目的　探索机械零件可靠性优化设计的一种新方法 .方法　在造价、体积、强度和重量等约束条件下 ,机械零件的可靠性优化设计问题是一个具有多局部极值、非线性及具有整数和实数变量的混合优化问题 .本文以发动机气门弹簧为例 ,应用遗传算法建立了机械零件可靠性优化设计的一种新方法 .结果　应用自适应遗传算法建立的新方法比常规优化方法更适合于并行处理优化计算 .结论　将自适应遗传算法用于机械零件的可靠性设计效果是明显的 Aim To explore a new method of reliability optimization design of mechanical components.Methods The reliability optimization design problem of mechanical components under the constraints of cost, volume, strength and weight is a multi-local optimization with multi-local extremum, nonlinear and integer and Aiming at the problem of hybrid optimization of real variables.This paper takes engine valve spring as an example and applies genetic algorithm to establish a new method to optimize the reliability of mechanical parts.Results The new method established by adaptive genetic algorithm is more suitable than parallel optimization Processing optimization calculation.Conclusion The application of adaptive genetic algorithm to the reliability design of mechanical parts is obvious