Laurent Perron

Search and strategies in OPL

OPL is a modeling language for mathematical programming and combinatorial optimization. It is the first language to combine high-level algebraic and set notations from mathematical modeling languages with a rich constraint language and the ability to specify search procedures and strategies that are the essence of constraint programming. This paper describes the facilities available in OPL to specify search procedures. It describes the abstractions of OPL to specify both the search tree (search) and how to explore it (strategies). The paper also illustrates how to use these high-level constructs to implement traditional search procedures in constraint programming and scheduling.

Constraint Programming in OPL

OPL is a modeling language for mathematical programming and combinatorial optimization problems. It is the first modeling language to combine high-level algebraic and set notations from modeling languages with a rich constraint language and the ability to specify search procedures and strategies that is the essence of constraint programming. In addition, OPL models can be controlled and composed using OPLSCRIPT, a script language that simplifies the development of applications that solve sequences of models, several instances of the same model, or a combination of both as in column-generation applications. This paper illustrates some of the functionalities of OPL for constraint programming using frequency allocation, sport-scheduling, and job-shop scheduling applications. It also illustrates how OPL models can be composed using OPLSCRIPT on a simple configuration example.

Structured vs. unstructured large neighborhood search: a case study on job-shop scheduling problems with earliness and tardiness costs

Large Neighborhood Search (LNS) [8] is a local search paradigm based on two main ideas to define and search large neighborhoods. The first key idea of LNS is to define its neighborhoods by fixing a part of an existing solution. The elements of the solution that are fixed are usually explicit or implicit variables of the model. For example, in a scheduling model, one may choose to fix the values of the start times of each activity (explicit variables) or one may add additional constraints that force one activity to be scheduled before another (implicit disjunctive variables). The rest of the variables are released: they are free to change values. The neighborhood is hence defined by all possible extensions of the fixed partial solution. Because a number of variables are released at a time, the neighborhoods defined are usually large, larger than typical local search neighborhoods.

Combining Forces to Solve the Car Sequencing Problem

Car sequencing is a well-known difficult problem. It has resisted and still resists the best techniques launched against it. Instead of creating a sophisticated search technique specifically designed and tuned for this problem, we will combine different simple local search-like methods using a portfolio of algorithms framework. In practice, we will base our solver on a powerful LNS algorithm and we will use the other local search-like algorithms as a diversification schema for it. The result is an algorithm is competitive with the best known approaches.

Solving a Network Design Problem

Industrial optimization applications must be “robust” i.e., they must provide good solutions to problem instances of different size and numerical characteristics, and continue to work well when side constraints are added. This paper presents a case study that addresses this requirement and its consequences on the applicability of different optimization techniques. An extensive benchmark suite, built on real network design data, is used to test multiple algorithms for robustness against variations in problem size, numerical characteristics, and side constraints. The experimental results illustrate the performance discrepancies that have occurred and how some have been corrected. In the end, the results suggest that we shall remain very humble when assessing the adequacy of a given algorithm for a given problem, and that a new generation of public optimization benchmark suites is needed for the academic community to attack the issue of algorithm robustness as it is encountered in industrial settings.

Alternate modeling in sport scheduling

Sports scheduling is a classical problem in the field of combinatorial optimization. One of the first successful methods to solve a complex instance was implemented using constraint programming. In this article, we explore an alternate and lighter way of modeling the round-robin part of the problem. We show this model can be enriched by additional propagations that complement the all different constraint.

Integration of AI and OR Techniques in Constraint Programming for Combinatorial Optimization Problems

Invited Talks.- Airline Scheduling: Accomplishments, Opportunities and Challenges.- Selected Challenges from Distribution and Commerce in the Airline and Travel Industry.- 30 Years of Constraint Programming.- Long Papers.- Constraint Integer Programming: A New Approach to Integrate CP and MIP.- New Filtering for the Constraint in the Context of Non-Overlapping Rectangles.- Multi-stage Benders Decomposition for Optimizing Multicore Architectures.- Fast and Scalable Domino Portrait Generation.- Gap Reduction Techniques for Online Stochastic Project Scheduling.- Integrating Symmetry, Dominance, and Bound-and-Bound in a Multiple Knapsack Solver.- Cost Propagation - Numerical Propagation for Optimization Problems.- Fitness-Distance Correlation and Solution-Guided Multi-point Constructive Search for CSPs.- Leveraging Belief Propagation, Backtrack Search, and Statistics for Model Counting.- The Accuracy of Search Heuristics: An Empirical Study on Knapsack Problems.- A Novel Approach For Detecting Symmetries in CSP Models.- Amsaa: A Multistep Anticipatory Algorithm for Online Stochastic Combinatorial Optimization.- Optimal Deployment of Eventually-Serializable Data Services.- Counting Solutions of Knapsack Constraints.- From High-Level Model to Branch-and-Price Solution in G12.- Simpler and Incremental Consistency Checking and Arc Consistency Filtering Algorithms for the Weighted Spanning Tree Constraint.- Stochastic Satisfiability Modulo Theories for Non-linear Arithmetic.- A Hybrid Constraint Programming / Local Search Approach to the Job-Shop Scheduling Problem.- Short Papers.- Counting Solutions of Integer Programs Using Unrestricted Subtree Detection.- Rapidly Solving an Online Sequence of Maximum Flow Problems with Extensions to Computing Robust Minimum Cuts.- A Hybrid Approach for Solving Shift-Selection and Task-Sequencing Problems.- Solving a Log-Truck Scheduling Problem with Constraint Programming.- Using Local Search to Speed Up Filtering Algorithms for Some NP-Hard Constraints.- Connections in Networks: A Hybrid Approach.- Efficient Haplotype Inference with Combined CP and OR Techniques.- Integration of CP and Compilation Techniques for Instruction Sequence Test Generation.- Propagating Separable Equalities in an MDD Store.- The Weighted Cfg Constraint.- CP with ACO.- A Combinatorial Auction Framework for Solving Decentralized Scheduling Problems (Extended Abstract).- Constraint Optimization and Abstraction for Embedded Intelligent Systems.- A Parallel Macro Partitioning Framework for Solving Mixed Integer Programs.- Guiding Stochastic Search by Dynamic Learning of the Problem Topography.- Hybrid Variants for Iterative Flattening Search.- Global Propagation of Practicability Constraints.- The Polytope of Tree-Structured Binary Constraint Satisfaction Problems.- A Tabu Search Method for Interval Constraints.- The Steel Mill Slab Design Problem Revisited.- Filtering Atmost1 on Pairs of Set Variables.- Extended Abstract.- Mobility Allowance Shuttle Transit (MAST) Services: MIP Formulation and Strengthening with Logic Constraints.