Yan Chen

The electrification of agricultural machinery has become an important trend. Active Disturbance Rejection Control (ADRC) shows considerable potential in agricultural motor control due to its low model dependence and strong anti-disturbance capability. However, the Extended State Observer (ESO) of traditional ADRC is limited by bandwidth, making it difficult to effectively capture high-frequency impact disturbances such as torque fluctuations during straw cutting, which results in reduced efficiency and increased energy consumption. To address this, this paper proposes an improved ADRC scheme: designing a Super-Twisting Extended State Observer (STESO) by integrating Super-Twisting technology to enhance disturbance observation capability; incorporating a Quasi-Proportional Resonant (QPR) controller into the Error Feedback Control Law (SEF) to compensate for the shortcoming of disturbance suppression beyond ESO bandwidth; and proposing a decoupling strategy to reduce the difficulty of parameter tuning and optimize control performance. Simulations and experiments on the Permanent Magnet Synchronous Motor (PMSM) of an automatic seeder demonstrate that the proposed method can effectively suppress various disturbances, reduce speed regulation errors, and not deteriorate dynamic responses.

To enhance the electromagnetic performance of flat-wire permanent magnet synchronous motors, three different groove structures were designed for the rotor, and a multi-objective optimization algorithm combining a genetic algorithm (GA) with the TOPSIS method was proposed. Firstly, an 8-pole 48-slot flat-wire motor model was established, and the cogging torque was analytically calculated to compare the motor’s performance under different groove schemes. Secondly, global multi-objective optimization of the rotor groove dimensions was performed using a combined simulation approach involving Maxwell, Workbench, and Optislang, and the optimal rotor groove size structure was selected using the TOPSIS method. Finally, a comparative analysis of the motor’s performance under both rated-load and no-load conditions was conducted for the pre- and post-optimization designs, followed by verification of the mechanical strength of the optimized rotor structure. The research results demonstrate that the combined optimization approach utilizing the genetic algorithm and the TOPSIS method significantly enhances the torque characteristics of the motor. The computational results indicate that the average torque is increased to 165.32 N·m, with the torque ripple reduced from 28.37% to 13.32% and the cogging torque decreased from 896.88 mN·m to 187.9 mN·m. Moreover, the total distortion rates of the air-gap magnetic flux density and the no-load back EMF are significantly suppressed, confirming the rationality of the proposed motor design.