Prediction of Rock Burst Hazard Level Based on Improved Whale Optimization Algorithm-Least Square Support Vector Machine (IWOA-LSSVM)

• Published in: 2022 2nd International Conference on Electrical Engineering and Mechatronics Technology (ICEEMT) pp. 446 - 455
• Main Authors: Xu, Yaosong, Ren, Rixin, Wang, Shuyue, Wang, Zhizhong
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.07.2022
• Subjects: Data models; Hazards; Improved Whale Optimization Algorithm (IWOA); Kernel; Kernel Principal Components Analysis (KPCA); Least Square Support Vector Machine (LSSVM); Prediction algorithms; Predictive models; Rock burst prediction; Rocks; Safety engineering; Support vector machines
• DOI: 10.1109/ICEEMT56362.2022.9862701

To improve the accuracy of rock burst level prediction and overcome the difficulty in determining the parameters of traditional prediction models, this paper proposed a prediction model of rock burst hazard level based on the Least Square Support Vector Machine (LSSVM) optimized by an Improved Whale Optimization Algorithm (IWOA). The Kernel Principal Components Analysis (KPCA) algorithm was used to reduce the dimensions of risk factors that affect the occurrence of rock bursts. The Whale Optimization Algorithm (WOA) was improved through the introduction of Tent chaotic mapping, the improvement into a nonlinear convergence factor, the storage of the optimal position in memory, and the introduction of the joint strategy of Cauchy disturbance to improve the optimization performance. Related parameters of LSSVM were optimized and a prediction model of rock burst hazard level based on IWOA-LSSVM was set up. With the measured data as the sample, the experiment made a comparison with other models and conducted an analysis. The results showed that the prediction accuracy of the model established was 93.3%, 7.61% higher than that of the WOA-LSSVM model, which verified the effectiveness of the model proposed for the prediction of rock bursts.