algorithms for train scheduling on a single line
play

Algorithms for train scheduling on a single line Laurent DAUDET 1 PhD - PowerPoint PPT Presentation

Sep 02, 2022 •242 likes •1.62k views

Algorithms for train scheduling on a single line Laurent DAUDET 1 PhD advisor: Frdric MEUNIER 1 1 CERMICS, Centre dEnseignement et de Recherche en Mathmatiques et Calcul Scientifique, ENPC December 22nd, 2017 Laurent DAUDET PhD Defense

  1. One-hour schedules maximizing HGV shuttles Model The model Variables � � : scheduled departure times of all trains. d = d Eur j , d Fr j , d PAX , d HGV j j j : j th departure time of train of type A. → d A n : number of scheduled HGV shuttles. Mathematical model Max ( d , n ) ∈ X n where X = set of constraints → can be expressed with linear constraints (only non immediate constraint: security headway [Serafini and Ukovich, 1989]). ⇒ Mixed Integer Linear Program Solved by commercial solver CPLEX. Laurent DAUDET PhD Defense December 22nd, 2017 8 / 40

  2. One-hour schedules maximizing HGV shuttles Numerical results A current schedule Instance: 4 Eurostars, 5 PAX shuttles, and 1 freight train. Laurent DAUDET PhD Defense December 22nd, 2017 9 / 40

  3. One-hour schedules maximizing HGV shuttles Numerical results A current schedule Instance: 4 Eurostars, 5 PAX shuttles, and 1 freight train. ⇒ Maximum 4 HGV shuttles in the schedule Laurent DAUDET PhD Defense December 22nd, 2017 9 / 40

  4. One-hour schedules maximizing HGV shuttles Numerical results A current schedule Instance: 4 Eurostars, 5 PAX shuttles, and 1 freight train. ⇒ Maximum 4 HGV shuttles in the schedule Laurent DAUDET PhD Defense December 22nd, 2017 9 / 40

  5. One-hour schedules maximizing HGV shuttles Numerical results Some schedule improvements Parameters T : length of the period. L : 12-minute time-window. η : full minute discretization. C Eur : 30-minute gap between grouped Eurostars. Laurent DAUDET PhD Defense December 22nd, 2017 10 / 40

  6. One-hour schedules maximizing HGV shuttles Numerical results Some schedule improvements Parameters T : length of the period. L : 12-minute time-window. η : full minute discretization. C Eur : 30-minute gap between grouped Eurostars. Constraint relaxed Instance Improvement Length cyclic period T : 1 h → 4 h Up to 25% Loading platforms L : 12 min → 0 min Up to 50% Full minute discretization η : 1 min → 1 s Up to 33% Agreements with Eurostar C Eur : 30 min → [27 min-33 min] Up to 14% Laurent DAUDET PhD Defense December 22nd, 2017 10 / 40

  7. One-hour schedules maximizing HGV shuttles Numerical results Some other problems Compute one-hour schedules with minimum delays. Compute one-hour schedules with maximum number of HGV shuttles and minimum delays. → Stochastic Optimization, Sample Average Approximation. Laurent DAUDET PhD Defense December 22nd, 2017 11 / 40

  8. One-hour schedules maximizing HGV shuttles Numerical results Some other problems Compute one-hour schedules with minimum delays. Compute one-hour schedules with maximum number of HGV shuttles and minimum delays. → Stochastic Optimization, Sample Average Approximation. Laurent DAUDET PhD Defense December 22nd, 2017 11 / 40

  9. Joint scheduling and pricing problem Problem General context 1 One-hour schedules maximizing HGV shuttles 2 Joint scheduling and pricing problem 3 Minimizing the waiting time for a one-way shuttle service 4 Laurent DAUDET PhD Defense December 22nd, 2017 11 / 40

  10. Joint scheduling and pricing problem Problem Why such a problem? Departures and prices computed jointly in airline companies. → Increase of customers’ satisfaction and company’s revenue. Same objective for rail transportation. Toy problem to challenge this idea. Laurent DAUDET PhD Defense December 22nd, 2017 12 / 40

  11. Joint scheduling and pricing problem Problem Why such a problem? Departures and prices computed jointly in airline companies. → Increase of customers’ satisfaction and company’s revenue. Same objective for rail transportation. Toy problem to challenge this idea. Laurent DAUDET PhD Defense December 22nd, 2017 12 / 40

  12. Joint scheduling and pricing problem Problem Why such a problem? Departures and prices computed jointly in airline companies. → Increase of customers’ satisfaction and company’s revenue. Same objective for rail transportation. Toy problem to challenge this idea. Laurent DAUDET PhD Defense December 22nd, 2017 12 / 40

  13. Joint scheduling and pricing problem Problem Why such a problem? Departures and prices computed jointly in airline companies. → Increase of customers’ satisfaction and company’s revenue. Same objective for rail transportation. Toy problem to challenge this idea. Laurent DAUDET PhD Defense December 22nd, 2017 12 / 40

  14. Joint scheduling and pricing problem Problem The problem (1/2) One-way trip. Laurent DAUDET PhD Defense December 22nd, 2017 13 / 40

  15. Joint scheduling and pricing problem Problem The problem (1/2) One-way trip. Company wants to fix departures d of S trains and prices p . Each train has finite capacity C . Q customers want to purchase tickets for this trip. Buy tickets that satisfies them the most, or leave without purchasing. Laurent DAUDET PhD Defense December 22nd, 2017 13 / 40

