Preface
Organization
Contents
Methodologies and Technologies
Multiple Heuristics with Reinforcement Learning to Solve the Safe Shortest Path Problem in a Warehouse
1 Introduction
2 Detailed Problem
2.1 A Warehouse and the Risk Induced by Current Activity
2.2 Our Problem: Searching for a Safe Shortest Path Inside the Warehouse
2.3 Some Structural Results
2.4 A Consequence: Risk Versus Distance Reformulation of the SSPP Model
2.5 Discussion About the Complexity
3 Solving the SSPP When the Path is Fixed
3.1 Generate Decisions with Three Methods
3.2 The Filtering Issue - Speeding up the Heuristic
4 Solving the SSPP - Proposed Algorithms
5 Speeding Algorithms Through Statistical Learning Techniques
5.1 Bounding the Number of States Generated
6 Numerical Experiments
7 Conclusion
References
Minimizing the Non-value Task Times: A Pickup and Delivery Problem with Two-Dimensional Bin-Packing
1 Introduction
2 Related Works
3 Case Study
3.1 Manufacturing Tool Repair Service
3.2 Solution Proposed
4 2D-BPP Model
5 Computational Results
5.1 Results
5.2 Model Efficiency
5.3 Integrated Solution with VRPSDPTW and 2D-BPP
6 Conclusions and Future Work
References
Predictive Maintenance Optimization Under Stochastic Production in Complex Systems
1 Introduction
2 Related Work
2.1 Maintenance Optimization in Complex Systems
2.2 Maintenance Optimization Methods
3 Problem Description and Formulation
3.1 DPMO and MILP Model
3.2 SPMO and Chance-constrained Programming Model
4 Approximations of the Chance-Constrained Model
4.1 Approximation Using Discrete Information
4.2 Approximation Using Continuous Information
4.3 Relationship Between the Approximation Approaches
5 Computational Experiments
5.1 Parameter Settings of Instances, Uncertainty, and Indicators
5.2 Comparison of Our Approaches
5.3 Impact of Probabilities in Chance Constraints
6 Conclusions and Future Work
References
Automated City Segmentation for Pollution Threshold Attribution: The Example of New Cairo
1 Introduction
2 Literature Review
3 Methodology
3.1 Dataset Formation
3.2 Sorting Technique
3.3 Increasing the Boundaries
3.4 Plotting on Maps
3.5 Pollution Measurement
4 Results
4.1 Area of Test
4.2 Datasets Formation
4.3 Sorting Technique
4.4 Increasing the Boundaries
4.5 Plotting on Maps
4.6 Air Pollution
4.7 Noise Pollution
4.8 Comparison of Measurements and Threshold
5 Conclusion
References
Customer Satisfaction and Company Revenue: A Solution Approach for the Passenger Management on Buses
1 Introduction
2 Problem Description
3 Dynamic Programming Formulation
4 Linear Programming Approximation
5 Booking Limit Policy
6 Numerical Results
7 Conclusions
References
Reinforcement Learning Algorithms: Categorization and Structural Properties
1 Introduction
2 Classes of MDP Problems and Suitable RL Algorithms
2.1 Model-Free and Model-Based RL
2.2 Policy Optimization and Value Learning
2.3 Finite and Infinite Horizon Problems
2.4 Discrete and Continuous State Spaces
2.5 Discrete and Continuous Action Sets
2.6 Discrete and Continuous Time Problems
3 Further Properties of RL Algorithms
3.1 On-Policy and Off-Policy RL
3.2 Deterministic and Stochastic Policies
3.3 Bias-Variance Tradeoff
3.4 Exploration-Exploitation Tradeoff
3.5 Hyperparameter Tuning and Robustness
3.6 Learning Stability
4 Summary
5 Conclusion
References
Comparison of Multi-objective Linear Programming Solutions Using Performance Metrics Based on Data Envelopment Analysis Models
1 Introduction
2 Applied Data Envelopment Analysis Models
2.1 Integer Slack-Based Measure Model (INT-SBM)
2.2 Super-Efficiency DEA Model (SE-DEA)
3 Performance Metrics Based on DEA Models
3.1 Cardinality Metric (CM)
3.2 Accuracy Metric (AC)
3.3 Diversity Metric (DM)
4 Analysis of MOLP Solutions Applying the Proposed DEA Metrics
4.1 Analysis of the Non-Convex Region of a Pareto Frontier
4.2 Performance Analysis of Pareto Frontiers for a Real Case Study
5 Conclusions
References
Comparing Power Flow Models in Tree Networks with Stochastic Load Demands
1 Introduction
2 Model Description
2.1 Queuing Model of EV-Charging
2.2 Distribution Network Model
2.3 Power Flow Equations
2.4 Summary
3 Analytical Results
3.1 Duality in Tree Networks
3.2 Inner Region of Feasible Region
4 Numerical Results
4.1 Line Network
4.2 Tree Network
5 Summary
6 Proofs
6.1 Proof of Lemma 1
6.2 Proof of Lemma 2
References
Robust Optimization for Operating Room Scheduling with Uncertain Surgical Durations: Impact of Risk-Aversion on Delay
1 Introduction
2 Literature Review
3 Operating Room Scheduling
3.1 Robust Optimization
3.2 Surgical Duration Uncertainty
4 Numerical Analysis
4.1 Data
4.2 Results
5 Conclusion
References
Matheuristic Local Search for the Placement of Analog Integrated Circuits
1 Introduction
2 Related Work
3 Problem Formulation
4 ILP Model and Extensions
4.1 Baseline Model
4.2 Improving the Performance of the Solver
5 Matheuristic as a Local Search
5.1 Intensification
5.2 Diversification
6 Experiments
6.1 Methodology and Data
6.2 Performance with Redundant Constraints
6.3 Matheuristics on Synthetic Data
6.4 Improvement on Real Life Instances
7 Conclusion
References
Applications
An Urban-Scale Application of the Problem of Designing Green Tourist Trips with Time Windows
1 Introduction
2 Literature Review
3 A Multi-criteria Method for the POI Evaluation
4 Mathematical Formulation
5 Case-Study Setting and Computational Results
6 Conclusion
References
Optimization of the Storage Location Assignment Problem Using Nested Annealing
1 Introduction
2 Literature Review
3 Simulated Annealing
4 Problem Formulation
4.1 Objective Function
4.2 Picking-Log Distance
4.3 Reassignment Distance
5 Optimization Algorithm
5.1 Assignment Sampling Using Markov Chain Monte Carlo (MCMC) and Hamming Distances
5.2 TSP Optimization and Cost Caching
5.3 Heatmap-Based Approximation
5.4 Nested Annealing
5.5 Restarts
6 Experiments
6.1 Overview
6.2 Parameters
6.3 Datasets
6.4 Experiment Results
7 Conclusion
A Appendix
References
Inventory Management Optimization in Multi-Stage Supply Chains Under Uncertainty
1 Introduction
2 Mathematical Background
2.1 B-Spline Functions ch13DeB78
2.2 The CRLS Problem
3 The MSSC Model
4 The DRMPC Approach
4.1 Local MMCOP for Ai
4.2 Some Remarks on the Cost Functional Ji,k
4.3 The Constraints u-i,k and u+i,k
5 Reformulation of the MMCOP
6 Feasibility and Stability of the DRMPC
7 Simulation Results
8 Concluding Remarks
References
Author Index