【摘要】 针对正流量挖掘机在动臂下降工况下，当多路阀动臂联阀芯前0～30%位移量过程中存在动臂抖动的问题，提出了一种基于神经网络模型的多路阀阀芯结构优化设计方法。在分析正流量挖掘机液压系统及多路阀阀芯结构的基础上，运用AMESim建立多路阀动臂联系统模型；根据AMESim仿真数据确定与动臂抖动相关的阀芯结构参数，通过正交实验法得到各参数组合条件下的动臂下降速度响应，实现其结构参数组合的神经网络表达；采用进化神经网络优化训练过程的初始权重和阈值，并应用优化后的神经网络得到阀芯节流槽最优结构参数。实验结果表明：在动臂下降工况阀芯前0～30%位移量过程中，动臂大腔压力脉动下降37%,动臂下降抖动问题得到有效解决，该方法对改进正流量挖掘机操作性能具有实际指导意义。更多还原

【Abstract】 A method for the optimized design of the multi-way valve spool structure based on the neural network model is proposed. This optimized design method is intended to address the issue of excavator boom jitter, with the valve core moving 0-30%. The AMESim model of the multi-way valve boom hydraulic system was established, with the hydraulic system of positive flow excavators and the structural features of multi-way valve spools analyzed. The structural parameters of the valve core related to the boom shake are determined and the neural network expression of the valve core structure parameter combination was realized by using AMESim simulation data. Finally, the initial weights and thresholds of the evolutionary neural network optimization training process are adopted, and the optimal structural parameters of the valve core throttle groove are obtained by applying the optimized neural network. It is proved in the experiment that during the 0-30% displacement process of the valve core, the jitter problem during the boom lowering process has been effectively solved and the pressure pulsation of the large chamber of the boom decreased by 37%. This method has shown a significant effect on improving the operating performance of the positive flow excavator.更多还原