  16. Joint scheduling and pricing problem Problem The problem (1/2) One-way trip. Company wants to fix departures d of S trains and prices p . Each train has finite capacity C . Q customers want to purchase tickets for this trip. Buy tickets that satisfies them the most, or leave without purchasing. Laurent DAUDET PhD Defense December 22nd, 2017 13 / 40

  17. Joint scheduling and pricing problem Problem The problem (1/2) One-way trip. Company wants to fix departures d of S trains and prices p . Each train has finite capacity C . Q customers want to purchase tickets for this trip. Buy tickets that satisfy them the most, or leave without purchasing. Laurent DAUDET PhD Defense December 22nd, 2017 13 / 40

  18. Joint scheduling and pricing problem Problem The problem (1/2) One-way trip. Company wants to fix departures d of S trains and prices p . Each train has finite capacity C . Q customers want to purchase tickets for this trip. → Buy tickets that satisfy them the most, or leave without purchasing. Laurent DAUDET PhD Defense December 22nd, 2017 13 / 40

  19. Joint scheduling and pricing problem Problem The problem (1/2) One-way trip. Company wants to fix departures d of S trains and prices p . Each train has finite capacity C . Q customers want to purchase tickets for this trip. → Buy tickets that satisfy them the most, or leave without purchasing. Objective Maximize the revenue of the company. Laurent DAUDET PhD Defense December 22nd, 2017 13 / 40

  20. Joint scheduling and pricing problem Problem The problem (2/2) Each customer i has a preferred departure time: random variable χ i belongs to economic class b i (e.g. business, tourist, low-cost, ...) “value of time” v b for economic class b Laurent DAUDET PhD Defense December 22nd, 2017 14 / 40

  21. Joint scheduling and pricing problem Problem The problem (2/2) Each customer i has a preferred departure time: random variable χ i belongs to economic class b i (e.g. business, tourist, low-cost, ...) “value of time” v b for economic class b Laurent DAUDET PhD Defense December 22nd, 2017 14 / 40

  22. Joint scheduling and pricing problem Problem The problem (2/2) Each customer i has a preferred departure time: random variable χ i belongs to economic class b i (e.g. business, tourist, low-cost, ...) “value of time” v b for economic class b Laurent DAUDET PhD Defense December 22nd, 2017 14 / 40

  23. Joint scheduling and pricing problem Problem The problem (2/2) Each customer i has a preferred departure time: random variable χ i belongs to economic class b i (e.g. business, tourist, ...) “value of time” v b for economic class b We assume v 1 ≤ v 2 ≤ · · · Customers of class 1 make their choice first, then 2, ... Laurent DAUDET PhD Defense December 22nd, 2017 14 / 40

  24. Joint scheduling and pricing problem Problem The problem (2/2) Each customer i has a preferred departure time: random variable χ i belongs to economic class b i (e.g. business, tourist, ...) “value of time” v b for economic class b We assume v 1 ≤ v 2 ≤ · · · Customers of class 1 make their choice first, then 2, ... Discrete choice model Each customer i and product j (departure d j at price p j ) → Random utility U ij ( d , p ) representing satisfaction. Laurent DAUDET PhD Defense December 22nd, 2017 14 / 40

  25. Joint scheduling and pricing problem Model The model (1/2) Denote by ξ vector representing uncertainty. Revenue of company → R ( d , p , ξ ) where ξ has been revealed. → Easy to compute (simulation, Linear Programming). Objective function → f ( d , p ) = E [ R ( d , p , ξ )] . → Remark: f ( d , p ) well defined (for all ( d , p ) ∈ X , R ( d , p , · ) measurable and E [ R ( d , p , ξ )] < ∞ ) and Var [ R ( d , p , ξ )] < ∞ . Laurent DAUDET PhD Defense December 22nd, 2017 15 / 40

  26. Joint scheduling and pricing problem Model The model (1/2) Denote by ξ vector representing uncertainty. Revenue of company → R ( d , p , ξ ) where ξ has been revealed. → Easy to compute (simulation, Linear Programming). Objective function → f ( d , p ) = E [ R ( d , p , ξ )] . → Remark: f ( d , p ) well defined (for all ( d , p ) ∈ X , R ( d , p , · ) measurable and E [ R ( d , p , ξ )] < ∞ ) and Var [ R ( d , p , ξ )] < ∞ . Laurent DAUDET PhD Defense December 22nd, 2017 15 / 40

  27. Joint scheduling and pricing problem Model The model (1/2) Denote by ξ vector representing uncertainty. Revenue of company → R ( d , p , ξ ) where ξ has been revealed. → Easy to compute (simulation, Linear Programming). Objective function → f ( d , p ) = E [ R ( d , p , ξ )] . → Remark: f ( d , p ) well defined (for all ( d , p ) ∈ X , R ( d , p , · ) measurable and E [ R ( d , p , ξ )] < ∞ ) and Var [ R ( d , p , ξ )] < ∞ . Laurent DAUDET PhD Defense December 22nd, 2017 15 / 40

  28. Joint scheduling and pricing problem Model The model (1/2) Denote by ξ vector representing uncertainty. Revenue of company → R ( d , p , ξ ) where ξ has been revealed. → Easy to compute (simulation, Linear Programming). Objective function → f ( d , p ) = E [ R ( d , p , ξ )] . → Remark: f ( d , p ) well defined (for all ( d , p ) ∈ X , R ( d , p , · ) measurable and E [ R ( d , p , ξ )] < ∞ ) and Var [ R ( d , p , ξ )] < ∞ . Laurent DAUDET PhD Defense December 22nd, 2017 15 / 40

