Research On The Key Technologies Of Error Prediction And Compensation In Leadscrew Grinding

This paper comes from the ShanDong University-JiNing Cooperation Fund "The research and development of the CNC machine tool which include scroll motor system of high performance and the construction of the platform." The project is aimed at improving the grinding precision of the ball screw produced by WenShang screw Ltd., ShanDong Province. Ball screw is the key component of position or drive in CNC machine tool and other precise mechanical equipments. The improvement of its processing precision especially pitch accuracy has been the focus of our country’s research projects.Firstly, this paper introduce the advantages of improving processing accuracy with error compensation techniques and its application in screw processing in our country and foreign countries respectively. Then we summarize the advantages and disadvantages of them to provide reference for our further research. After that, we analyze the error resources in screw grinding and put an emphasize on the factors which influence the screw pitch error most and their impact. Next, we use analytical solution and ANSYS finite element method respectively to obtain the temperature field and thermal deformation in the grinding process. All of the results provide a basis for the error prediction and compensation in the following.Then the BP neural network is used to construct identification model for the prediction of the screw pitch error. The inputs are the difference of the first derivative of error at last time and the second one, also including the force deformation and thermal deformation at current time, the output is the current pitch error. The results of simulation show that this method is effective and feasible. Considering the transmission delay of the signal, fluctuation of the voltage signal, inertia of the mechanical system in actual processing, we come up with an adaptive control strategy for the smooth transition of compensation process and to reduce the influence caused by unknown status.This method and result could be used to the machining control of other parts, also it can provide theoretical support for the improvement of the accuracy of Precision machinery parts and components, which would lay a solid foundation for our county’s basic equipment manufacturing.