Research on Accident Prediction of Cable Tunnel Based on FT-GS-SVR Algorithm

The accident category in the cable tunnel is mainly judged according to whether a single environmental parameter exceeds a certain threshold, which has the defects of high accident false alarm rate and insufficient intelligence. To solve this problem, this paper proposes a cable tunnel accident prediction model based on fuzzy theory grid search and algorithm optimization regression support vector machine (FT-GS-SVR). Firstly, In order to solve the problem that the historical fault data set only retains the collected information and fault categories, the data set is standardized. Secondly, it is fuzzed and clarified by the fuzzy membership method to obtain a training data set with grading indicators. Finally, a support vector regression model is established. In this model, the cross-validation grid search algorithm is used to globally optimize the penalty coefficient C and the kernel parameter g, so as to obtain the optimal accident prediction model. Based on the fuzzy GS-SVR prediction model combined with the historical fault data set, the results show that: the accuracy rate of the 12 test sets of the model in this paper is 100%, and the error of the classification warning size is within ± 0.3. Besides, the proposed model can accurately predict tunnel accidents according to real-time input data and data changes, as well as graded early warning in advance.