Adaptive Servo System Parameter Tuning with Success-Driven Differential Evolution

Addressing issues in system instability and low efficiency during the parameter tuning process of servo systems using traditional optimization algorithms, this paper proposes an improved Differential Evolution (DE) algorithm. Although the traditional Relay Feedback Method (RFB) can obtain theoretical parameters, it is susceptible to irregular disturbances within the system, leading to parameter fluctuations and affecting tuning accuracy. To enhance tuning efficiency and system robustness, this study employs the RFB results as the initial population for the improved DE algorithm and introduces a Success-Driven Mutation Strategy (SDMS) to accelerate algorithm convergence. The SDMS records and reuses historically successful mutation steps to guide the algorithm in searching along effective directions in subsequent iterations. Through simulation experiments, the improved DE algorithm’s rapid convergence and the effectiveness of the optimized parameters are validated. The algorithm significantly enhances the system’s dynamic performance and disturbance rejection capabilities, demonstrating its potential value in practical engineering applications.