Research On Rotor Position Estimation Method Of Artificial Heart Pump Motor

According to the data in recent years,the growth rate of patients with heart failure is faster and faster,and the mortality rate is also increasing year by year.This is due to the particularity of the heart,which leads to many difficulties in the treatment of heart failure.Heart transplantation is an effective method.Artificial heart pump can effectively solve the problem of insufficient donor and save the lives of many patients with advanced heart disease.It has become an effective method to treat heart disease.Brushless DC motor is widely used in various fields because of its small size,light weight,high efficiency and reliable operation.In this paper,an integrated design of an axial flow heart pump,a rotating impeller and a permanent magnet is adopted,so that the permanent magnet can drive the blade to rotate,thus promoting blood flow.Obtaining rotor position information is an important part of motor control of cardiac pump,and it is also widely studied.Because of the special working environment of artificial heart pump and the high requirement of implantability,this paper adopts a sensorless motor control strategy and uses intelligent control method to estimate the rotor position.The main work is as follows:(1)To study the human blood circulation system,mainly including the human blood circulation mechanism,the natural heart working mechanism,find out the relationship between heart rate and motor speed,lay a physiological foundation for the next research work.At the same time,the structure,working principle and mathematical model of cardiac pump motor are introduced and analyzed.(2)The traditional back EMF zero-crossing method uses the opposite EMF zero-crossing method,which needs to construct the neutral point of the motor,resulting in errors.Therefore,this paper adopts the line back EMF zero-crossing method,which reduces the complexity of the peripheral circuit.The conclusion thatthe zero crossing point of line back EMF is the commutation point of motor is verified by simulation results.(3)A terminal sliding mode observer is designed to solve the phase delay problem of traditional sliding mode,which not only eliminates the use of low-pass filter in traditional sliding mode observer,reduces commutation error caused by phase delay,but also speeds up its convergence speed and achieves better tracking accuracy and control performance.The simulation results show that the terminal sliding mode observer can observe the rotor position more accurately,reduce the phase delay and reduce the chattering.(4)In order to overcome the influence of parameter uncertainties and external disturbances,the terminal sliding mode observer of RBF neural network is designed.Firstly,the terminal sliding mode surface is introduced,which has fast convergence and good observation accuracy,and reduces the phase lag problem.Secondly,the control strategy of the observer is designed by using RBF neural network,and the sliding mode variable is used as the input of the neural network and the output is control.The system strategy simplifies the control structure.RBF terminal sliding mode observer combines the advantages of RBF control and terminal sliding mode control,optimizes the control signal and weakens the chattering phenomenon.The simulation results show that the linear back EMF curve of the RBF terminal sliding mode observer is smoother and more accurate than that of the terminal sliding mode observer.The speed error estimation of the RBF terminal sliding mode observer is smaller and its performance is better.With the increase of the heart rate,the speed of the motor can be adjusted rapidly and has good dynamic performance,which is consistent with the physiological regulation mechanism of the human body.