Prediction Of Filter Flow Rate Of Wafer Cleaning Machine Based On RBF Neural Networks Of Particle Swarm

Wafer cleaning technology is very important for electronic industry,especially for semiconductor industry.In the process of manufacturing integrated circuits and semiconductor products,almost every process involves cleaning.The cleaning quality of wafers is directly related to the yield of semiconductors and integrated circuits.As far as wafer cleaning technology is concerned,wet cleaning is the most commonly used cleaning technology.The key of wet cleaning technology is to control the flow rate and concentration of chemicals in the cleaning machine.By controlling the flow rate of the filter of the cleaning machine,the production efficiency of the product can be improved effectively,so as to avoid the waste of the cost caused by the premature replacement of the filter and the excessive or too little flow rate.Waste,reduce the rate of waste.In this paper,Particle Swarm Optimization(PSO)and improved RBF neural network(PSOIRBF)are used to predict filter flow.The main research work is as follows:(1)This paper introduces the background and application value of filter flow prediction for wafer cleaning machine,as well as several main wafer cleaning technologies at home and abroad at present,and briefly introduces the structure characteristics and working principle of wafer cleaning machine.(2)the structure,merits and demerits of all kinds of typical neural network prediction models,as well as the application results of the former people are simply analyzed.(3)the theoretical knowledge of RBF neural network is introduced,its advantages and disadvantages are analyzed,and the RBF neural network model is improved in order to solve the problems that RBF neural network is easy to fall into the local optimum and the precision is not high.(4)A particle swarm optimization(PSO)RBF neural network algorithm is proposed,and a nonlinear model is used to verify the algorithm,and the shortcomings of the algorithm are analyzed.(5)aiming at the shortcomings of RBF neural network based on particle swarmoptimization,the improved RBF neural network algorithm based on particle swarm optimization(PSO)is verified by nonlinear model,and the prediction results with different number of neurons are discussed.Compared with genetic algorithm and unimproved algorithm,the validity of the algorithm is verified.Finally,the model is applied to the flow prediction of the filter of wafer cleaning machine,and the simulation experiment is carried out.The applicability of the algorithm in the filter flow prediction of wafer cleaning machine is proved.