The invention relates to a production scheduling and equipment maintenance collaborative optimization method based on a Stackelberg game, and the method comprises the steps: building a Stackelberg game model, taking a production scheduling party as a leader, and taking an equipment maintenance party as a follower; constructing a production leader agent and a maintenance follower agent; executing two-stage reverse training based on meta learning: firstly training a meta maintenance follower agent to enable the agent to quickly adapt to any
production strategy and generate an optimal response; taking the current to-be-trained production leader agent as a reference
fine tuning element agent to obtain a paired maintenance follower agent, and training the paired production leader agent under the support of the optimal response of the paired maintenance follower agent; and sequentially deciding according to input production task data and
equipment state data by utilizing the trained agent pair, and outputting an approximate Stackelberg
equilibrium solution to obtain a collaborative optimization result. Compared with the prior art, the method has the advantages of high efficiency, high reliability, accordance with actual working conditions and the like.