基于遗传算法的机车二系支承载荷调整优化方法

Optimization Model of Locomotive Secondary Spring Load Adjustment Based on Genetic Algorithm

为了改善机车轮轴重量分配的均匀性,从分析机车二系支承结构出发,建立二系载荷分布弹簧高度调整模型,提出应用遗传算法实现机车二系支承点加垫高度值的快速优化计算。针对机车二系支承载荷分布多维参数优化问题的特点,给出基于实数编码的改进遗传算法模型。对国产SS9,SS3b型机车实际应用,可使二系支承载荷分布最大误差进一步减小8%～15%,平均计算过程时间减少约20%。结果表明这种方法克服了基于误差梯度信息迭代搜索求解支承点加垫最佳值算法的不足,在获得一致的全局优化(近似)解,以及在减小运算时间、提高调簧算法的可靠性和实时性等方面具有明显的效果。

In order to improve the equality of the locomotive axle and wheel load on rail, this paper presents a mathematical model for secondary spring load adjustment on the basis of locomotive secondary suspension structure analysis, and introduces a method using genetic algorithm (GA) to determine the optimal padding height of secondary spring. Targeting to the multi-parameter optimization of the load distribution, an improved GA based on real parameter coding is proposed. Applications to SS9 and SS3b locomotives yield 8%-15% further reduction of maximum deviation of secondary spring load and about 20% less average computing time. The result shows that the new approach has enhanced ability to find overall optimal solution to the problem by overcoming the shortcomings of iterative methods dependent on error gradient information, and leads to faster real-time computing, more accurate and reliable result in secondary load test and spring adjustment.