基于BP神经网络的磨床力误差补偿方法

为建立磨削加工参数与磨削力导致的力变形误差之间的关系模型,提出基于神经网络的力误差建模和实时补偿方法。建立经遗传算法优化的BP神经网络以表征磨削参数与磨削力的关系;运用有限元方法对零件进行力学分析,建立磨削力与力变形量的关系模型;建立加工参数与切削力误差映射模型,预测误差补偿量,进行实时补偿。实验结果表明:该切削力误差模型准确有效,具有较高的应用价值。

Force Error Compensation Method of Grinding Machine Based on BP Neural Network

In order to establish the relationship model between grinding parameters and deformation error of workpiece caused by grinding force, force error modeling and real time compensation method based on neural network and finite element simulation were proposed. A BP neural network optimized by genetic algorithm was established to characterize the relationship between grinding parameters and grinding force. Then, the mechanical analysis of the parts was carried out by using the finite element method, and the relationship model between grinding force and force deformation was established. Finally, the error mapping model of the machining parameters and cutting force was established to predict the amount of error compensation and make real time compensation. Experimental results show that the error model of the cutting force is accurate and effective, and has high application value.