Modeling Of Spatial Contour Error And Compensation With Cross-Coupled Controller In Multi-axis Machining

The development of machining industry is mainly focused on high speed and high precision. And the spatial contour error would be increased as the feedrate increasing in the high-speed and high-precision machining process. Also, the contour distortion would be induced for the parameter mismatch between axes. All the problems above cannot be settled by the traditional servo control algorithm. So the factors relative with the change of spatial contour error in the machining is analyzed, and the cross-coupled control scheme which could reduce the error obviously is presented. The main contents are as follows:The research is first focused on the servo system of NC system. The relationship between the track-lag error of single axis and construction of the servo control system is obtained. Then the factors which affect the spatial contour error are elicited. From the results of the analysis, the spatial contour error in the multi-axis machining is influenced by the set value of feedrate and the coupled relationship between axes. But the primary factor is the feedrate. On the basis of the factors affecting the spatial contour error, a spatial contour error model based on the cross-coupled control is established, and an improvement of the model is also introduced to simple the calculation. Also, the improved model arithmetic can be used in the three-axis. Considering the nonlinear coupled relationship of the machine with rotational-axis, the error of the translational portions is calculated only. However, a proportional distribution method is used for the error compensation, since the method is simple to predigest the computational complexity and error is also effectively controlled.The emphasis is the parameter adjusting for the cross-coupled controller. Here an adaptive control scheme based on the neural network is used. And there are two types of network which is BP neural network and single neural cell. The advantage of the first is high precision, but the disadvantage is real-time quality caused by the complex calculation. The second one is just the opposite. It has a simple calculation but a less precise than the first and more precise than the cross-coupled controller using the classical PID scheme.The conclusion is that contour precision is increased much more by the cross coupled controller added on the servo system by the simulating circularity, corner, and spiral curves. And the spatial contour accuracy has got the much better optimization with the controller based on BP neural network.