cν
(1−√
1+4(1− c)cν
1−ν
)
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 50: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/50.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Saddle point approximation
for integral of the form:
IN (h(z),g(z)) =∮
Γ0
eNh(z)g(z)dz2π iz
,
where h(z) = ln(1−νz)− ln(1− z)−ρ lnz .Critical point, h′(z−) = 0:
z− = 1+(1−ν)
2cν
(1−√
1+4(1− c)cν
1−ν
)
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 51: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/51.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Particle current
jASEP = limL→∞
J/L
=cp(1+ (1−2c)µ)
2µ +2c(p(1−µ)−µ)−
cp√
(1−µ)(1−4(1− c)c(p−µ)−µ)
2µ +2c(p(1−µ)−µ)
Μ=0
Μ=0.95
Μ=0.995
Μ=0.9995
0.2 0.4 0.6 0.8 1.0
0.1
0.2
0.3
0.4
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 52: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/52.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Particle current
jASEP = limL→∞
J/L
=cp(1+ (1−2c)µ)
2µ +2c(p(1−µ)−µ)−
cp√
(1−µ)(1−4(1− c)c(p−µ)−µ)
2µ +2c(p(1−µ)−µ)
Μ=0
Μ=0.95
Μ=0.995
Μ=0.9995
0.2 0.4 0.6 0.8 1.0
0.1
0.2
0.3
0.4
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 53: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/53.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Occupation number distribution in ZRP like system:
P(n) = zn− exp(−h(z−))/(1−ν) ,n > 0
P(0) = exp(−h(z−))
Cluster size distribution in gTASEP:
P(n) = zn− (1− z−)−1
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 54: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/54.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Occupation number distribution in ZRP like system:
P(n) = zn− exp(−h(z−))/(1−ν) ,n > 0
P(0) = exp(−h(z−))
Cluster size distribution in gTASEP:
P(n) = zn− (1− z−)−1
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 55: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/55.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Occupation number distribution in ZRP like system:
P(n) = zn− exp(−h(z−))/(1−ν) ,n > 0
P(0) = exp(−h(z−))
Cluster size distribution in gTASEP:
P(n) = zn− (1− z−)−1
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 56: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/56.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Validity range of saddle point approximation
Consider a limit
µ → 1,ν → 1,p =µ−ν
1−ν= const.
Letλ := (1−ν)−1→ ∞.
How large can it be for the saddle point analysis to be valid, givenhk ∼ λ
k−12 :
limN→∞
∣∣∣∣∣N1−k/2hk
hk/22
∣∣∣∣∣= 0⇒ λ/N2→ ∞.
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 57: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/57.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Transition regime,λN−2 = const
Deform contour Z (M,N) =−∮
Γ1eNh(z) dz
2π iz .
Choose the right integration scale z = 1+ eiϕ√ρλ
to get.
h (z) =−2√
ρ
λcosϕ +O(1/λ )
Then we obtain
Z (M,N) =−1√ρλ
∫ 2π
0e−2N
√ρ/λ cosϕ+iϕ dϕ
2π' θ
2MI1(θ),
where
θ = 2N√
ρ
λ
is the scaling parameter controlling the KPZ-DA transition andρ = N/L
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 58: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/58.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Transitional distribution over the macroscopic scale
Occupation number distribution:
P(0)' 1− θ
2NI0(θ)
I1(θ), P(M)' θ
2N1
I1(θ).
P (n)' θ2
4NMI1(θ√
1− nM
)I1 (θ)
√1− n
M, 0 < n < M,
Cluster fraction distribution (χ = n/M, M → ∞):
Prob(χ = 1) =1
I0(θ),
Prob(χ < x) =θ
2I0 (θ)
∫ x
0
I1(θ√1− y
)√1− y
dy .
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 59: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/59.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Transitional distribution over the macroscopic scale
Occupation number distribution:
P(0)' 1− θ
2NI0(θ)
I1(θ), P(M)' θ
2N1
I1(θ).
P (n)' θ2
4NMI1(θ√
1− nM
)I1 (θ)
√1− n
M, 0 < n < M,
Cluster fraction distribution (χ = n/M, M → ∞):
Prob(χ = 1) =1
I0(θ),
Prob(χ < x) =θ
2I0 (θ)
∫ x
0
I1(θ√1− y
)√1− y
dy .
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 60: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/60.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Transitional distribution over the macroscopic scale
Occupation number distribution:
P(0)' 1− θ
2NI0(θ)
I1(θ), P(M)' θ
2N1
I1(θ).
