基于ES-RF模型的材料裂纹故障分析研究

Research on Material Crack Fault Analysis Based on the ES-RF Model

摘要:

材料裂纹故障是飞机维修检查中常见故障之一,该故障的发生会影响飞机的正常运行,降低航空公司运行的效率。针对材料裂纹故障中发生难以通过目击直接检查,目前采用的手工检查方式耗时长,准确率低等问题,对A、D两种类型的裂纹故障提出ES-RF(专家系统-决策树)模型,对材料裂纹故障进行分析研究。分析过程中首先使用专家系统验证材料裂纹故障发生与多个检测参数的相关性,随后使用决策树算法对材料裂纹故障进行分析。分析结果表明使用ES-RF模型对材料裂纹故障评估的训练准确率98.98%,测试准确率97.93%.

Abstract:

Material crack failure is a common fault encountered during aircraft maintenance inspections. The occurrence of this fault can affect the normal operation of aircraft and reduce the operational efficiency of airlines. Addressing the difficulty in directly detecting material crack failures through visual inspection, and the issues associated with current manual inspection methods such as long duration and low accuracy, an ES-RF (Expert System-Random Forest) model is proposed for analyzing and studying material crack failures. The analysis process first utilizes an expert system to verify the correlation between the occurrence of material crack failures and multiple inspection parameters, and then employs a random forest algorithm to analyze the material crack failures. The analysis results indicate that the ES-RF model achieves a training accuracy of 98.98% and a testing accuracy of 97.93% in evaluating material crack failures.