Reservoir slope deformation early warning method based on deep learning algorithm

The invention relates to the technical field of reservoir safety, in particular to a reservoir slope deformation early warning method based on a deep learning algorithm. Comprising the steps of determining basic data of a slope to be pre-warned; determining a primary prediction model and a plurality of secondary prediction models; obtaining current feature data of a slope to be pre-warned; inputting the current feature into a primary/secondary prediction model to obtain a primary/secondary prediction result; weighting the primary prediction result and the plurality of secondary prediction results by using the initial fusion weight to serve as a current landslide prediction result; and comparing the current actual result of the slope to be pre-warned with the current landslide prediction result, and adjusting the fusion weight according to the comparison result as the fusion weight of the slope to be pre-warned in the next pre-warning. According to the method, a plurality of alternative models can be adopted f