  29. Joint scheduling and pricing problem Model The model (2/2) Mathematical model Max ( d , p ) ∈ X f ( d , p ) = E [ R ( d , p , ξ )] where R ( d , p , ξ ) is revenue and X is set of constraints. Laurent DAUDET PhD Defense December 22nd, 2017 16 / 40

  30. Joint scheduling and pricing problem Model The model (2/2) Mathematical model Max ( d , p ) ∈ X f ( d , p ) = E [ R ( d , p , ξ )] where R ( d , p , ξ ) is revenue and X is set of constraints. Compute d and p without knowing ξ ! Laurent DAUDET PhD Defense December 22nd, 2017 16 / 40

  31. Joint scheduling and pricing problem Methods Sample Average Approximation ( ξ 1 , . . . , ξ Ω ) of Ω independent and identically distributed realizations ⇒ ξ ω is not random variable! We approximate objective function f ( d , p ) = E [ R ( d , p , ξ )] by Laurent DAUDET PhD Defense December 22nd, 2017 17 / 40

  32. Joint scheduling and pricing problem Methods Sample Average Approximation ( ξ 1 , . . . , ξ Ω ) of Ω independent and identically distributed realizations ⇒ ξ ω is not random variable! We approximate objective function f ( d , p ) = E [ R ( d , p , ξ )] by Laurent DAUDET PhD Defense December 22nd, 2017 17 / 40

  33. Joint scheduling and pricing problem Methods Sample Average Approximation ( ξ 1 , . . . , ξ Ω ) of Ω independent and identically distributed realizations ⇒ ξ ω is not random variable! We approximate objective function f ( d , p ) = E [ R ( d , p , ξ )] by f Ω ( d , p ) = 1 ˆ � R ( d , p , ξ ω ) Ω ω Laurent DAUDET PhD Defense December 22nd, 2017 17 / 40

  34. Joint scheduling and pricing problem Methods Sample Average Approximation ( ξ 1 , . . . , ξ Ω ) of Ω independent and identically distributed realizations ⇒ ξ ω is not random variable! We approximate objective function f ( d , p ) = E [ R ( d , p , ξ )] by f Ω ( d , p ) = 1 ˆ � R ( d , p , ξ ω ) Ω ω New approximated optimization program Max ( d , p ) ∈ X ˆ f Ω ( d , p ) Laurent DAUDET PhD Defense December 22nd, 2017 17 / 40

  35. Joint scheduling and pricing problem Methods Sample Average Approximation ( ξ 1 , . . . , ξ Ω ) of Ω independent and identically distributed realizations ⇒ ξ ω is not random variable! We approximate objective function f ( d , p ) = E [ R ( d , p , ξ )] by f Ω ( d , p ) = 1 ˆ � R ( d , p , ξ ω ) Ω ω New approximated optimization program Max ( d , p ) ∈ X ˆ f Ω ( d , p ) We denote by → v ∗ = Max ( d , p ) ∈ X f ( d , p ) v Ω = Max ( d , p ) ∈ X ˆ → ˆ f Ω ( d , p ) Laurent DAUDET PhD Defense December 22nd, 2017 17 / 40

  36. Joint scheduling and pricing problem Methods SAA properties Proposition, Shapiro et al., 2009 We have (i) E [ˆ f Ω ( d , p )] = f ( d , p ) , for all ( d , p ) ∈ X , (ii) ˆ f Ω ( d , p ) converges to f ( d , p ) w.p. 1, for all ( d , p ) ∈ X , (iii) E [ˆ v Ω ] ≥ v ∗ , and v Ω converges to v ∗ w.p. 1. (iv) ˆ v Ω ] is an upper bound on v ∗ (iii). E [ˆ → With the Central Limit Theorem, we can compute (1 − α ) -confidence interval. For any solution (¯ p ) ∈ X , f (¯ p ) ≤ v ∗ � � d , ¯ d , ¯ = Max ( d , p ) ∈ X f ( d , p ) ⇒ lower bound on v ∗ . → With (i) and the Central Limit Theorem, we can compute (1 − α ) -confidence interval. Laurent DAUDET PhD Defense December 22nd, 2017 18 / 40

  37. Joint scheduling and pricing problem Methods SAA properties Proposition, Shapiro et al., 2009 We have (i) E [ˆ f Ω ( d , p )] = f ( d , p ) , for all ( d , p ) ∈ X , (ii) ˆ f Ω ( d , p ) converges to f ( d , p ) w.p. 1, for all ( d , p ) ∈ X , (iii) E [ˆ v Ω ] ≥ v ∗ , and v Ω converges to v ∗ w.p. 1. (iv) ˆ v Ω ] is an upper bound on v ∗ (iii). E [ˆ → With the Central Limit Theorem, we can compute (1 − α ) -confidence interval. For any solution (¯ p ) ∈ X , f (¯ p ) ≤ v ∗ � � d , ¯ d , ¯ = Max ( d , p ) ∈ X f ( d , p ) ⇒ lower bound on v ∗ . → With (i) and the Central Limit Theorem, we can compute (1 − α ) -confidence interval. Laurent DAUDET PhD Defense December 22nd, 2017 18 / 40

