Automatic Synthesis of Switching Controllers for Linear Hybrid Automata

Massimo Benerecetti,Marco Faella,Stefano Minopoli
DOI: https://doi.org/10.48550/arXiv.1103.4584
2011-08-05
Abstract:In this paper we study the problem of automatically generating switching controllers for the class of Linear Hybrid Automata, with respect to safety objectives. We identify and solve inaccuracies contained in previous characterizations of the problem, providing a sound and complete symbolic fixpoint procedure, based on polyhedral abstractions of the state space. We also prove the termination of each iteration of the procedure. Some promising experimental results are presented, based on an implementation of the fixpoint procedure on top of the tool PHAVer.
Logic in Computer Science,Formal Languages and Automata Theory,Systems and Control,Optimization and Control
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to automatically generate a switching controller for Linear Hybrid Automata (LHA) to ensure the safety goals of the system. Specifically, the researchers aim to develop an algorithm that can calculate a strategy for a given safety goal, so that the controller can keep the system within the safe area in any situation (no matter how the continuous variables evolve or when the uncontrollable transitions occur). ### Problem Background 1. **Linear Hybrid Automata (LHA)**: - LHA is a dynamic system used to model the interaction between physical systems and digital controllers. - The state of the system consists of discrete states and continuous variables. The discrete states represent control modes, and the continuous variables represent physical quantities such as temperature, speed, etc. - In each discrete state, the evolution of continuous variables is controlled by the differential inclusion \( \dot{x} \in P \), where \( \dot{x} \) is the vector of first - order derivatives of all variables, and \( P \) is a convex polyhedron. 2. **Reachability Problem**: - For LHA, the reachability problem is to calculate the set of states that can be reached from the initial state at any time. - It has been proven that the reachability problem of LHA is undecidable, but the reachable state set can be accurately calculated within a finite number of steps (bounded - time - domain reachability). 3. **Switching Controller Synthesis**: - This paper focuses on the fact that the discrete transitions in LHA are divided into two categories: controllable and uncontrollable. - The goal is to calculate a strategy so that the controller can meet the given safety goal (i.e., keep the system within the safe area) in the case where the environment selects variable trajectories and uncontrollable transitions. ### Main Contributions of the Paper 1. **Proposing a New Symbolic Fixed - Point Process**: - A symbolic fixed - point process of the state space based on polyhedral abstraction is provided, which solves the inaccuracies in previous methods. - The termination of each iteration is proven. 2. **Improving the Reach While Avoiding (RWA) Operator**: - The RWA operator is used to calculate the regions that can be reached while avoiding certain regions. - The new algorithm is applicable to the case of non - convex regions, which is more common in practical applications. 3. **Experimental Results**: - The above - mentioned fixed - point process is implemented based on the tool PHAVer, and promising experimental results are shown. ### Summary This paper solves the problem of automatic generation of switching controllers in linear hybrid automata by proposing a new symbolic fixed - point process and an improved RWA operator, ensuring that the system can meet safety goals in any situation. This result provides a solid foundation for future research and a benchmark for the development of approximation techniques.