Research Of Oil Cooling Machine’s High Precision Temperature Control Algorithm

With the development of technology，the mold manufacturing precision demand ofthe device become higher and higher.Whether can it produce high quality moldsdepends on the CNC machining precision. The most important factor affecting theaccuracy of CNC machining is the error of processing temperature. According to therefrigeration principles of the oil-cooled machine, through the heat exchange mediumto stabilize the temperature of machine tools,the liquid refrigerant. Temperature-controlhas a non-linear, strong coupling, time-varying delay and other characteristics. It iseasy to realize that the accurate temperature control for improving performance of oilcooler is very important.This paper firstly analyzes the development status of oil cooler temperature controland the significance of the research, and then studies the circulation system ofrefrigeration oil cooler.The mathematical models of oil cooler part and the wholesystem are established by using mathematical tools, which finally get the temperaturecontrol function.The paper also researches the particle swarm algorithm, introduces shrinkagefactor to improve the performance of the particle swarm optimization algorithm. Thenthe paper uses the particle swarm optimization algorithm to tuning the PID controllerparameters, and gets the simulation and analysis of results.Finally, the paper presents anew PSO-RBF hybrid control algorithm: utilize particle swarm optimization algorithmto global optimization to tuning RBF neural network parameters, then utilize RBFneural network to tuning PID parameters, which combines the particle swarmoptimization outstanding global local search capabilities and performance optimizationRBF neural network. Paper describes the new algorithm of PID control parameterstuning step, and simulation comparison test. The results of simulation show that:compared to other PID parameter self-tuning algorithm, the PSO RBF networkparameters PID parameter tuning control algorithm works well, have good robustnessand adaptability, the new algorithm improved convergence time and accuracy of oilcooler temperature control.