  38. Joint scheduling and pricing problem Methods SAA properties Proposition, Shapiro et al., 2009 We have (i) E [ˆ f Ω ( d , p )] = f ( d , p ) , for all ( d , p ) ∈ X , (ii) ˆ f Ω ( d , p ) converges to f ( d , p ) w.p. 1, for all ( d , p ) ∈ X , (iii) E [ˆ v Ω ] ≥ v ∗ , and v Ω converges to v ∗ w.p. 1. (iv) ˆ v Ω ] is an upper bound on v ∗ (iii). E [ˆ → With the Central Limit Theorem, we can compute (1 − α ) -confidence interval. For any solution (¯ p ) ∈ X , f (¯ p ) ≤ v ∗ � � d , ¯ d , ¯ = Max ( d , p ) ∈ X f ( d , p ) ⇒ lower bound on v ∗ . → With (i) and the Central Limit Theorem, we can compute (1 − α ) -confidence interval. Laurent DAUDET PhD Defense December 22nd, 2017 18 / 40

  39. Joint scheduling and pricing problem Methods A first heuristic: a sequential heuristic → Try to mimic natural way of scheduling and then fixing prices. 1. Compute departure times d with optimization problem maximizing utilities U ij with prices p = 0 . Laurent DAUDET PhD Defense December 22nd, 2017 19 / 40

  40. Joint scheduling and pricing problem Methods A first heuristic: a sequential heuristic → Try to mimic natural way of scheduling and then fixing prices. 1. Compute departure times d with optimization problem maximizing utilities U ij with prices p = 0 . � Max d ∈ X d E [ U ij ( d , 0 )] i , j Laurent DAUDET PhD Defense December 22nd, 2017 19 / 40

  41. Joint scheduling and pricing problem Methods A first heuristic: a sequential heuristic → Try to mimic natural way of scheduling and then fixing prices. 1. Compute departure times d with optimization problem maximizing utilities U ij with prices p = 0 . � Max d ∈ X d E [ U ij ( d , 0 )] i , j 2. Compute prices p with previous two-stage recourse program (with fixed departure times ˜ d ). Max p ∈ X p ˆ f Ω (˜ d , p ) Laurent DAUDET PhD Defense December 22nd, 2017 19 / 40

  42. Joint scheduling and pricing problem Methods A second heuristic: a Gauss-Seidel heuristic 1. Initialize ( d , p ) to some ( d 0 , p 0 ) . 2. Let m ∈ Z + and J 1 ∪ J 2 ∪ · · · ∪ J m be a partition of { 1 , . . . , S } . 3. Generate sequence of feasible solutions ( d k , p k ) : Laurent DAUDET PhD Defense December 22nd, 2017 20 / 40

  43. Joint scheduling and pricing problem Methods A second heuristic: a Gauss-Seidel heuristic 1. Initialize ( d , p ) to some ( d 0 , p 0 ) . 2. Let m ∈ Z + and J 1 ∪ J 2 ∪ · · · ∪ J m be a partition of { 1 , . . . , S } . 3. Generate sequence of feasible solutions ( d k , p k ) : Laurent DAUDET PhD Defense December 22nd, 2017 20 / 40

  44. Joint scheduling and pricing problem Methods A second heuristic: a Gauss-Seidel heuristic 1. Initialize ( d , p ) to some ( d 0 , p 0 ) . 2. Let m ∈ Z + and J 1 ∪ J 2 ∪ · · · ∪ J m be a partition of { 1 , . . . , S } . 3. Generate sequence of feasible solutions ( d k , p k ) : Laurent DAUDET PhD Defense December 22nd, 2017 20 / 40

  45. Joint scheduling and pricing problem Methods A second heuristic: a Gauss-Seidel heuristic 1. Initialize ( d , p ) to some ( d 0 , p 0 ) . 2. Let m ∈ Z + and J 1 ∪ J 2 ∪ · · · ∪ J m be a partition of { 1 , . . . , S } . 3. Generate sequence of feasible solutions ( d k , p k ) : � ∈ argmax d , p f � d k +1 J 1 , p k +1 � ( d , d k J 2 . . . , d k J m ) , ( p , p k J 2 . . . , p k � J m ) J 1 � ∈ argmax d , p f � d k +1 J ℓ , p k +1 � ( d k +1 J 1 , . . . , d , . . . , d k J m ) , ( p k +1 J 1 , . . . , p , . . . , p k J m ) � J ℓ Laurent DAUDET PhD Defense December 22nd, 2017 20 / 40

  46. Joint scheduling and pricing problem Methods A second heuristic: a Gauss-Seidel heuristic 1. Initialize ( d , p ) to some ( d 0 , p 0 ) . 2. Let m ∈ Z + and J 1 ∪ J 2 ∪ · · · ∪ J m be a partition of { 1 , . . . , S } . 3. Generate sequence of feasible solutions ( d k , p k ) : � ∈ argmax d , p f � d k +1 J 1 , p k +1 � ( d , d k J 2 . . . , d k J m ) , ( p , p k J 2 . . . , p k � J m ) J 1 � ∈ argmax d , p f � d k +1 J 2 , p k +1 � ( d k +1 J 1 , d , d k J 3 , . . . , d k J m ) , ( p k +1 J 1 , p , p k J 3 , . . . , p k J m ) � J 2 Laurent DAUDET PhD Defense December 22nd, 2017 20 / 40

