JCSE, vol. 20, no. 3, pp.230-242, September, 2026

DOI: http://dx.doi.org/10.5626/JCSE.2026.20.3.230

Improved Immune Clonal Algorithm-Driven Tool Path Optimization and Machining Accuracy Control

Yangyang Li and Heng Sha
Engineering Training Center, Shandong Huayu University of Technology, Dezhou, China

Abstract: Tool path optimization aims to plan the tool motion path to balance machining efficiency and geometric accuracy, while machining accuracy control ensures that the actual contour is consistent with the design model, and both jointly determine the surface quality and dimensional accuracy of the workpiece. Conventional tool path optimization and machining accuracy control methods have limitations in dynamic error compensation, easily causing process instability in free-form surface and whole array machining. To address this issue, this paper proposes a tool path optimization and machining accuracy control model based on the immune clonal algorithm. This study adopts an improved immune clonal algorithm to generate the optimal tool path and designs fast rejection and straddle tests to remove redundant overlapping path segments. This study establishes a dynamic error prediction model for real-time prediction of tool dynamic errors, on the basis of which a closed-loop compensation algorithm is constructed to achieve machining accuracy control. Experimental results demonstrate that the proposed prediction model possesses high precision, with a determination coefficient of 0.97 and a maximum prediction error of merely 0.005 mm. In free-form surface and array hole machining, the proposed model achieves satisfactory contour accuracy, surface quality and machining efficiency. After model compensation, the position error is reduced by 31.58% and the dimensional error drops to 0.014 mm. This model exhibits superior performance in global trajectory optimization, dynamic error prediction and real-time compensation, offering an intelligent approach for high-precision and high-efficiency machining of complex components.

Keyword: No keyword

Full Paper:   1 Downloads, 3 View