Ma Xiao-ping

Rock burst chaotic prediction on multivariate time series and LSSVR

State variables reconstructed by multivariate time series were used as LSSVR model inputs to predict the future value of rock burst monitor variables. First, the chaotic prediction principle on multivariate reconstruction and LSSVR was given. Then given the effects of reconstruction parameters on reconstruction results and LSSVR parameters on prediction error, genetic algorithm was adopted to determine reconstruction and LSSVR parameters simultaneously to ensure chaotic prediction accuracy. Finally, in Matlab2009b environment, based on the effectiveness verify of our method by Lorenz chaos system, Microseism time series were simulated to predict rock burst. The results show that the rock burst prediction method on multivariate time series reconstruction and LSSVR can accurately predict monitoring variables in advance to forecast rock burst even in the case of relatively short history data.

Fault Diagnosis of Mine Hoist Braking System Based on Wavelet Packet and Support Vector Machine

This paper concerns mine hoist braking system fault diagnosis with the combination of wavelet packet and support vector machine. It is motivated by the scarce of fault samples in mine hoist such requiring very high security system. A novel approach is presented in order to diagnose blockage piston in cylinder, a typical fault of mine hoist braking system. This method mainly consists of three steps: (1) apply 3 levels wavelet package to construct and reconstruct signal of brake distance-time , extract fault feature vectors (2) set up training samples (3) establish a SVM fault classifier to complete fault diagnosis. Experimental results show that SVM method can effectively accomplish the blockage piston in cylinder fault diagnosis of braking system and has a high adaptability for fault diagnosis in the case of smaller number of samples.

Predictive functional control based on differential evolution algorithm and its dynamic performance analysis

An optimization method of predictive function control (PFC) parameters that based on modified differential evolution (DE) is provided. Differential evolution is a new evolutionary computation technology and exhibits good performance on optimization. Differential evolution algorithm as a relatively new evolutionary computation technique has a good optimization. Therefore, the modified differential evolution which is proposed to solve the optimization problems. The new algorithm uses initialization and the scale factor and crossover probability to improve PFC control performance in terms of model mismatch and parameters optimization. Simulation results show that the performance of the optimized DE PFC controller is superior to that of the conventional PFC controller.

Vibration trend prediction based on gray LSSVM combination model for mine main ventilator

A novel combination prediction model based on gray theory and Least Squares Support Vector Machines(LSSVM) is put forward. Firstly, GM(1,1) model is adopted to forecast the trend item in non-stationary time series; Secondly, LSSVM model is used to predict the residual sequences of the GM(1,1); Finally, the prediction values are computed by adding the trend item and residual prediction values. Furthermore, this model is used to predict the vibration trend of mine main ventilator. The results show that this combination model can get the best predicting precision. Therefore, this model can satisfy the engineering application requirement.