Model of gas-bearing coal permeability prediction based on APSO-WLS-SVM

In order to predict gas-bearing coal permeability accurately and prevent gas accidents effectively, a weighted least square support vector machine(WLS-SVM) algorithm based on Adaptive Particle Swarm Optimization(APSO) is proposed. Based on the related theory and literature research on gas-bearing coal permeability, the effective stress, gas pressure, temperature and compressive strength were selected as the main characteristic indexes. The APSO algorithm was used to optimize the combination parameters(C, σ) of the WLS-SVM model and APSO-WLS-SVM gas-bearing coal permeability prediction model was established. According to the field data, 40 sets of data were used as training samples and the remaining 10 groups were predicted samples. Then the model was trained and tested, and its prediction results were compared with the results of PSO-WLS-SVM and WLS-SVM respectively. The results show that the prediction effect of APSO-WLS-SVM model is better than the other two models, improves the prediction performance and generalization ability of coal permeability.