Numerical Interpolation Based On The Genetic Algorithm And Neural Network

Today, the ultra-precision machining technology of aspheric surface optic parts has become the key technology that various nations strive for develop in the field of ultra-precision machining, but numerical control interpolation algorithm is one of the core technology for ultra-precision numerical control.In the thesis, the general situation of the domestic and abroad developments in the field of the numerical control interpolation algorithm for aspheric surface machining is reviewed firstly, for the poorness of recent numerical control interpolation algorithm, an numerical control interpolation algorithm that based on the Genetic Algorithm(GA) and Neural Network(NN) is introduced. The Algorithm mainly adopted the parallel performance of neural network and the character that can simulate arbitrary non-circular function, and cut the time of interpolation greatly. So interpolate for any curve or for space disperse become possible. Aim at the neural network get local optimum easily, the speed of convergence is slow, the quantity of training excessive and so on. This paper adopted the GA optimize the NN firstly, and then the NN Algorithm. Because GA possesses the traits of can global random search, the robustness is strong, been use briefly and broadly, it didn't use path search, and use probability search, didn't care inherence rule of problem itself, can search the global optimum points effectively and rapidly in great vector space of complicated, many peak values, cannot differentiable. So it can offset the shortages of NN study Algorithm, can reduce the possibility that the minimum value get into local greatly, the speed of convergence can improve, interpolation time shorten greatly, the quantity of training reduce.Through compare the simulink results of GA-BP interpolation modeling with tradition methods, and compare the interpolation results with tradition method, the method that can improve the precision of numerical control interpolation and can diminish the quantity of calculate and interpolation cycle is been proofed. At end, do the experiment on the improved equipment, the feasibility of the method is validated.