This research is an optimization method and system for power dispatching of equipment in dyeing and finishing plants. Taking the TTRI Yunlin Branch as an example, the trained AI model can predict the sensor parameter data of the equipment in the factory area, can accurately predict the time when the equipment is overloaded, and feed the prediction result back to the dispatching system to reset the electricity consumption schedule in stages according to the severity, so that the production line can continue to run without exceeding specifications.