基于故障机理的液压泵多信息烈度特征状态评估方法

以液压泵滑靴磨损故障为例，提出一种基于故障机理的多信息烈度特征状态评估新方法。该方法从滑靴磨损机理出发，利用滑靴副压紧系数值对滑靴不同磨损程度对应的液压泵工作状态进行区域划分；通过振动烈度的频域计算方法提取泵壳体振动、出口流量及压力三种信号的烈度特征因子，分析三种烈度特征因子对滑靴磨损程度的敏感性，并建立特征因子样本集；利用最小二乘法进行数据拟合，得到三种烈度特征因子与液压泵工作状态的对应量化关系，结合BP神经网络和D-S证据理论建立基于多信息决策融合算法的状态评估模型。通过测试样本验证了模型的有效性，结果表明该模型具有较高的评估精度。

Multi-information Intensity Characteristic State Assessment Method of Hydraulic Pums Based on Fault Mechanism

Taking the wear fault of hydraulic pump slipper as an example, a new method for assessing the status of multi-information intensity features was proposed based on fault mechanism. Starting from the wear mechanism of the slippers, this method divided the working states of the hydraulic pumps corresponding to the different wear degrees bused on of the slippers based on the compacting coefficients of the slippers. The intensity characteristic factors of pump vibrations, outlet flows and pressure signals were extracted by frequency domain calculation method of vibration intensity. The sensitivity of three intensity characteristic factors to the wear degrees of the slippers was analyzed, and the characteristic factor sample sets were established. The least squares method was used to fit the data to obtain the corresponding quantified relationship among the three intensity characteristic factors and the working states of the hydraulic pump. The BP neural network and D-S evidence theory were used to establish the state assessment model based on the multi-information decision fusion algorithm, and the validity of the model was verified by test samples. It indicates that the model has higher assessment accuracy.