Formal Abstraction of Stochastic Systems: Bringing Theory into Practice

Formal Abstraction of Stochastic Systems: Bringing Theory into Practice

Gevaerts, C.J.P.

Esmaeil Zadeh Soudjani, S. (mentor)
Abate, A. (mentor)

Mechanical, Maritime and Materials Engineering

Delft Center for Systems and Control

2014-01-24

Recent developed theories in the field of formal abstraction of stochastic systems have provided new methods for solving the safety and reach-avoid problems. These new methods have been extended to fit controlled stochastic systems, resulting in a policy that optimizes the safety and reach-avoid probabilities. The goal of this thesis is to put all these methods into practice by integrating them in one Graphical User Interface. Practical issues involving limitations on computation time and data storage are addressed. These issues have led to the creation of new theories regarding the abstraction process. This thesis puts forward innovative and efficient ways to calculate the abstraction error. This efficient implementation of the newly developed theories increases their applicability in academic research. A case study is presented to show the possibilities and validity of the developed Graphical User Interface.

Markov Chains
Formal Verification
Markov Processes
Abstractions
Approximations
Safety Problem
Reach-Avoid Problem

http://resolver.tudelft.nl/uuid:a040fce4-02f3-46b2-9d47-0c2052f38642

Student theses

master thesis

(c) 2014 Gevaerts, C.J.P.