Research On Torque Ripple Reduction Method Of Induction Motor For Electric Vehicle Based On Random Position Pulse Width Modulation

As a new energy vehicle,electric vehicle is developing rapidly in the field of automobile industry,and it has the trend of replacing fuel vehicle gradually.With the advantages of simple structure,high efficiency and low cost,asynchronous motor is widely used as vehicle motor of electric vehicle.However,the torque ripple has a great impact on the development of electric vehicles.In order to solve the problem of torque ripple of induction motor for electric vehicle,this paper starts with the frequency spectrum distribution of harmonics,and adopts the technology of random PWM pulse width modulation,so that the cluster harmonic energy output by inverter can be evenly distributed in a wide frequency band.The results show that this method can effectively suppress harmonic and torque ripple.The main contents of this paper are as follows:Firstly,the vector control model of induction motor is established,and the reason of harmonic cluster in the integral frequency multiplication of switching frequency of SVPWM is analyzed.The output torque model is established and the influence of harmonics on torque ripple is studied.Secondly,three different random PWM modulation strategies are analyzed and compared.According to the advantages and disadvantages of three different random PWM modulation strategies,the appropriate random PWM modulation strategy is selected for this method.Thirdly,the influence of different random number probability distribution on random number generation is studied,several methods of random number generation are analyzed,and a random number generator based on Markov chain is established,which is compared with the ordinary random number generator,and the superiority of the generator is verified.Then,a kind of random pulse position SVPWM PWM strategy based on Markov chain is introduced,the influence of different random number range on harmonic suppression is analyzed,three different random pulse position structures are designed,the influence of different random pulse position SVPWM on time harmonic is studied,and compared with traditional SVPWM,the correctness of the method is verified.At the same time,in order to improve the speed of the algorithm,this paper introduces a random pulse position SVPWM modulation strategy based on neural network.Finally,the experiment platform is built to compare the traditional SVPWM and the random pulse position SVPWM based on neural network.The results verify the correctness of the theoretical analysis and simulation model.