Research On Control Of Electric Compressor Driven By Permanent Magnet Brushless DC Motor

Everything in present society wouldn't proceed without green development.And it is no exception in the process of reform and research of an automobile.As a necessary trend for the automobile industry,not only has the development of electric vehicles helped to reduce the consumption of petroleum resource,but also it has contributed to easing the increasingly serious pollution.Electric cars known as electric drives which has the advantages of economy,environmental protection,energy saving,low noise,simple structure convenient maintenance and so on.Air conditioning compressor is deemed as the core of the refrigeration system of automobile air conditioner,which is made up of three parts of a controler,a permanent magnet brushless DC motor(PMBDCM)and a compressor.What distinguishes PMBDCM from a traditional brush DC motor is its communication mode---non-mechanical electronic commutation.Therefore,it proves that the performance of both a brush DC motor and PMBDCM are excellent in speed control,the latter has more simple structure as a consequence of electronic commutation including reliable operation,a simple structure and low cost maintenance in practical application.However,a large number of documents show that the position sensors of the PMBDCM are deficient and cannot be applied to some special occasions.In order to solve the situation,the use of electronic circuit rather than the mechanical structure of PMBDCM's position sensorless control has become the essential method of motor control,in which the operating performance of the air-conditioning compressor in the electric car can be greatly improved.Hence,position sensorless control of PMBDCM is emphasis and difficulty of the air-conditioning compressor control.The main contents of this paper are as follows:1.This paper introduces the working principle of the electric compressor and then further analyzes the important part for its structure and working principle,introduces the mathematical model and two speed control strategy of PMBDCM.2.The basic principle of support vector machine is introduced.Support vector machine(SVM)method is a practical algorithm based on the statistics,by choosing the appropriate kernel function,the low-dimensional nonlinear function is mapped to the high-dimensional space,and the principle of structural risk minimization is used to solve the convex Secondary planning and dimensionality problem.Therefore for the multivariable,nonlinear and time-varying of PMBDCM position sensorless system,combined with support vector machine(SVM)for the regression function of nonlinear function,An intelligent control algorithm based on regression least square support vector machines(LSSVM)is proposed.3.The principle of rotor position detection is analyzed,The relationship between the phase voltage of the PMBDCM and the rotor position signal is derived from the voltage equation.So the position sensorless control problem is transformed into the nonlinear system identification process.4.The LSSVM model based on RBF kernel function is constructed to solve sensorless control problem of PMBDCM,However,the model has two parameters(?,?)to be determined,where ? controls the model complexity and approximation error,and the value is too large or too small will result in the sample data is too fitting or less fit.The reasonable determination of two parameter values is the key to solve the starter sensor starting control of PMBDCM.5.In order to solve the problem of the above two parameters,the paper introduces the principle of cross validation(CV)and particle swarm optimization(PSO)based on the regression type least squares support vector machine(LSSVM)model in the RBF kernel function,thus,the PSO-LSSVM model is constructed.In the MATLAB platform,the writing program combined with the simulation data to prove the PSO-LSSVM model can be controlled feasibility,and also achieve the precise control of PMBDCM sensorless starting from the theory.6.This paper introduces the traditional PID control algorithm,and then proposes a parameter self-tuning PID speed control method based on least squares support vector machines.The simulation module of PMBDCM control system is constructed by using Simulink simulation model library in MATLAB environment.The control model is double closed loop control structure: the current loop composed of hysteresis current controler is used as inner loop;the speed loop is constructed by the combination of PSO-LSSVM and PID controler as outer loop.In conclusion,this paper adopts motor starting control method based on KCV-PSO-LSSVM and speed control of the PID controler based on the improved SVM parameters and in the MATLAB platform and its simulink simulation platform to prove that both are feasible.Thus,the sensorless control of the electric vehicle air conditioner compressor based on PSO-LSSVM is realized.