Permutation Encoding for Automatic Reconstruction of Connections in Closed-Loop Control System using Evolutionary Algorithm

Search-based software engineering aims to apply different search-based techniques to software engineering problems. Automation of software development is one such problem. In this paper we evaluate the permutation-based individual encoding for automatic reconstruction of measurement connections in a closed-loop control system using evolutionary algorithm and model checking. Using the permutation-based encoding greatly increases the difficulty of the considered problem, but makes it much closer to the real world scenarios. The results show that even the simple (1+1) evolutionary algorithm can successfully solve the realistic optimization problem with large search space size, although it struggles to find the optimal solution within reasonable time on the hardest problem instance.