Developing Deep Neural Net Controllers to Assure System Stability with Non-Zero Equilibrium Points
Conference
He, X. (2024). Developing Deep Neural Net Controllers to Assure System Stability with Non-Zero Equilibrium Points
. Proceedings of the International Conference on Software Engineering and Knowledge Engineering, SEKE, 289-294. 10.18293/SEKE2024-002
He, X. (2024). Developing Deep Neural Net Controllers to Assure System Stability with Non-Zero Equilibrium Points
. Proceedings of the International Conference on Software Engineering and Knowledge Engineering, SEKE, 289-294. 10.18293/SEKE2024-002
Cyber-physical systems (CPS) have become increasingly important in the functioning of our society. In recent years, machine learning (ML) approaches start to become an attractive choice to design CPS controllers for better performance and adaptability. How to assure the correctness of this type of new controllers is extremely difficulty and remains a grand research challenge. This paper presents a convex optimization-based technique for computing the stability regions of deep neural net (DNN) controllers for CPS. The technique has been successfully applied to three benchmark systems.
