Based On The Cnc Machine Tool Thermal Error Compensation Of The Combined Neural Network Modeling Studies

This paper establishes the CNC machine thermal error mathematical model based on neural network, analyzes the current research status of CNC machine thermal error modeling at home and abroad, studies the CNC machine thermal error combination model using the improved pso and fuzzy theory combined neural net,and compares with the prediction effect with the two kinds of combination modeling method. The result shows that the effect of combined modeling method is significantly better than the neural network modeling method.The main research contents:1)Elaborate the CNC machine thermal error causes,the present research situation and methods of thermal error compensation technology,and points out the error modeling is the key factor of deciding compensation effect.2)Improve and derive the BP neural network model with particle swarm algorithm optimization,and compare with the matlab software simulation models.3)Simulate with the fuzzy theory and neural network model.Results show that using the fuzzy neural network for modeling has a good prediction effect.4)Use the Proteus simulation software to develop the CNC machine thermal error compensation temperature gathering system. Use Proteus software simulation developed CNC machine thermal error compensation temperature acquisition system,which put the AT89C51 as core,using the digital temperature sensor(DS18B20).Designed the system hardware,software and debug it .