State space quantitative analysis of timed coloured petri nets for scheduling improvements
- Autores: Miguel Antonio Mujica Mota
- Directores de la Tesis: Miguel Angel Piera (dir. tes.)
- Lectura: En la Universitat Autònoma de Barcelona ( España ) en 2011
- Idioma: español
- Tribunal Calificador de la Tesis: Idalia Flores de la Mota (presid.), Juan José Ramos González (secret.), Gasper Music (voc.)
- Materias:
- Texto completo no disponible (Saber más ...)
- Resumen
Digital simulation is a well accepted approach which applied in industrial scenarios results suitable to avoid the expenses that would be caused by experimenting with the real systems. This approach carries out experiments over computer models that represent the system in order to obtain information such as the values of performance indicators or configurations that improve the performance of the real system. The information obtained from the experiments facilitates the analysts to understand the cause and effect relationships of the processes within the system so that it is possible to take decisions that improve the performance of the system under study.
In recent years simulation has evolved in order to pursue a wedding with optimization techniques. In particular focus has been put on the coloured Petri net formalism because it has capabilities that make it suitable to model concurrent systems and it facilitates the understanding of the causality present in industrial systems. The reachability graph or state space is a tool that has been traditionally used to verify models' properties in order to determine the behaviour of the modelled system. Furthermore, the reachability graph can also be used with industrial models as the search space for determining the different configurations of the modelled system. This tool allows not only to verify the properties of the model but also to search for particular configurations that result useful from the engineering point of view. Particularly it can be used to determine the configurations that improve performance values of manufacture or logistic models such as makespan, total tardiness, costs, etc. Therefore the use of state space analysis together with the timed coloured Petri net modelling formalism allows taking advantage of the abstraction capabilities of the formalism and the analytical potential of the state space analysis. If the latter approach is properly implemented, decision support tools can be developed in industrial environments in order to manage common value-adding activities such as scheduling, resource planning or resource allocation procedures.
The work towards the PhD has been developed focusing on the development of state space analysis techniques for timed coloured Petri net models in order to obtain results in short time.
The implementation of key techniques such as the ones developed in the transition evaluation and the information storage algorithms allowed the development of an efficient algorithm which generates and stores the information produced during the state space analysis of timed coloured Petri net models. The compact timed state space is fundamental for the development of the analysis algorithms presented in the articles of this thesis; all the developed techniques took into account the type of state space implemented in this work. All the implementations were experimentally validated using very well known academic benchmarks and the obtained results were satisfactorily compared to available implementations.
