The application provides a multi-stack fuel
cell hydrogen energy comprehensive power supply
system cooperative optimization method and
system, optimal working points of each stack are obtained through offline calculation; source and load are predicted in a prediction
time domain, a first-stage optimization is performed with the sum of working costs in the prediction
time domain and optimal working
point tracking performance indexes of each stack being minimum as an objective, and optimal outputs of each device in the prediction
time domain are obtained; wherein the working costs include:
system device operation and maintenance costs, wind and light abandonment penalty costs, grid interaction costs and stack performance degradation costs; source and load are predicted in a control time domain, a second-stage optimization is performed with the optimal output of the first stage as a benchmark value, with the sum of deviations of the optimal output of the second stage from the benchmark value and output fluctuations of each stack being minimum as an optimization objective function, and final
optimal scheduling results are obtained; the application improves the system economy, and improves the overall working efficiency of the stack and the system life.