Abstract Cyber-physical systems (CPSs) are used in many safety-critical areas, making it crucial to ensure their safety. However, with CPSs increasingly dynamically deployed and reconfigured during runtime, their safety analysis becomes challenging. For one thing, reconfigurable CPSs usually consist of multiple agents dynamically connected during runtime. Their highly dynamic system topologies are too intricate for traditional modeling languages, which, in turn, hinders formal analysis. For another, due to the growing size and uncertainty of reconfigurable CPSs, their system models can be huge and even unavailable at design time. This calls for runtime analysis approaches with better scalability and efficiency. To address these challenges, we propose a scenario-based hierarchical modeling language for reconfigurable CPS. It provides template models for agent inherent features, together with an instantiation mechanism to activate single agent’s runtime behavior, communication configurations for multiple agents’ connected behaviors, and scenario task configurations for their dynamic topologies. We also present a path-oriented falsification approach to falsify system requirements. It employs classification-model-based optimization to explore search space effectively and cut unnecessary system simulations and robustness calculations for efficiency. Our modeling and falsification are implemented in a tool called . Experiments have shown that it can largely reduce modeling time and improve modeling accuracy, and perform scalable CPS falsification with high success rates in seconds.

Cyber-physical systems (CPSs) integrate discrete computations with continuous physical processes and can be highly nonlinear and nondeterministic. Unlike the verification of CPS, which is difficult to handle, the falsification of CPS fulfills certain requirements from testing by seeking witness behavior of these systems and is easier to conduct. However, existing falsification techniques may fail to support the general complex CPS in practice because they usually focus on certain restricted classes of systems. In this article, we present a path-oriented, derivative-free approach to falsify safety properties in nonlinear and nondeterministic CPS. In our approach, we model the behavior of CPS by hybrid automata. Then, we enumerate candidate paths of hybrid automata (HA), transform the feasibility of candidate paths into optimization problems, and solve these optimization problems by our newly proposed classification model-based, derivative-free optimization algorithm. We also provide two novel pruning techniques to further improve the efficiency and efficacy of our approach: 1) a nested optimization structure with better model refinements for continuous search space pruning and 2) a hardly feasible path prefixes guided backtracking for discrete search space pruning. We implement our approach into a tool called PDF. Our experiments showed that PDF supported the safety falsification of CPS in all of our benchmarks, and it achieved success rates no lower than 95% in only seconds on 22/28 of the benchmarks.

Hybrid systems are widely used in safety-critical areas. Hybrid optimal control synthesis, which aims to generate an optimal sequence of control inputs for a given task, is one of the most important problems in the field. The classical Gradient-based methods are efficient but they require the system under control should be differentiable. Sampling-based methods have no such limitations, but the ability of existing ones to solve complex control missions is restricted.