Active Disturbance Rejection Control of Five-Phase Motor Based on Parameter Setting of Genetic Algorithm

Abstract

Five-phase induction motors have the characteristics of high torque density, low torque ripple, and flexible control, making them suitable for medium- and low-voltage power supply situations. However, with the expansion of application scenarios, five-phase motors need to cope with increasingly complex operating conditions. Five-phase motors for propeller propulsion will face various complex sea conditions during actual use, and five-phase motors for electric vehicles will also face various complex road conditions and operating requirements during use. Therefore, as a propulsion motor, its speed control system must have strong robustness and anti-disturbance performance. The use of traditional PI algorithms has problems, such as poor adaptability and inability to adapt to various complex working conditions, but the use of an active disturbance rejection controller (ADRC) can effectively solve these problems. However, due to the significant coupling between the variables of induction motors and the large number of parameters in the ADRC, tuning the parameters of the ADRC is complex. Traditional empirical tuning methods can only obtain a rough range of parameter values and may have significant errors. Therefore, this paper uses ADRC based on genetic algorithm(GAADRC) to tune the parameters of the control and design an objective function based on multi-objective optimization. The parameters to be adjusted were obtained through multiple iterations. The simulation and experimental results indicate that GAADRC has lower startup overshoot, faster adjustment time, and lower load/unload speed changes compared to the empirically tuned PI controller and ADRC. Meanwhile, using a genetic algorithm for motor ADRC parameter tuning can obtain optimal control parameters while the control parameter range is completely uncertain; therefore, the method proposed in this paper has strong practical value.

Composite ADRC Speed Control Method Based on LTDRO Feedforward Compensation

“…Reference [ 7 ] employed the TD3 algorithm for the dynamic adjustment of controller parameters, thereby achieving innovative trajectory tracking for aircraft. GAADRC, introduced by reference [ 8 ], harnesses the genetic algorithm for the tuning of motor ADRC parameters through a multi-objective optimization-based target function, culminating in the identification of optimal ADRC control parameters after several iterations. An innovative ADRC design, grounded in IMA and aiming for the determination of optimal ADRC parameters, was unveiled by reference [ 9 ].…”