Dynamic and Stochastic Multi-Project Planning, 2015 Lecture Notes in Economics and Mathematical Systems Series, Vol. 673

This book deals with dynamic and stochastic methods for multi-project planning. Based on the idea of using queueing networks for the analysis of dynamic-stochastic multi-project environments this book addresses two problems: detailed scheduling of project activities, and integrated order acceptance and capacity planning. In an extensive simulation study, the book thoroughly investigates existing scheduling policies. To obtain optimal and near optimal scheduling policies new models and algorithms are proposed based on the theory of Markov decision processes and Approximate Dynamic programming. Then the book presents a new model for the effective computation of optimal policies based on a Markov decision process. Finally, the book provides insights into the structure of optimal policies.

1. Introduction.- 2. Problem Statements.- 3. Literature Review.- 4. Continuous-time Markov Decision Processes.- 5. Generation of Problem Instances.- 6. Scheduling Using Priority Policies.- 7. Optimal and Near Optimal Scheduling Policies.- 8. Integrated Dynamic Order Acceptance and Capacity Planning.- 9. Conclusions and Future Work.

Philipp Melchiors is a consultant for an Operations Research focused consulting company. Prior to his current position he worked as research and teaching assistant at the TUM School of Management, Technische Universität München. During this time he wrote his Ph.D. thesis on "Dynamic and stochastic multi-project planning".

Introduces new analytical models for optimal multi-project management based on decision rules in dynamic-stochastic environments Presents new insights into the structure of optimal policies Describes extensive experimental investigations into the performance of well-known heuristics for multi-project scheduling in dynamic-stochastic environments Introduces new approaches for high quality computing policies which outperform existing heuristic policies