  47. Joint scheduling and pricing problem Methods A second heuristic: a Gauss-Seidel heuristic 1. Initialize ( d , p ) to some ( d 0 , p 0 ) . 2. Let m ∈ Z + and J 1 ∪ J 2 ∪ · · · ∪ J m be a partition of { 1 , . . . , S } . 3. Generate sequence of feasible solutions ( d k , p k ) : � ∈ argmax d , p f � d k +1 J 1 , p k +1 � ( d , d k J 2 . . . , d k J m ) , ( p , p k J 2 . . . , p k � J m ) J 1 � ∈ argmax d , p f � d k +1 J 2 , p k +1 � ( d k +1 J 1 , d , d k J 3 , . . . , d k J m ) , ( p k +1 J 1 , p , p k J 3 , . . . , p k J m ) � J 2 . . . � ∈ argmax d , p f � d k +1 J m , p k +1 � ( d k +1 J 1 , . . ., d k +1 J m − 1 , d ) , ( p k +1 J 1 , . . ., p k +1 � J m − 1 , p ) J m Laurent DAUDET PhD Defense December 22nd, 2017 20 / 40

  48. Joint scheduling and pricing problem Results The results Academic instance: Realistic distribution for preferred departure times. Gumbel distribution for random part of utilities. Instance Lower bounds (20 min) Frontal SAA Seq. heur. G.-S. heur. Q S C 100 3 33 1159.6 ± 0.8 1162.7 ± 1.1 1165.5 ± 1.1 200 5 40 4290.7 ± 2.2 Laurent DAUDET PhD Defense December 22nd, 2017 21 / 40

  49. Joint scheduling and pricing problem Results The results Academic instance: Realistic distribution for preferred departure times. Gumbel distribution for random part of utilities. Instance Lower bounds (20 min) Frontal SAA Seq. heur. G.-S. heur. Q S C 100 3 33 1159.6 ± 0.8 1162.7 ± 1.1 1165.5 ± 1.1 200 5 40 4290.7 ± 2.2 Laurent DAUDET PhD Defense December 22nd, 2017 21 / 40

  50. Joint scheduling and pricing problem Results The results Academic instance: Realistic distribution for preferred departure times. Gumbel distribution for random part of utilities. Instance Lower bounds (20 min) Frontal SAA Seq. heur. G.-S. heur. Q S C 100 3 33 1159.6 ± 0.8 1162.7 ± 1.1 1165.5 ± 1.1 200 5 40 4290.7 ± 2.2 4353.1 ± 2.8 4461.1 ± 2.2 Laurent DAUDET PhD Defense December 22nd, 2017 21 / 40

  51. Joint scheduling and pricing problem Results The results Academic instance: Realistic distribution for preferred departure times. Gumbel distribution for random part of utilities. Instance Lower bounds (20 min) Frontal SAA Seq. heur. G.-S. heur. Q S C 100 3 33 1159.6 ± 0.8 1162.7 ± 1.1 1165.5 ± 1.1 200 5 40 4290.7 ± 2.2 4353.1 ± 2.8 4461.1 ± 2.2 Laurent DAUDET PhD Defense December 22nd, 2017 21 / 40

  52. Joint scheduling and pricing problem Results The results Academic instance: Realistic distribution for preferred departure times. Gumbel distribution for random part of utilities. Instance Lower bounds (20 min) Frontal SAA Seq. heur. G.-S. heur. Q S C 100 3 33 1159.6 ± 0.8 1162.7 ± 1.1 1165.5 ± 1.1 200 5 40 4290.7 ± 2.2 4353.1 ± 2.8 4461.1 ± 2.2 Instance Upper bounds (1 h) Frontal SAA Q S C 100 3 33 1847.4 ± 769.8 200 5 40 5957.1 ± 1339.7 Laurent DAUDET PhD Defense December 22nd, 2017 21 / 40

  53. Joint scheduling and pricing problem Results The results Academic instance: Realistic distribution for preferred departure times. Gumbel distribution for random part of utilities. Instance Lower bounds (20 min) Frontal SAA Seq. heur. G.-S. heur. Q S C 100 3 33 1159.6 ± 0.8 1162.7 ± 1.1 1165.5 ± 1.1 200 5 40 4290.7 ± 2.2 4353.1 ± 2.8 4461.1 ± 2.2 Instance Upper bounds (1 h) Frontal SAA Q S C 100 3 33 1847.4 ± 769.8 200 5 40 5957.1 ± 1339.7 Other heuristic: Lagrangian relaxation ⇒ does not improve Frontal SAA for big instances. Laurent DAUDET PhD Defense December 22nd, 2017 21 / 40

  54. Joint scheduling and pricing problem Results Conclusion Joint scheduling and pricing: higher revenues. Upper bounds: not very precise... Laurent DAUDET PhD Defense December 22nd, 2017 22 / 40

  55. Minimizing the waiting time for a one-way shuttle service Problem General context 1 One-hour schedules maximizing HGV shuttles 2 Joint scheduling and pricing problem 3 Minimizing the waiting time for a one-way shuttle service 4 Laurent DAUDET PhD Defense December 22nd, 2017 22 / 40

  56. Minimizing the waiting time for a one-way shuttle service Problem Why such a problem? Trucks arrive continuously on terminals. Huge waiting lines during peak hours. ⇒ Design schedules to decrease congestion (waiting times). Laurent DAUDET PhD Defense December 22nd, 2017 23 / 40

  57. Minimizing the waiting time for a one-way shuttle service Problem Why such a problem? Trucks arrive continuously on terminals. Huge waiting lines during peak hours. ⇒ Design schedules to decrease congestion (waiting times). Laurent DAUDET PhD Defense December 22nd, 2017 23 / 40

  58. Minimizing the waiting time for a one-way shuttle service Problem Why such a problem? Trucks arrive continuously on terminals. Huge waiting lines during peak hours. ⇒ Design schedules to decrease congestion (waiting times). Laurent DAUDET PhD Defense December 22nd, 2017 23 / 40

