基于三点模型的柴油机热工参数预测

准确预测热工参数劣化趋势是对柴油机实施预测维修的前提。已有预测技术均视不同时段信息所具有的建模价值相等,忽略了近期信息比远期信息更能反映设备当前运行态势这一客观事实,导致预测模型难以准确描述柴油机热工参数的真实变化规律。针对该问题,提出一种基于三点模型的柴油机热工参数预测方法,并采用新陈代谢法同步更新建模数据,以提高建模数据的利用效率及预测精度。与灰色系统GM(1,1)模型预测精度的对比结果表明,三点模型能够有效处理不同时段信息在反映设备运行态势能力上的差异,得到更为准确的预测结果。

Study on Performance Prediction of Diesel EngineBased on Three-Point Model

It is the precondition for implementing the predictive maintenance of diesel engine that accurately predict the performance of the machine. Available methods take no account of the fact that latest data can reflect the operating condition of diesel more accurately than early one, since the value of the data collected in different time is regard as equal. The variation treads of diesel thermal parameter can be hardly predicted based on the available methods for the foregoing reason. In this paper, the paper proposes a novel thermal parameter prediction method for diesel engine based on the three-point model. In order to improving the sample data use efficiency and predicting accuracy, the metabolism algorithm is utilized to update the modeling data synchronously. The predicting effect of three-point model is contrast with grey model GM(1,1). The results show that three-point model can treat the difference between the data collected in different time availably, and the predictions are moreSapproximate toStheSreal treat.