Diagnosis And Simulation Of Leakage Fault In Hydraulic System Of Crane Outrigger

Truck crane is a type of lifting machinery that is easy to move in working place.The outrigger hydraulic system is the most critical part to ensure the stability of the crane operation,and it is always in a high-load and variable working state.If there is insufficient pressure or soft legs due to leakage during operation,it is likely to cause a major safety accident.As one of the most prone faults,hydraulic internal leakage is often difficult to detect and diagnose.Therefore,it is important to diagnose and simulate the internal leakage of outrigger hydraulic system.In order to realize the fault diagnosis of the internal leakage in the hydraulic system of the crane leg,the hydraulic internal leakage experiment is carried out,and the hydraulic cylinder inlet pressure data was studied through the data-driven method.The method of fault diagnosis for the hydraulic leakage location and the hydraulic leakage amount were proposed.Then the internal leakage simulation model was established to study the impact of the leakage of various parts on the performance of the outrigger system,and the accuracy of the leakage assessment method was verified.The main contents are as follows:(1)With the purpose of diagnosing the location of the leak,a diagnosis method based on the two-dimensional time proximity map and the improved Le Net-5 network is proposed.The time segment method is automatically extracted through the two-dimensional time proximity map,and then the improved Le Net-5 network is used to complete the diagnosis.This method can automatically extract time domain segments and extract features.Compared with Le Net-5 network,the simplified one has a simpler structure.It can effectively prevent overfitting and avoid errors caused by small data volume.Its prediction accuracy is better than SVM,BP neural network and ensemble learning methods.(2)Aiming at the problem of internal leakage assessment,an assessment method based on temporal feature segmentation and dimensionality reduction spectral clustering is proposed.Through time feature segmentation,the problem of phase changes in hydraulic data and different features of each phase can be effectively solved.Extract the time domain and wavelet domain features of the data and remove non-sensitive information through PCA.Then,the method of dimensionality reduction spectrum clustering is used for fault diagnosis.This method requires less data and has a good diagnostic effect on different parts.Compared with other unsupervised learning methods,it shows that it has both stability and high accuracy.(3)The internal leakage simulation study of the outrigger hydraulic system with the parameters of the actual working conditions is carried out.The internal leakage failure of the vertical hydraulic cylinder and the reversing valve under the working condition is simulated,and the influence factors of the internal leakage is analyzed.The influence of different components and different degrees of internal leakage faults on the expansion and pressure holding performance of outriggers is study.The internal leakage assessment method based on dimensionality reduction spectrum clustering is verified by simulation data,which shows that the method has good adaptability.Based on the above-mentioned outrigger hydraulic system fault diagnosis and health assessment method,the outrigger health assessment system is designed.Finally,the design of each functional module,database and algorithm library is completed.The three major functions of data management,health assessment and user management are realized.The purpose of evaluating the hydraulic system of crane legs is achieved.