  59. Minimizing the waiting time for a one-way shuttle service Problem Why such a problem? Trucks arrive continuously on terminals. Huge waiting lines during peak hours. ⇒ Design schedules to decrease congestion (waiting times). Laurent DAUDET PhD Defense December 22nd, 2017 23 / 40

  60. Minimizing the waiting time for a one-way shuttle service Problem The problem One-way trip. Infinitesimal users arriving continuously (demand known in advance). Company wants to schedule S shuttles of capacity C . Peculiar loading process. ff Laurent DAUDET PhD Defense December 22nd, 2017 24 / 40

  61. Minimizing the waiting time for a one-way shuttle service Problem The problem One-way trip. Infinitesimal users arriving continuously (demand known in advance). Company wants to schedule S shuttles of capacity C . Peculiar loading process. ff Laurent DAUDET PhD Defense December 22nd, 2017 24 / 40

  62. Minimizing the waiting time for a one-way shuttle service Problem The problem One-way trip. Infinitesimal users arriving continuously (demand known in advance). Company wants to schedule S shuttles of capacity C . Peculiar loading process. ff Laurent DAUDET PhD Defense December 22nd, 2017 24 / 40

  63. Minimizing the waiting time for a one-way shuttle service Problem The problem One-way trip. Infinitesimal users arriving continuously (demand known in advance). Company wants to schedule S shuttles of capacity C . Peculiar loading process. ff Laurent DAUDET PhD Defense December 22nd, 2017 24 / 40

  64. Minimizing the waiting time for a one-way shuttle service Problem The loading process Laurent DAUDET PhD Defense December 22nd, 2017 25 / 40

  65. Minimizing the waiting time for a one-way shuttle service Problem The loading process Laurent DAUDET PhD Defense December 22nd, 2017 25 / 40

  66. Minimizing the waiting time for a one-way shuttle service Problem The loading process Laurent DAUDET PhD Defense December 22nd, 2017 25 / 40

  67. Minimizing the waiting time for a one-way shuttle service Problem The problem One-way trip. Infinitesimal users arriving continuously (demand known in advance). Company wants to schedule S shuttles of capacity C . Peculiar loading process. Laurent DAUDET PhD Defense December 22nd, 2017 26 / 40

  68. Minimizing the waiting time for a one-way shuttle service Problem The problem One-way trip. Infinitesimal users arriving continuously (demand known in advance). Company wants to schedule S shuttles of capacity C . Peculiar loading process. Objectives - Minimize maximum waiting time of users � Problem P max - Minimize average waiting time of users � Problem P ave → Waiting time: time between arrival time on terminal and departure of shuttle. Laurent DAUDET PhD Defense December 22nd, 2017 26 / 40

  69. Minimizing the waiting time for a one-way shuttle service Model The model (1/3) Variables d j : departure time j th shuttle. y j : cumulative loads for shuttles 1 to j . Parameters S : number of shuttles. C : capacity. ν : loading rate (loading x users takes a time ν x ). T : time horizon. D : [0 , T ] → R + : cumulative demand known a priori (oracle) → we assume that D ( · ) is upper semicontinuous. Laurent DAUDET PhD Defense December 22nd, 2017 27 / 40

  70. Minimizing the waiting time for a one-way shuttle service Model The model (1/3) Variables d j : departure time j th shuttle. y j : cumulative loads for shuttles 1 to j . Parameters S : number of shuttles. C : capacity. ν : loading rate (loading x users takes a time ν x ). T : time horizon. D : [0 , T ] → R + : cumulative demand known a priori (oracle) → we assume that D ( · ) is upper semicontinuous. Laurent DAUDET PhD Defense December 22nd, 2017 27 / 40

  71. Minimizing the waiting time for a one-way shuttle service Model The model (1/3) Variables d j : departure time j th shuttle. y j : cumulative loads for shuttles 1 to j . Parameters S : number of shuttles. C : capacity. ν : loading rate (loading x users takes a time ν x ). T : time horizon. D : [0 , T ] → R + : cumulative demand known a priori (oracle) → we assume that D ( · ) is upper semicontinuous. Laurent DAUDET PhD Defense December 22nd, 2017 27 / 40

  72. Minimizing the waiting time for a one-way shuttle service Model The model (2/3) Arrival time of user y → function τ ( · ) , pseudo-inverses of D ( · ) : τ ( y ) = inf { t ∈ [0 , T ]: D ( t ) ≥ y } . Objective function � y j 1 g ave ( d , y ) = � ( d j − τ ( y )) dy D ( T ) y j − 1 j Laurent DAUDET PhD Defense December 22nd, 2017 28 / 40

  73. Minimizing the waiting time for a one-way shuttle service Model The model (2/3) Arrival time of user y → function τ ( · ) , pseudo-inverses of D ( · ) : τ ( y ) = inf { t ∈ [0 , T ]: D ( t ) ≥ y } . Objective function � y j 1 g ave ( d , y ) = � ( d j − τ ( y )) dy D ( T ) y j − 1 j Laurent DAUDET PhD Defense December 22nd, 2017 28 / 40

  74. Minimizing the waiting time for a one-way shuttle service Model The model (2/3) Arrival time of user y → function τ ( · ) , pseudo-inverses of D ( · ) : τ ( y ) = inf { t ∈ [0 , T ]: D ( t ) ≥ y } . Objective function � y j 1 g ave ( d , y ) = � ( d j − τ ( y )) dy D ( T ) y j − 1 j Similar function g max ( d , y ) for maximum waiting time. Laurent DAUDET PhD Defense December 22nd, 2017 28 / 40

