基于优化灰色神经网络模型的内燃机气缸套磨损量预测的研究

摘要

气缸套磨损量是衡量汽车发动机寿命的重要因素.针对传统灰色预测模型的缺陷,采用均值修正策略,对样本数据进行预处理,实现灰色预测模型的优化；将灰色预测模型与BP网络相结合,建立了优化灰色神经网络预测模型；实例分析结果表明,优化模型能更精确地预测发动机气缸套的磨损量,为发动机的寿命预测与维修提供了更有效的方法.%Cylinder liner wear is an important factor to measure the life of engine. In view of the defect of traditional grey prediction model (GPM), a mean correction strategy is adopted to conduct the preprocessing of sample data and realize the optimization of GPM. By combining GPM with BP network, the optimized gray neural network prediction model is established. The results of real case analysis show that the optimized model can more accurately predict the wear of cylinder liners, providing a more effective way of engine life prediction and maintenance.