Research On Thermal Error Compensation Of CNC Machine Tools Based On Comprehensive Intelligence

The error compensation method has the characteristics of low cost and easy implementation.Therefore,it becomes an important method to improve the machining accuracy of CNC machine tools.In recent years,the vast number of experts both at home and abroad have gained a certain degree of progress,but the precise construction of the error compensation model is still the most important factor affecting the error compensation effect,and it has become the bottleneck of the machining precision of CNC machine tools.The data of thermal error has the characteristics of correlation and nonlinearity.The least square method used now is difficult to fit it effectively.In order to improve the accuracy of the mathematical model,the BP neural network is used to fit the nonlinear data.The rough set is used to reduce the related data.They can effectively solve the precise establishment of mathematical model in the error compensation method.Because the BP network is prone to fall into the local optimal value in the learning process,it combines the ant colony algorithm with the network to achieve the global optimal value.Thermal error compensation method for machine tools based on the rough set theory,the ant colony algorithm and the BP neural network is proposed.The main tasks of this article are as follows:The basic knowledge of rough set theory,ant colony algorithm and artificial neural network,as well as the application in error compensation are analyzed,and the combination of the three is carefully analyzed.The error sources of NC machine tools are analyzed.In view of the coupling characteristics between error sources,rough set theory is used to reduce the input characteristics of error sources.In view of the shortcomings of BP network in the process of learning,the initial weights and thresholds of BP network are assigned by ant colony algorithm,and a new network learning method based on ant colony learning algorithm is proposed.Finally,the learned rules are applied to CNC machine tools.Innovative contributions of this article are as follows:1 By using the rough set theory to reduce the number of input variables of the BP network,the structure of the network is simplified and the work efficiency of the network is improved.At the same time,the installation of the hardware is also convenient and the cost is reduced.2 Ant colony algorithm is used to train BP network.The training times of BP network are shortened greatly,and the prediction accuracy of network is greatly improved.3 Embedding the law contained in weights and thresholds in machine tools can be recognized in the machine tool,so it has practical application value.