Cylindricity error evaluation based on an improved artificial gorilla troop optimizer

Abstract The cylindricity error evaluation of shaft parts is a nonlinear optimization problem, and no specific calculation formula exists. To evaluate the cylindricity error efficiently and accurately, an improved artificial gorilla troop optimizer (IGTO) is proposed, which is applied to the cylindr...
Ausführliche Beschreibung

Abstract The cylindricity error evaluation of shaft parts is a nonlinear optimization problem, and no specific calculation formula exists. To evaluate the cylindricity error efficiently and accurately, an improved artificial gorilla troop optimizer (IGTO) is proposed, which is applied to the cylindricity error evaluation. First, in the generation of the initial solution of the artificial gorilla troop optimizer (GTO), using the uniformity and ergodicity of the tent map to rearrange the initial population increases the diversity and improves the randomness and irregularity. Second, the Levy flight is integrated into the development stage of the GTO. The vitality of the moving population position is increased by moving the step size of the population optimization. The position update formula of the slime mold algorithm is used along with the Levy flight, which adjusts different searched modes to dynamically adjust the balance between global and local searches of the algorithm and improves the optimization accuracy and stability of the algorithm. Finally, the current optimal population is perturbed using the Cauchy mutation strategy. According to the Cauchy mutation operation, the disturbance step size can be adjusted adaptively to avoid the population falling into the local optimum and to improve the convergence speed of the algorithm. The IGTO was compared with four advanced meta-heuristic algorithms on eight test functions. The results indicate that the IGTO has advantages regarding computational accuracy and iteration speed. The IGTO was used to simulate the cylindricity error evaluation of the three sets of data. It can also quickly find the central axis of the minimum zone and obtain a more accurate cylindricity error value than other algorithms. For Dataset 1, the accuracy improved by 1.6%, that for Dataset 2 performed as well as in previous algorithms, and that for Dataset 3 increased by 44%. Ausführliche Beschreibung