A life prediction method of hydraulic pump based on grey support vector was presented for the shortcomings of low precision forecasting model that based on simple method. This method processed the origin data with grey accumulated generating operation to strengthen the regularity of data, confirming the insert data dimension and choosing the parameters of model with minimum final prediction error criteria,and using grey support vector machines to realize hydraulic pumplife prediction, and through reducing data with inverse accumulated generating operation to obtain the result of prediction. Spectroscopic analysis data of hydraulic oil was selected as life characteristic information of hydraulic pump. The presented model was used to predict life of hydraulic pump and the predicted result was compared with that of grey model and support vector machines. The result shows that this model has the best prediction accuracy,and the accuracy is 99. 37% ,so an effective approach for evaluating performance and forecasting life of hydraulic pump was offered.%针对单方法所建液压泵寿命预测模型精度较低的缺陷,提出基于灰色理论和支持向量机的组合预测模型的液压泵寿命预测方法.该方法通过灰色累加生成操作对原始序列进行数据处理,以增强数据的规律性；运用最小最终误差预测准则确定嵌入维数,选择模型的参数；采用支持向量机进行预测,利用灰色累减生成操作还原数据,得到预测结果.选取液压油的光谱分析数据作为液压泵的寿命特征信息,采用该模型对液压泵进行寿命预测,并与灰色模型、单一支持向量机模型进行预测性能对比.结果表明,灰色支持向量机预测性能最优,精度达到99.37％,为液压泵性能评估和寿命预测提供一种更为有效的方法.
展开▼
