Switched Reluctance Motor has the advantages of simple structure,stable and durable,low speed operation and high torque,making it an ideal motor choice.However,due to the special double convex pole structure of SRM,the internal magnetic circuit is highly saturated and the "switch" form of the motor makes it very difficult to calculate the output torque,thus producing a large torque pulsation and limiting the application scope of the motor.This paper presents an enhancement of structure,a design of the drive circuit,and a control strategy for switching motor.Through the deep study of the four-phase 8/6pole switching reluctance motor,we can not only better understand its body structure,but also further explore its working mechanism,and use mathematical modeling technology to realize the motor modeling and control strategy optimization.Aiming at the defects of common torque modeling method in SRM,this paper proposes a torque model using neural network method.First,steady state analysis and torque data were obtained in ANSYS Maxwell finite element analysis software.An examination of the deficiencies of the conventional BP algorithm is conducted in this paper.Proposing to optimize the initial weights and thresholds of BP neural network by exploiting the rapid optimization features of northern goshawk search algorithm in a global context.The torque modeling of northern goshawk has been demonstrated to be capable of producing a swift network convergence rate and a remarkable level of fitting accuracy.Secondly,the direct torque control is used to subdivide the sector and optimize the zero vector magnetic chain,a new switch table is designed,and the table is used to build the system simulation model,study its influence on voltage vector switching,and analyze its effect on winding conversion.It is shown that sector refinement and DTC of trivector magnetic chains can effectively improve the efficiency of winding commutation and inhibit torque pulsation more effectively,thus improving the efficiency of winding commutation.For sliding,mode control can suppress torque pulsation,and we propose a modified direct torque control method for SRM.This method can effectively convert the speed difference into the expected torque,thus effectively suppressing the torque fluctuations.The feasibility of the scheme is verified by simulation experimentsFinally,by using the DSP hardware experiment platform,we can effectively test the feasibility and reliability of the proposed solution,so as to achieve the expected goal.The analysis of the acquired experimental waveforms demonstrates the utility of the proposed strategy for torque pulse inhibition in SRM. |