Pipe gallery fire detection method and system

The invention provides a pipe gallery fire detection method, which comprises the steps of acquiring a pipe gallery sample and a neural network used for pipe gallery fire detection, wherein the pipe gallery sample comprises a training set and a test set; inputting the training set into the neural network for training to obtain a training model, inputting the test set into the training model, and determining training parameters and a detection model; and detecting whether a fire occurs in the real-time pipe gallery sample through the detection model. A detection model is determined and related feature data are analyzed through establishment of a neural network and data training, so that whether a pipe gallery fire occurs or not is predicted more accurately and quickly, fire prevention work is done in advance, feasible suggestions can be provided for safety and operation maintenance of the comprehensive pipe gallery, the safety of the comprehensive pipe gallery is improved, and casualties and property loss are r