P (n)' θ2
4NMI1(θ√
1− nM
)I1 (θ)
√1− n
M, 0 < n < M,
Cluster fraction distribution (χ = n/M, M → ∞):
Prob(χ = 1) =1
I0(θ),
Prob(χ < x) =θ
2I0 (θ)
∫ x
0
I1(θ√1− y
)√1− y
dy .
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 61: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/61.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Large deviation function
Introduce deformed Markov matrix
Mγ
n,n′ = Mn,n′ exp(γN (n,n′)
),
where N (n,n′) is the number of particle jumps in theone-step transition from n′ to n.The log of its largest eigenvalue Λ0 (γ) is the rescaled cumulantgenerating function of total number of particle jumps
lnΛ0 (γ) = limt→∞
ln⟨eγYt
⟩t
.
Its Legendre transform is the large deviation function:
limt→∞
t−1 lnP(Yt/t > y) = supγ
(yγ− lnΛ0(γ))
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 62: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/62.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Large deviation function
Introduce deformed Markov matrix
Mγ
n,n′ = Mn,n′ exp(γN (n,n′)
),
where N (n,n′) is the number of particle jumps in theone-step transition from n′ to n.The log of its largest eigenvalue Λ0 (γ) is the rescaled cumulantgenerating function of total number of particle jumps
lnΛ0 (γ) = limt→∞
ln⟨eγYt
⟩t
.
Its Legendre transform is the large deviation function:
limt→∞
t−1 lnP(Yt/t > y) = supγ
(yγ− lnΛ0(γ))
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 63: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/63.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Large deviation function
Introduce deformed Markov matrix
Mγ
n,n′ = Mn,n′ exp(γN (n,n′)
),
where N (n,n′) is the number of particle jumps in theone-step transition from n′ to n.The log of its largest eigenvalue Λ0 (γ) is the rescaled cumulantgenerating function of total number of particle jumps
lnΛ0 (γ) = limt→∞
ln⟨eγYt
⟩t
.
Its Legendre transform is the large deviation function:
limt→∞
t−1 lnP(Yt/t > y) = supγ
(yγ− lnΛ0(γ))
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 64: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/64.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
What would we expect in scaling limit?
Large deviation hypothesis:
P(Yt/t > y)' exp(
taLz G
(y − yab
))
limt→∞
t−1 ln⟨eγYt
⟩= γy +aL−zG (γbLz), (L→ ∞,γbLz = const)
KPZ, z = 3/2, (Derrida-Lebowitz, 1998):
GDL(γ) =−Li5/2(B)
γ =−Li3/2(B)
DA limit, z = 2, CLT for random walk of particle of mass M,:
limt→∞
t−1 ln⟨eγYt
⟩= ln
(1−p+peγM
)'Mpγ +M2p(1−p)
γ2
2,
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 65: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/65.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Derrida-Lebowitz method
Bethe ansatz(1−νui
1−ui
)N
eNγ = (−1)M−1M
∏j=1
ui
uj,
Λ(γ) =M
∏i=1
(1−µui
1−νui
).
Reduction to polynomial:
P(u) = (1−νu)N B− (1−u)NuM ,
where B = (−1)M−1 eγN∏
Mj=1 uj .
Cauchy theoremM
∑i=1
f (uj) =∮
Γ0
f (u)P ′(u)
P(u)
du2π i
.
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 66: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/66.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Derrida-Lebowitz method
Bethe ansatz(1−νui
1−ui
)N
eNγ = (−1)M−1M
∏j=1
ui
uj,
Λ(γ) =M
∏i=1
(1−µui
1−νui
).
Reduction to polynomial:
P(u) = (1−νu)N B− (1−u)NuM ,
where B = (−1)M−1 eγN∏
Mj=1 uj .
Cauchy theoremM
∑i=1
f (uj) =∮
Γ0
f (u)P ′(u)
P(u)
du2π i
.
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 67: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/67.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Derrida-Lebowitz method
Bethe ansatz(1−νui
1−ui
)N
eNγ = (−1)M−1M
∏j=1
ui
uj,
Λ(γ) =M
∏i=1
(1−µui
1−νui
).
Reduction to polynomial:
P(u) = (1−νu)N B− (1−u)NuM ,
where B = (−1)M−1 eγN∏
Mj=1 uj .
Cauchy theoremM
∑i=1
f (uj) =∮
Γ0
f (u)P ′(u)
P(u)
du2π i
.
