Research On Machining Trajectory Control And Experimental Method Of Aspheric Parts

With the rapid development of modern high technology,aspheric optical components are applied more and more widely.At the same time,higher requirements are put forward for the processing accuracy and efficiency of aspheric parts.In fact,the aspheric parts processing technology with high efficiency and high quality has become a hot research direction in the world.The core of ultra-precision machining technology for aspheric optical parts is the subroutine of controlling the processing trajectory,which directly affects the processing accuracy and the cycle of machine tools and plays a vital role in achieving high-efficiency and high-precision processing of aspheric parts.Therefore,in order to solve the problems of low efficiency and low precision of aspheric surface processing,the tangent trajectory forming principle has been proposed.Since the existing NC system can not meet the requirement of tangent method in the fitting accuracy and data processing,the research on the control method of aspheric surface machining trajectory has been carried out based on the tangent method.Analyzing the characteristics of the contour curve of the optical aspheric surface gives the explanation of the relationship between the trajectory control of the aspheric surface and the accuracy of the surface,as well as the principle of the trajectory forming of the aspheric surface by tangent method.The limitations of the trajectory control of the tangent method by using the existing interpolation methods are also discussed.In order to meet the requirements of the tangent method for high-efficiency data processing and high-precision trajectory control ability,the neural network technique is proposed to optimize the control parameters of aspheric surface machining trajectory.Aiming at the problems of slow convergence speed and easily trapping into local minimum in practical application of artificial neural network,this paper will analyse and compare several improved methods of BP neural network,and propose to optimize the initial values of weights and thresholds by initial value assignment algorithm.By analyzing the control principle of tangent trajectory shaping,the influence rule of tangent trajectory control parameters on the processing trajectory curve is studied,and the mapping relationship between input and output of BP neural network is established.The three-layer network structure is set up to approximate the aspheric surface machining track curve,the neural network interpolation control model was established,and the interpolation data of the machining track curve was optimized,the simulation results has shown that the interpolation control model can complete the non-circular interpolation effectivelyIn the end,based on tangent trajectory shaping principle,the neural network interpolation control model is effectively combined with PC-UMAC motion control system for aspheric surface processing trajectory control and parameter optimization,and the neural network trajectory parameter optimization interface is developed and applied to aspheric surface processing test.The test results have shown that the neural network interpolation control method can save a lot of computing time of interpolation data and the aspheric surface components processed by milling and grinding have met the requirements of direct polishing.