  75. Minimizing the waiting time for a one-way shuttle service Model The model (3/3) Min d , y g ( d , y ) s.t. y j − y j − 1 ≤ C y j − 1 ≤ y j ≤ d j − 1 d j y S = D ( T ) ≥ τ ( y j ) + ν ( y j − y j − 1 ) d j y 0 = 0 . Laurent DAUDET PhD Defense December 22nd, 2017 29 / 40

  76. Minimizing the waiting time for a one-way shuttle service Piecewise constant demand When demand is a step function Theorem Assume that D ( · ) is a step function defined with K discontinuities, and that ν = 0 . There is an algorithm computing an optimal solution of P ave in O ( K 2 S ) . Laurent DAUDET PhD Defense December 22nd, 2017 30 / 40

  77. Minimizing the waiting time for a one-way shuttle service Piecewise constant demand Precision on ν = 0 Laurent DAUDET PhD Defense December 22nd, 2017 31 / 40

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend

Module 6: CPU Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms

Module 6: CPU Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms

Module 6: CPU Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms Multiple-Processor Scheduling Real-Time Scheduling Algorithm Evaluation Silberschatz, Galvin, and Gagne 1999 Applied Operating System

731 views • 36 slides
TRISTAN 2016, Aruba, June 2016 1 Real-time train rescheduling Train scheduling : routing and

TRISTAN 2016, Aruba, June 2016 1 Real-time train rescheduling Train scheduling : routing and

TRISTAN 2016, Aruba, June 2016 1 Real-time train rescheduling Train scheduling : routing and scheduling a set of trains on a railway network (including stations) at planning level nominal timetable Train rescheduling : nominal train

250 views • 13 slides
CPU Scheduling - I Different Scheduling Algorithms FCFS SJF Priority RR

CPU Scheduling - I Different Scheduling Algorithms FCFS SJF Priority RR

CSE 421/521 - Operating Systems Roadmap Fall 2011 CPU Scheduling Basic Concepts Lecture - V Scheduling Criteria &amp; Metrics CPU Scheduling - I Different Scheduling Algorithms FCFS SJF Priority RR Tevfik

323 views • 5 slides
Chapter 6: CPU Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms

Chapter 6: CPU Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms

Chapter 6: CPU Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms Multiple-Processor Scheduling Real-Time Scheduling Algorithm Evaluation Operating System Concepts Silberschatz, Galvin and Gagne 2002

496 views • 17 slides
CPU Scheduling - I Different Scheduling Algorithms Tevfik Ko ar Louisiana State

CPU Scheduling - I Different Scheduling Algorithms Tevfik Ko ar Louisiana State

CSC 4103 - Operating Systems Roadmap Spring 2008 CPU Scheduling Basic Concepts Lecture - V Scheduling Criteria CPU Scheduling - I Different Scheduling Algorithms Tevfik Ko ar Louisiana State University January 29 th ,

425 views • 4 slides
Module 5: CPU Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms

Module 5: CPU Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms

' $ Module 5: CPU Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms Multiple-Processor Scheduling Real-Time Scheduling Algorithm Evaluation &amp; % Silberschatz and Galvin c Operating System Concepts

331 views • 22 slides
1 Last class: CPU Scheduling Today: CPU Scheduling Algorithms and Systems 2

1 Last class: CPU Scheduling Today: CPU Scheduling Algorithms and Systems 2

1 Last class: CPU Scheduling Today: CPU Scheduling Algorithms and Systems 2 Scheduling Algorithms First-come, First-serve (FCFS) Non-preemptive Does not account for waiting time (or much else) Convoy problem

593 views • 25 slides
Module 5: CPU Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms

Module 5: CPU Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms

' $ Module 5: CPU Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms Multiple-Processor Scheduling Real-Time Scheduling Algorithm Evaluation &amp; % Operating System Concepts 5.1 Silberschatz and

350 views • 11 slides
Scheduling Algorithms of deciding which process in ready queue should be allocated to CPU.

Scheduling Algorithms of deciding which process in ready queue should be allocated to CPU.

2/15/2011 Scheduling Algorithms CPU Scheduling algorithms deal with the problem Scheduling Algorithms of deciding which process in ready queue should be allocated to CPU. Following are the commonly used scheduling algorithms: Gursharan

329 views • 8 slides
Uniprocessor Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms Three

Uniprocessor Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms Three

Uniprocessor Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms Three level scheduling 2 1 Types of Scheduling 3 Long- and Medium-Term Schedulers Long-term scheduler Determines which programs are admitted to

553 views • 26 slides
Uniprocessor Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms 2

Uniprocessor Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms 2

Uniprocessor Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms 2 Three level scheduling 3 Types of Scheduling g yp Long- and Medium-Term Schedulers g Long-term scheduler Determines which programs are

907 views • 63 slides
Scheduling Algorithms Gursharan Singh Tatla professorgstatla@gmail.com www.eazynotes.com 1

Scheduling Algorithms Gursharan Singh Tatla professorgstatla@gmail.com www.eazynotes.com 1

Scheduling Algorithms Gursharan Singh Tatla professorgstatla@gmail.com www.eazynotes.com 1 15-Feb-2011 Scheduling Algorithms CPU Scheduling algorithms deal with the problem of deciding which process in ready queue should be allocated to

745 views • 46 slides
Clock-Driven Scheduling (in-depth) Precompute static schedule off-line Task Scheduler: (e.g.

Clock-Driven Scheduling (in-depth) Precompute static schedule off-line Task Scheduler: (e.g.

