Offline Optimization Approach for Designing a Backstepping Speed Observer for PMSMs

This paper introduces an optimization-based approach for designing a backstepping rotor speed observer for permanent magnet synchronous machines (PMSMs). The approach addresses the challenge of determining suitable gains for convergence of stabilizing functions by using an offline optimization technique. The technique involves minimizing a cost function iteratively using a search algorithm for nonlinear functions. The cost function aims to reduce tracking errors of the observer. The proposed technique successfully determines the gains, which are validated through simulations, confirming the effectiveness of the observer design and the system's operation.