The invention discloses an adaptive electric vehicle SOC estimation method based on big data. The time, longitude, latitude, SOC value, vehicle speed, total mileage value of the odometer, and total output current of the battery pack are collected in advance from the vehicle. and the total output voltage of the battery pack as the training data set, calculate the acceleration, distance, and energy consumption values according to the time, speed, latitude and longitude, current, and voltage values in the training data set, and then calculate the characteristic speed, acceleration, distance, and dependent variable energy consumption values It is used to construct an extremely random decision tree model, and then the SOC prediction model based on mileage and energy consumption is obtained from the total mileage value, energy consumption value and SOC value of the odometer, so that the SOC prediction model based on mileage and energy consumption is obtained according to the genetic algorithm. The SOC prediction model, every T time, the model will update the data in the training data set, so as to achieve the effect of adaptive prediction.