A New Method of Priority Analysis and Optimization for Accuracy Allocation of Machining Center Considering Error Polarity
摘要
Achieving ultra-high precision in CNC machining requires robust geometric error allocation. Traditional methods, limited by localized optimization and unidirectional tolerance assumptions, frequently fail at workspace boundaries. This paper presents a novel systematic method that introduces a bidirectional error model based on multi-body system theory. Leveraging global sensitivity analysis, the method employs genetic algorithm optimization, constrained by worst-case volumetric errors evaluated across 27,000 workspace points, to guarantee full-workspace accuracy. Simulations verify that all errors are contained within ± 0.01 mm. Compared to the traditional single-point approach, the proposed method reduces the non-conformance rate significantly (from > 60.89% to 0%) while strictly maintaining the accuracy target, significantly improving assembly precision and manufacturability.