发动机磨损诊断的神经网络技术研究

Research of Neutral Network Technique in Engines Wear Diagnosis

目的 利用神经网络技术诊断发动机的磨损故障 ,研究神经网络技术应用于机械故障诊断领域的发展方向 .方法 基于油液分析技术 ,对发动机常见磨损故障进行诊断 ,具体包括发动机磨损颗粒的识别、磨损元素浓度值的预测和磨损故障智能诊断专家系统 .结果 对基于油液分析的发动机故障诊断提供了新的方法 .结论 使用神经网络技术可以大大提高发动机磨损故障诊断的效率 ,说明该技术方法在发动机磨损故障诊断领域是有效的 。

Aim\ To diagnose engines′ wear by using neutral network technique, and the prospect of this technique in mechanical wear condition monitoring is discussed. Methods\ Engines′ wear are diagnosed based on oil analysis. The methods include predicting the concentration of oil of wear metallic elements, and the recognition of the ferrography. The effectiveness of the technique in recognizing the kind of wear is also discussed, studying the application of neutral network technique on the expert system of wear diagnosis. Results\ New methods to diagnose engines′ wear based on oil analysis are given. Conclusion\ Neutral network technique is an effective method in the diagnosis of engines′ wear, and this method is also useful for diagnosing the wear of other mechanical equipment.