Vibration Reliability Analysis of Drum Brake Using the Artificial Neural Network and Important Sampling Method

Abstract

This research aims to evaluate the calculation accuracy and efficiency of the artificial neural network-based important sampling method (ANN-IS) on reliability of structures such as drum brakes. The finite element analysis (FEA) result is used to establish the ANN sample in ANN-based reliability analysis methods. Because the process of FEA is time-consuming, the ANN sample size has a very important influence on the calculation efficiency. Two types of ANNs used in this study are the radial basis function neural network (RBF) and back propagation neural network (BP). RBF-IS and BP-IS methods are used to conduct reliability analysis on training samples of three different sizes, and the results are compared with several reliability analysis methods based on ANNs. The results show that the probability of failure of the RBF-IS method is closer to that of the Monte-Carlo simulation method (MCS) than those of other methods (including BP-IS). In addition, the RBF-IS method has better calculation efficiency than the other methods considered in this study. This research demonstrates that the RBF-IS method is well suited to structure reliability problems.

A reliability‐based robust optimization design for the drum brake using adaptive Kriging surrogate model

“…The commonly used surrogate model methods are RSM, 24,25,39 support vector machine, 27 Kriging surrogate model, 26,[28][29][30][31] and ANN. 8,22,23,32 Based on these theories, researches propose new ideas and advanced surrogate model methods such as VSM method, 11 deep learning model, [33][34][35] and advanced Kriging method. 12,36 In this subsection, several advanced surrogate model methods are briefly reviewed.…”

“…In order to solve the problem, many surrogate model methods are widely used for reliability problems in recent years. The commonly used surrogate model methods are RSM, 24,25,39 support vector machine, 27 Kriging surrogate model, 26,28–31 and ANN 8,22,23,32 …”

“…ANN has characteristics of strong nonlinear mapping ability, is widely used for reliability problem of complex structure. For the feedforward neural network, the output model is expressed as 23 $$$$\begin{equation}{G_A} = f(\sum {wX + \theta } )\end{equation}$$$$where w is the weight matrix between the hidden layer and the output layer, θ the threshold matrix, and $$f( \cdot )$$ is the transfer function between the hidden layer and output layer.…”

“…For the reliability problem of the drum brake, Yang et al 22 present a new frequency reliability performance function model of the drum brake using ANN, and the reliability sensitivity analysis is conducted by the advanced first-order second-moment method. Yang et al 23 propose an effective reliability analysis method based on ANN, and the calculation accuracy and efficiency of this method are evaluated by two examples and the drum brake reliability problem.…”

Study on reliability of emergency braking performance of high-speed and heavy-load monorail crane

“…and carried out vibration reliability by BP neural network, radial basis function neural network with important sampling method (RBF-IS) [29]. To address the problem of squeal instability of automotive brakes, Lyu et al [21,22] proposed the random and fuzzy model and the fuzzy random model for the squeal instability reliability analysis of automotive brake discs, and proposed an interval variable-based response surface method (RSM) to approximate the implicit relationship between the approximation system parameters and noise modes.…”