Iterative Learning Contour Compensation Algorithm Based On Modified Reference Input

At present,the social demand of consumer products represented by mobile phones and tablets is increasing,and with the proximity of market saturation,the competition among products becomes more and more fierce.The processing of parts such as mobile glass,shell and so on of these products has the characteristics of large yield and high repeatability.In the current stage of product processing,each processing of the same product is treated as a new product processing,does not have the learning function,in order to improve the current processing conditions,so that product processing in the direction of intelligence,this paper introduced iterative learning algorithm as a control algorithm to guide the processing of products,and design contour error estimation model,The introduction of cross-coupling control makes the precision and efficiency of the product at high speed be improved.Modeling and identifying the structure of the servo control system in CNC machine tools,through the design of the three-ring control system,by optimizing the parameters of the position ring controller,the system is stable and has good repeatability performance.Good repeatability is the premise of the implementation of the Learning control algorithm,which can well reflect the characteristics of mass repetitive processing.The contour error characterizes the machining precision of the servo system of CNC machine tool,which is mainly caused by the feed rate of the system and the coupling coordination between the axes.For the compensation of the following error,the single-axis iterative learning algorithm can achieve the effect of learning compensation,while the contour error due to the existence of multi-axis coupling factors,simple iterative learning for the single-axis follow-up performance is improved,the improvement of contour error is not obvious,which requires the introduction of cross-coupling control.In this paper,an estimation model of contour error is established,and the contour error estimation model can be compared for different feed rates.Using the contour error information collected,the single-axis iterative algorithm is used in combination with multi-axis contour control to improve the synchronization of multi-axis,and then reduce the contour error of the system.For the existing high frequency noise information,the offline 0 phase filter is used for processing.Through the experiment of XY platform,the validity of the algorithm using cross-coupled iterative learning compensation based on modified reference input is verified.