Study On Fault Diagnosis Of Full Hydraulic Breaking System Of Trackless Vehicles Based On SVM

The experience from home and abroad shows that mine trackless transportation equipment plays an important role in the operation of modern mine. Compared with the traditional mode of transport, mine trackless transport equipment with a large load, fast and other unique advantages, therefore, can greatly improve the mine transport capacity, production capacity, economic and social benefits. Among them,the trackless rubber tire vehicle has been widely used because of its good flexibility.In order to adapt to the coal mine under the special conditions, most of the trackless rubber tire vehicles adopt full hydraulic double circuit braking system. As the coal mine environment is relatively poor, trackless rubber tire vehicles have to bear a huge long-term workload, while to carry out frequent start and braking operations under high load, So the accumulator, charging valve, brakes and hydraulic lines will inevitably occur some failure, resulting in vehicle transport process loss of control,and more seriously it may cause mine personnel casualties, resulting in unnecessary loss of manpower. Therefore, the research on the fault diagnosis of the hydraulic braking system of the trackless rubber tire vehicle is of great value to the safe and efficient production of the coal mine.Firstly, the hydraulic brake system model was built in the AMEsim environment according to the structure and important working parameters of the hydraulic braking system of the trackless rubber tire vehicle. The pressure output curve of the front and rear axle accumulator and the front and rear axle brakes was obtained. The dynamic response performance of the system was analyzed by combining the working principle and actual working performance of the hydraulic braking system, and the rationality of the AMEsim simulation model was verified. At the same time, the output pressure data of the front and rear axle accumulator, front and rear axle brakes and the input pressure data of the brake pedal were obtained through the simulation model, which provided a reliable original data sample for support vector machine fault diagnosis.4 SVM regression models were established which based on SVM regression fault diagnosis theory and fault mode of full hydraulic braking system. Based on the model, the accumulator pressure of the front and rear axle, the brake pressure of the front and rear axle are predicted and checked, the modeling error and calibration error are in the 10-1 -10-1 orders of magnitude, so the good generalization performance of the SVM regression model is verified. Finally, the SVM regression model is used to predict the fault data, and the fault is judged according to the change of the residual value, which proves the feasibility of the SVM method in the fault diagnosis of the hydraulic braking system.In order to further improve the performance of fault diagnosis of SVM regression prediction model,and effectively improve the accuracy of diagnosis,the kernel parameter g and penalty parameter c of the SVM fault prediction model were separately optimized by cross validation, genetic algorithm and particle swarm algorithm. Under the optimal parameters g and parameter c, the SVM regression model is trained and predicted respectively, and obtained the SVM fault prediction model with better diagnostic effect. The prediction accuracy is 10-5 orders of magnitude, so the result is satisfactory.Finally, a testing and fault diagnosis system of full hydraulic braking system is designed with Labwindows/CVI development environment, which for real-time monitoring of the hydraulic braking system state through the friendly man-machine interface. ActiveX technology is used to call the MATLAB support vector machine regression model program, which can effectively complete the fault diagnosis and detection of the full hydraulic braking system.