Data-driven Fault Diagnosis And Health Assessment For Key Components Of Truck Crane

The truck crane is one of the large-scale special equipments integrating electro-hydraulic.Due to the high strength and high complexity of its working conditions,the performance of its key components will gradually decline over time.The specific reasons include operational leakage,friction,wear,impact,vibration,high temperature,corrosion and other factors.The performance degradation of key components will affect the health of the machine.If the spool is tightened,the return spring will be faulty or the spool displacement is not in place,causing the hydraulic oil flow in the control valve to be too large or too small.In severe cases,the system failures may occur,affecting normal production operations.In order to accurately assess the health status of the crane key components,reduce the failure rate and risk under actual operating conditions.This paper is aimed at the key components of the truck crane,such as outrigger control valve,rotary component and transmission system spindle.Based on the mechanism of health deterioration,the component fault diagnosis method and performance evaluation algorithm model are studied,Establishing non-linear mapping relationship between control flow data,sensor signals and health status during operation,so as to achieve the purpose of the health performance evaluation of the crane system.The main contents are as follows.(1)Health performance assessment for outrigger system control valves.Collecting the outlet pressure signal for wavelet domain feature and time domain feature extraction.For the characteristics of the control valve signal,the principal component analysis is used for dimensionality reduction,and the Mahalanobis distance is used to evaluate the extent to which the test data deviates from the baseline data.Finally,it is mapped to the health index space for measurement.By comparing with other methods,the robustness and generalization of the method are more prominent.(2)The health performance evaluation for the swing system,the 12-dimensional signals such as the handle tilt angle,the swing speed,the arm length,the rated crane weight,and extracting reduced data dimensions by random forest feature importance.The Gaussian kernel density estimation is constructed to characterize the characteristics of the data,and the Probability density obtained by Kullback-Leibler relative entropy to get health performance trends.Compared with other model methods,The method avoids data noise interference and can accurately achieve the purpose of health performance assessment.(3)Fault diagnosis for the spindle of the crane drive system.The acceleration signal of the main shaft is collected for the vibration signal.Firstly,the acceleration signal is denoised by wavelet packet,the power spectrum is drawn,and the initial fault diagnosis is performed according to the fault characteristic frequency.Then the wavelet packet energy is used as the feature vector,and the error correction output coding support vector machine is used for fault diagnosis.The experimental results show that the method can accurately diagnose the fault state of the transmission components.The advantage is to correct the deviation caused by the single classifier,and the diagnostic accuracy in the multi-classification problem is improved.(4)Design and develop a health assessment system.The system acquires control flow data and sensor signals through the acquisition card of the embedded controller,and draws a trend of health index change according to the establishment of a diagnostic evaluation model.At the same time,the system has the ability to expand other key components of the crane,providing a technical,methodological and theoretical basis for crane fault diagnosis and health assessment.Based on the above-mentioned key component fault diagnosis and health assessment methods,the health performance evaluation of multicomponent signals and mutual evaluation of multiple devices will be carried out in the future to further study the health maintenance of key components of cranes.