Application Of Data Mining In Excavator Diesel Engine State Detection

In the fault diagnosis of excavator diesel engine,the data mining technology is applied to the state detection of excavator diesel engine in order to solve the problems that the fault is not found in time and accurately and the data of diesel engine is not fully utilized.Using the technology and thought of data mining,this paper studies the data mining algorithm for this topic,mining the potential information hidden in the data,realizing the state detection,so as to reduce the maintenance cost and reduce the economic loss.Firstly,the common faults of diesel engine are analyzed,and eight key state parameters are selected.The state data obtained from the data acquisition system of excavator diesel engine itself are selected to construct the state data set for subsequent analysis.In view of the status data for data without a label,it represents the state information of the unknown,cannot be directly used for detecting state problems,put forward a kind of based on immune clone wolves-K-means fault diagnosis methods,data validation results show that the introduction of gray wolves of the immune clone algorithm can obtain better initial clustering center to improve the clustering effect of K-means,By comparing with the clustering results of faulty diesel engines,it is found that category 5 is the fault data,and then the SVM is used to detect the health state,and the accuracy is higher than that of only using SVM.For state data loss cannot be directly used for detecting state problems,put forward a kind of based on adaptive weighted clustering missing data interpolation method,to improve the distance calculation formula and use the immune clone wolves-K-means of missing data clustering,similar other complete data calculation and missing data,the distance to the missing data,the weight is calculated adaptively according to the distance,and the weight and its corresponding integrity are interpolated to the missing data.The data verification results show that the error of this method is smaller.Because the vibration data packet contains many characteristics,but the data acquisition system of excavator diesel engine itself does not collect the vibration data,so the vibration data is not used in fault diagnosis.In order to solve this problem,the vibration data acquisition platform is first built by using NI CDAQ9172,NI 9233 and vibration sensors.Realized on the diesel engine cylinder head vibration data collection,and then put forward a kind of based on wavelet packet decomposition and random forest fault diagnosis methods,the vibration signal is three layer wavelet packet decomposition and calculating each component energy eigenvector,finally using random forests for classification of data validation results show that after wavelet packet decomposition,extract the features of the data,improved accuracy.At last,the status detection system of excavator diesel engine based on Web is developed,which realizes the status detection of excavator diesel engine by combining the data mining algorithm model and data with the technology of data mining and thought.