Traditional sliding mode observers (SMOs) are prone to chattering, and existing improvement methods have limitations, such as increasing system complexity or suboptimal performance. To overcome these issues, this paper proposes an SMO based on the cooperative optimization of a sinusoidal saturation function and a dynamic gain, focusing on the switching function and observer gain. In this observer, the sign function is replaced with a sinusoidal saturation function, effectively suppressing chattering and improving estimation accuracy. Moreover, the need for a low-pass filter (LPF) is eliminated, thereby avoiding phase lag issues. Additionally, a variable gain function was designed, which is jointly controlled by the velocity error and state variables. This further reduces oscillations and errors, thereby improving the system's stability. Through Lyapunov stability analysis, the stability of the proposed observer is verified. Simulation results show that, compared to existing SMOs based on sign and saturation functions, the method proposed in this paper results in smaller speed chattering and position errors.

This paper proposes an improved position estimation method based on low-resolution hall sensor to meet the high-precision position feedback requirement of vector control. Firstly, this paper research on vector control theory and implementation of id=0 based motor control methods. Secondly, the space vector pulse width modulation technique is used to obtain the PWM inverter switching signal. Thirdly, the position estimation circuit uses an improved method to obtain high-resolution position information from the signal of the low-resolution position sensor. Compared to the traditional estimation method, the proposed one has the advantages of initial rotor position estimation and adaptation to Hall mounting error. The proposed position estimation method is verified by simulation conducted on a 2000.rpm BLDCM and it is confirmed that the proposed drive system can achieve high-performance current vector control.

For the full-speed sensorless control of permanent magnet synchronous motors (PMSM) used in robotic joints, conventional methods adopt a combined strategy of zero-low-speed and medium-high-speed control methods. This requires switching between methods during motor operation, which reduces the reliability and stability, making it unable to be used in robotic joint motors. In this paper, a novel full-speed range sensorless control method is proposed. The method consists of two ranges: the zero-speed and the operation range. In the zero-speed range, a high-frequency square wave injection (HFSI) method with only 20 square-wave pulses (SWPs) is employed, resulting in a short injection duration, which ensures low noise. Leveraging the bidirectional convergence property of the linear time-invariant enhanced phase-locked loop (LTI-EPLL), the initial position and NS polarity can be determined without additional signal injection, ensuring smooth startup and rotor standstill. In the operation range, based on the eletrical properties of the PMSM, a sliding-mode speed observer (SMSO) is constructed to directly estimate the speed, and the position is obtained via integration. Thus, full-speed range operation is achieved without the need to switch between methods during operation range. Finally, the feasibility of this method is validated in MATLAB/Simulink.

In the application of coal mining belt conveyors, dual-motor drives based on permanent magnet motors (PMM) are gaining increasing attention now. To achieve high-performance coordination control of the two motors, this paper proposes a finite control set model predictive speed control (FCS-MPSC) method to improve the dynamics and speed tracking performance of the motors. First, the features of the dual-motor drives used in conveyor belts are analyzed. On this basis, the requirements of the control strategies are illustrated. Second, a master-slave control strategy is developed after treating the PMMs at the tail end and head end as the master motor and slave motor, respectively. Third, the FCS-MPSC method is developed for both master and slave motors by using new predicting model. In this process, the issue that the speed property is not directly related to the manipulated variables are tackled. Moreover, in order to further improve the dynamics of the slave motor, a speed reference compensation strategy is proposed. Finally, the proposed FCS-MPSC method is validated through comparative simulation results.

Reluctance machines with DC field coil in the stator is a competitive candidate for electric vehicle propulsion due to the elimination of rare-earth permanent magnets (PM), robust structures and controllable excitation. However, their torque density cannot be compared with PM machines. Introducing rare-earth PMs into reluctance machines can be an effective way to boost torque density, but the costs will greatly increase as well. To solve this problem, this paper proposes a hybrid reluctance motor using assistive ferrite magnets in stator slots. The introduced ferrite magnets operate based on the flux modulation effect for torque density improvement. It is revealed by finite element analysis, with ferrite magnets in stator slots, machine torque density and efficiency can be improved by 20.2% and 5.1%, respectively.