论文部分内容阅读
针对大孔径管件的圆度误差问题,研制了一套在线检测系统,并提出了一种分组粒子群优化算法的圆度误差评定,将整个搜索空间分成若干个小组,各个小组单独搜索,迭代到一定步数,将各个小组混合,重新合成一个大的种群,在整个空间追随各自的极值和种群的最优值,避免了陷入局部最优解。在自行研制的检测系统中进行了应用,证明分组粒子群优化方法收敛速度快,精度高,适于圆度误差的评定,解决了大孔径管件圆度误差在线评定问题。
In order to solve the roundness error problem of large-aperture pipe fittings, an on-line inspection system was developed and a roundness error assessment of the particle swarm optimization algorithm was proposed. The whole search space was divided into several groups. Each group was searched separately and iterated After a certain number of steps, all the groups are mixed to recombine a large population, and follow their respective extreme values and the optimal population values in the entire space to avoid falling into a local optimal solution. It is proved that the PSO method has the advantages of fast convergence and high precision, which is suitable for the evaluation of the roundness error and solved the problem of online evaluation of the circularity error of the large-aperture pipe.