There are problems of difficult identification of the traveling wave head and difficult extraction of inherent dominant frequency in hybrid DC transmission system fault location. Thus a fault location scheme based on a gate recurrent unit (GRU) model optimized by a wavelet energy spectrum and the sparrow search algorithm (SSA) is proposed. First, the correlation between spectrum energy and fault distance is analyzed, and wavelet packet decomposition is used to extract the wavelet packet energy spectrum feature vector as the input to the GRU model. Second, the GRU model is built and trained to mine deep-seated fault information in the time series, and the parameters of the GRU model are optimized using the iterative optimization of the SSA, so as to realize the rapid and accurate location of the fault distance. Finally, a hybrid three-terminal DC transmission system model is built in PSCAD/EMTDC. The experimental results show that this method has high positioning accuracy, strong anti-interference ability and generalizability, and has a certain resistance to transition resistance.