Ground subsidence prediction method and equipment based on active RC circuit model, and medium

According to the land subsidence prediction method and device based on the active RC circuit model and the medium, real subsidence observation data of land subsidence monitoring points are collected, and a subsidence observation sequence is formed; constructing an active RC circuit model according to the collected land subsidence observation sequence, wherein the model is a first-order simplest RC circuit; a nonlinear least square method is used for parameter estimation, an optimal model parameter vector is calculated, an active RC circuit with bias voltage is built, and the voltage time change process of the two ends of a capacitor in the circuit correspondingly describes a land subsidence change rule; a circuit model is used for enhancing settlement data, a genetic algorithm (GA) is used for performing parameter optimization on a long short-term memory network (LSTM), and the enhanced data are input into an optimized LSTM model to realize ground settlement prediction. According to the invention, the change