Optimization of Timed Automata Models Using Mixed-Integer Programming
Research on optimization of timed systems, as e.g. for computing optimal schedules of manufacturing processes, has lead to approaches that mainly fall into the following two categories: On one side, mixed integer programming (MIP) techniques have been developed to successfully solve scheduling problems of moderate to medium size. On the other side, reachability algorithms extended by the evaluation of performance criteria have been employed to optimize the behavior of systems modeled as timed automata (TA). While some successful applications to real-world examples have been reported for both approaches, industrial scale problems clearly call for more powerful techniques and tools.
The work presented in this paper aims at combining the two types of approaches: The intention is to take advantage of the simplicity of modeling with timed automata (including modularity and synchronization), but also of the relaxation techniques and heuristics that are known from MIP. As a first step in this direction, the paper describes a translation procedure that automatically generates MIP representations of optimization problems formulated initially for TA. As a possible use of this translation, the paper suggests an iterative solution procedure, that combines a tree search for TA with the MIP solution of subproblems. The key idea is to use the relaxations in the MIP step to guide the tree search for TA in a branch-and-bound fashion.
KeywordsBranch-and-Bound Techniques Discrete Optimization Mixed-Integer Programming Scheduling Timed Automata
Unable to display preview. Download preview PDF.
- 4.Brooke, A., Kendrick, D., Meeraus, A., Raman, R.: GAMS - A User’s Guide. GAMS Development Corp., Washington (1998)Google Scholar
- 5.Bruns, R.: Scheduling. In: Baeck, T., Fogel, D.B., Michalewicz, Z. (eds.) Handbook of Evolutionary Computation, pp. F1.5: 1–9. Inst. of Physics Publishing (1997)Google Scholar
- 6.CPLEX. Using the CPLEX Callable Library. ILOG Inc., Mountain View, CA (2002)Google Scholar
- 7.Engell, S., Maerkert, A., Sand, G., Schultz, R.: Production planning in a multiproduct batch plant under uncertainty. In: Progress in Industrial Mathematics, pp. 526–531. Springer, Heidelberg (2002)Google Scholar
- 8.Fehnker, A.: Citius, Vilius, Melius - Guiding and Cost-Optimality in Model Checking of Timed and Hybrid Systems. Dissertation, KU Nijmegen (2002)Google Scholar
- 10.Groetschel, M., Krumke, S.O., Rambau, J. (eds.): Online Optimzation of Large Scale Systems. Springer, Heidelberg (2001)Google Scholar
- 17.Williams, H.P.: Model Building in Mathematical Programming. Wiley, Chichester (1999)Google Scholar