Investigation Of Optimization Algorithm For Polishing Path Based On Particle Swarm Optimization

Optical technology plays an important role in all aspects of modern society,and the application of optical technology is inseparable from the optical elements as carriers and tools.In the optical system,aspheric and free-form surface elements have many advantages over the traditional spherical elements,but also have greater processing difficulty.The deterministic processing technology(CCOS)based on computer-controlled small grinding head can be called the second generation optical processing technology.The third generation polishing technologies,such as magnetorheological polishing,jet polishing,air bag polishing and ion beam polishing,have been developed based on computer-controlled mode.The difference lies in the different polishing heads used.At present,CCOS technology is still the mainstream optical processing technology,which can be connected with more advanced polishing technology to improve the processing efficiency.Polishing process takes up most of the time in optical surface machining,and polishing trajectory planning directly affects the efficiency of machining and the quality of workpiece after machining.Therefore,the research on polishing trajectory planning is of great significance to the development of optical surface machining.Spiral trajectory is a very good polishing trajectory.In order to explore the optimal distribution of spiral trajectory,particle swarm optimization algorithm is introduced to find the optimal solution in the solution space.In this paper,the area of contact between the spherical tool head and workpiece is modeled,and a material removal profile model is established for polishing the spherical tool head along the curve.The model is suitable for spiral machining trajectory.Taking K9 glass as an example,the accuracy of the model is verified by experiments.On this basis,considering the interference between adjacent tracks,a single object particle swarm optimization algorithm with machining quality as the objective function is established.The change trend of Ji calculation results is studied when the parameters of the algorithm and the number of spiral segments increase.The algorithm can ensure the coverage of spiral and filter the trajectory distribution which may lead to edge collapse.The MOPSO/D,which combines PSO and Tchebycheff decomposition,realizes the multi-objective optimization algorithm of polishing trajectory with polishing quality and polishing efficiency as the target value.The algorithm decomposes the multi-objective optimization problem into several single-objective sub problems.Each particle and its neighborhood constitute particle swarm optimization algorithm for single sub problem.It can solve complex problems without gradient of objective function Secondly,it can solve all single-objective subproblems through a set of operations,make full use of the mechanism of particle swarm cooperation and share,and can keep the diversity of solutions,and get better Pareto solution sets.In order to solve the problem of high cost in calculating the polishing trajectory profile,the neural network algorithm is introduced to improve the efficiency of the algorithm.The new algorithm can support a larger population of particles and more iteration steps in a shorter time.In order to solve the problem of minimum and the cost of target function calculation which are easy to meet when PSO is used to solve the multi peak problem,mopso/d is improved by roulette operator and extreme disturbance operator.The local minimum problem is optimized.The improved algorithm is compared with the original algorithm,and the improved MOPSO/D has the advantages.