CPSC-663: Real-Time Systems Clock-Driven Scheduling Clock-Driven Scheduling (in-depth) Precompute static schedule off-line Task Scheduler: (e.g. at design time): can afford i := 0; k := 0; expensive algorithms. &lt;set timer to expire at

303 views • 7 slides
CPU Scheduling Schedulers in the OS Structure of a CPU Scheduler Scheduling =

CPU Scheduling Schedulers in the OS Structure of a CPU Scheduler Scheduling =

CPSC 410 / 611 : Operating Systems CPU Scheduling Schedulers in the OS Structure of a CPU Scheduler Scheduling = Selection + Dispatching Criteria for scheduling Scheduling Algorithms FIFO/FCFS SPF / SRTF

342 views • 16 slides
5 CPU Scheduling (1)

5 CPU Scheduling (1)

. . . . . . . . . Scheduling Algorithms Scheduling Criteria Basic Concepts . . . . 5 CPU Scheduling (1) 2009 09 01

706 views • 56 slides
CPU Scheduling Schedulers in the OS Structure of a CPU Scheduler Scheduling =

CPU Scheduling Schedulers in the OS Structure of a CPU Scheduler Scheduling =

CPSC 313: Intro to Computer Systems Intro to Scheduling CPU Scheduling Schedulers in the OS Structure of a CPU Scheduler Scheduling = Selection + Dispatching Criteria for scheduling Scheduling Algorithms FIFO/FCFS

863 views • 7 slides
Adoption of the Pennsylvania Avenue Alignment as the Preliminary Preferred Alternative for the

Adoption of the Pennsylvania Avenue Alignment as the Preliminary Preferred Alternative for the

Adoption of the Pennsylvania Avenue Alignment as the Preliminary Preferred Alternative for the Approach to the DTX Board Agenda Item 7 SAN N FRANC NCISCO COU OUNT NTY TRANS NSPORTATION N AUTHORITY September 11, 2 2018 The RAB Study

376 views • 8 slides
Co Corridors: s: Expanding the U.S. Passenger Market Amtrak + Rail Passengers Association

Co Corridors: s: Expanding the U.S. Passenger Market Amtrak + Rail Passengers Association

Developing Ne New Amtrak Co Corridors: s: Expanding the U.S. Passenger Market Amtrak + Rail Passengers Association Developing New Amtrak Corridors: Expanding the U.S. Passenger Market September 23rd 2:30 PM Eastern 2:30 PM

504 views • 47 slides
SafeComp 20, 15 Sept 2020 Model-Centered Assurance For Autonomous Systems Susmit Jha, John

SafeComp 20, 15 Sept 2020 Model-Centered Assurance For Autonomous Systems Susmit Jha, John

SafeComp 20, 15 Sept 2020 Model-Centered Assurance For Autonomous Systems Susmit Jha, John Rushby, N. Shankar Computer Science Laboratory SRI International Menlo Park, California, USA Jha, Rushby, Shankar; SR I Model-Centered Assurance 1

320 views • 21 slides
Intelligent vehicles and road transportation systems (ITS) Week 9 : Multi-vehicle cooperative and

Intelligent vehicles and road transportation systems (ITS) Week 9 : Multi-vehicle cooperative and

ME470 Intelligent vehicles and road transportation systems (ITS) Week 9 : Multi-vehicle cooperative and collaborative scenarios Denis Gingras January 2015 1 1-fvr.-15 D Gingras ME470 IV course CalPoly Week 9 Course outline Week 1

776 views • 63 slides
Concepts for In-Use Locomotive Regulation Workshop Day 1 October 29, 2020 Cmo escuchar la

Concepts for In-Use Locomotive Regulation Workshop Day 1 October 29, 2020 Cmo escuchar la

Concepts for In-Use Locomotive Regulation Workshop Day 1 October 29, 2020 Cmo escuchar la Listening to Language interpretacin de un idioma Interpretation 1. En los controles de la reunin o el seminario web, 1. In your meeting/webinar

750 views • 55 slides
Norio TOMII o Professor at Chiba Institute of Technology (Computer Science) Board member of Japan

Norio TOMII o Professor at Chiba Institute of Technology (Computer Science) Board member of Japan

Norio TOMII o Professor at Chiba Institute of Technology (Computer Science) Board member of Japan Transport Safety Board n Member of Council of Transport Policy n President of IAROR International Association of n Railway Operations Research

362 views • 3 slides
HY HYPE PERL RLOOP OOP: : CONN NNECTING ECTING THE HE GRE GREAT T LAK AKES ES ME

HY HYPE PERL RLOOP OOP: : CONN NNECTING ECTING THE HE GRE GREAT T LAK AKES ES ME

HY HYPE PERL RLOOP OOP: : CONN NNECTING ECTING THE HE GRE GREAT T LAK AKES ES ME MEGAR GAREGION EGION Lawrence Hall, Northeast Ohio Areawide Coordinating Agency (NOACA) Ohio Freight Conference August, 2018 ABOUT NOACA

475 views • 24 slides
Britain's New High Speed Railway Carole Bardell, Head of Health, Safety and Security, HS2 Ltd 12

Britain's New High Speed Railway Carole Bardell, Head of Health, Safety and Security, HS2 Ltd 12

Britain's New High Speed Railway Carole Bardell, Head of Health, Safety and Security, HS2 Ltd 12 th March 2018 Aims for Today Get you up to speed on: HS2 Phase One The way we work; Safe at heart and The Supply Chain Standard

464 views • 29 slides

More recommend