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 68: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/68.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Exact results.
Integral expressions
lnΛ0(γ) = (µ−ν)∮
Γ0
ln[1− B(1−νu)N
(1−u)NuM
](1−µu)(1−νu)
du2π i
.
γ =1−ν
M
∮Γ0
ln
(1− B(1−νu)N
(1−u)NuM
)(1−u)(1−νu)
du2π i
.
Series representations (term by term integration)
lnΛ0(γ) =−(µ−ν)∞
∑n=1
Bn
n
(Ln−2Mn−1
)F1 (1−nM;1−nN,1;2−nL;ν ,µ) ,
γ =−1−ν
M
∞
∑n=1
Bn
n
(Ln−1Mn−1
)2F1
(1−Mn,1−Nn ;1−nL;ν
).
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 69: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/69.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Exact results.
Integral expressions
lnΛ0(γ) = (µ−ν)∮
Γ0
ln[1− B(1−νu)N
(1−u)NuM
](1−µu)(1−νu)
du2π i
.
γ =1−ν
M
∮Γ0
ln
(1− B(1−νu)N
(1−u)NuM
)(1−u)(1−νu)
du2π i
.
Series representations (term by term integration)
lnΛ0(γ) =−(µ−ν)∞
∑n=1
Bn
n
(Ln−2Mn−1
)F1 (1−nM;1−nN,1;2−nL;ν ,µ) ,
γ =−1−ν
M
∞
∑n=1
Bn
n
(Ln−1Mn−1
)2F1
(1−Mn,1−Nn ;1−nL;ν
).
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 70: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/70.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Asymptotic forms
KPZ regime,λ/N2→ 0,λ 1/4N3/2γ = const
lnΛ(γ) = γJ∞ +aL−zGDA(γbLz),
a ∼ λ 1/4 and b ∼ λ−1/4as λ → ∞.
Transition regime
lnΛ(γ) = γpM +N−2p(1−p)Gθ (N2ργ),
Gθ (t) =θ2
4
∞
∑k=1
I2(kθ)Bk
k, t =−θ
2
∞
∑k=1
I1(kθ)Bk
k
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 71: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/71.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Cumulants
Mean current: J 'Mp−p(1−p)ρθ
2I2(θ)I1(θ) ,
Diffusion coefficient: ∆ = p(1−p)[
I1(2θ)I 21 (θ)
(I2(2θ)I1(2θ) −
I2(θ)I1(θ)
)]
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 72: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/72.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Cumulants
Cumulants in the transition regime scale as cn ∼ N2(n−1)
unlike cn ∼ N3/2(n−1) in the KPZ regime and cn ∼ Nn.c2
θ0 20 40 60 80 100
0.0
0.2
0.4
0.6
0.8
1.0
1.2
c3
θ20 40 60 80 100
-0.04
-0.03
-0.02
-0.01
0.00 c4
θ0 20 40 60 80 100
0.000
0.005
0.010
0.015
0.020
Universal cumulant
ratio:R(θ) =c23
c2c4=
(G
(3)θ
(0))2
G′′θ
(0)G(4)θ
(0)→ 2(3/2−8/33/2)
2
15/2−24/√
3+9/√
2' 0.41517
.
R(θ)
θ
20 40 60 80 100
0.1
0.2
0.3
0.4
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 73: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/73.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Stationary stateAsymptotic analysisFluctuations of particle current
Limiting forms of transitional LDF
Gθ (t)'−θ t2
+38
√θ
2πGDL
(t
√8π
θ
), θ → ∞
Gθ (t)' t2
2− θ2t
8, θ → 0
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming
![Page 74: Generalized TASEP on the edge of jamming. · Generalized TASEP on the edge of jamming. AlexanderPovolotsky Joint Institute for Nuclear Research, Dubna & Higher School of Economics,](https://reader030.vdocuments.us/reader030/viewer/2022041103/5f0275c17e708231d4045ecb/html5/thumbnails/74.jpg)
ModelsLong-time behaviour of gTASEP.
Summary
Summary
The KPZ breaks when the saddle point method fails.The KPZ scaling function keeps the form up to the diffusivescale, all the change being in model dependent constants.We obtained the LDF interpolating between Gaussian andDerrida-Lebowitz
Outlook
The transient regime (from Gauss to Tracy-Widom) — inprogressCombinatorial structure iof gTASEP. (Should the RSK bemodified?)
RPAC-2015 , Oxford 2015 Generalized TASEP on the edge of jamming