The invention discloses a battery SOC robustness evaluation method based on PLSTM sequence mapping. The battery SOC robustness evaluation method comprises the following specific steps: step 1, normalizing training and test data; 2, slicing a sliding window sequence; step 3, constructing and training an auto-encoder based on PLSTM; step 4, constructing and optimizing a sequence mapping model basedon PLSTM; and step 5, using the test data to complete SOC evaluation. On the basis of a basic LSTM unit, process information is introduced into input and gating, the PLSTM is provided, the sequence mapping model is constructed and trained on the basis of pre-training of the auto-encoder on the basis of the PLSTM unit, and SOC robustness evaluation is completed. According to the evaluation model, the association between the battery state and SOC can be learned, and the influence of the charging and discharging process on the SOC can be learned, so that high-accuracy SOC evaluation can be completed under the adverse conditions of variable sampling frequency, position load profile and the like, and the evaluation model has relatively high practical engineering application value.