Research On Drum Brake Performance Prediction Based On Deep Learning

According to traffic accident-related statistics,it can be known that the accident caused by brake performance failure accounts for about 45%of the vehicle factors,and it is obvious that the research on brake performance changes and related test methods is of great significance for traffic safety and personal safety.In this paper,a drum brake performance prediction method based on deep learning is proposed by analyzing the inspection methods,standards,projects and technologies of the braking performance of traffic accident vehicles under the current standards.Firstly,by parametrically modeling the drum brake,a three-dimensional solid model between the brake construction parameters,motion parameters and brake clearance transformation law is constructed.The kinematic simulation using the parametric model is used,and on this basis,the calculation method and judgment basis of the braking performance under certain initial conditions are studied.Secondly,the parameterized model motion simulation and braking performance judgment are formed according to the parameters,which can provide training sets and verification sets for subsequent deep learning.On this basis,the deep learning convolutional neural network is used to construct a prediction model of brake performance,and the input variables of the model are the initial braking speed,actuator delay,braking force growth time,maximum deceleration,average deceleration,and pushrod clearance,and the output result is the braking distance,which meets the current national standard detection and identification standards.Finally,the real vehicle road test data of the same type of brake is collected and sorted,the brake performance evaluation index data is obtained as the test set,and the convolutional neural network model constructed above is used to test and verify,and the effectiveness of the prediction model is verified by comparing the predicted value of the braking distance with the true value,and the results show that the model has a good effect and has a certain scientific,effective and reference.Experimental results show that the root mean square error of the prediction model is within 0.3.This method provides the inspection basis for the relevant inspectors,improves the scientificity of their work and the authority of the test results;At the same time,it provides a technical path reference for the prediction of braking performance of other types of brakes,and can also provide theoretical basis and practical experience for the research on the manufacture,assembly and performance deterioration of brakes.