Research On Analysis Method Of Mine Rockburst Characteristics Based On Information Fusion Technology

Rockburst is a dynamic,spontaneous and uncontrolled geological hazard that is common during the excavation of deep rock mass engineering.When it happens,it will cause damage to underground structures and equipment,threaten the safety of staff and equipment.Due to the non-linearity of the rockburst and the highly complex mechanism,the analysis of a single signal is not enough.To accurately predict the rockburst,need to analyze many of the rockburst characteristics.Firstly,to analyze the characteristics of the representative static indicators selected.Because of many indicators,the analysis method needs to have a high effct on reduction.So,the non-dimensional data of the PCA method is improved.By comparing the improved and traditional methods,found it improves data reduction ability.So,use it to analyze the static indicators,finally extracts the predictive indicators.Then,multiple sensors(acoustic emission,infrared radiation and electromagnetic radiation)are used to dynamically monitor a variety of characteristic quantities(vibration,temperature,pressure,flow,etc.).Acoustic emission adopts an improved PCA,remove useless information,select acoustic emission rockburst.Analyze the electromagnetic radiation use improved algorithm,the FIR-improved wavelet joint denoising algorithm is used to denoise it signal.Comparing the simulation experiment and denoising effect evaluation index of others,found the improved algorithm has a good effect on the signal.Finally,a rockburst prediction model based on support vector machine is established,and the parameters of model are optimized by improved particle swarm algorithm.The improved algorithm balances the search capability and convergence speed,improves the accuracy of the rockburst prediction model,and the simulation experiment verifies the correctness of the model.Figure 28;Table 17;Reference 56...