Research on risk prediction of bridge support maintenance construction based on machine learning

The risk prediction in the process of bridge support maintenance and construction is very important to the stability and safety of the whole bridge structure. Taking a bridge support maintenance construction project as a case, this study analyzed and sorted out the risk factors from three aspects of construction environment, support maintenance construction and construction management through expert investigation and literature reading based on the machine learning algorithm prediction model, established the risk prediction index system, and adopted the grid search algorithm to optimize the hyperparameters of the prediction model. The model evaluation indexes R2 and MAE were used to evaluate the prediction results. The results show that the prediction model of the neural network algorithm can predict the construction risk more accurately than that of the random forest algorithm. However, the performance of the random forest algorithm on the test set is poor due to the over-fitting of data on the training set, and the performance of the model needs to be improved by combining related optimization algorithms.