Abstract: The conventional methods of depth peaking for supercritical units mainly focus on input variable conditions, ignore the coordination of peak balancing, and fall into the problem of frequent peak balancing. Therefore, an optimization method for deep peaking operation of supercritical unit based on tabu search heuristic algorithm is designed. The primary FM dead zone of supercritical units is divided, and the range between the lower FM dead zone and the upper FM dead zone is determined according to the static characteristics of primary FM, so as to avoid the unit falling into the situation of frequent peak regulation. Based on the heuristic of tabu search, the optimization model of unit peak load balancing operation was constructed, the fitness of the peak data of supercritical units was sorted, and the adaptive peak load balancing set of tabu search was established to ensure the global optimization of deep peak load balancing operation. The depth peaking load of supercritical unit is optimized, and the unit output is adjusted according to the state of charge and maximum output power of supercritical unit, so as to realize the depth peaking of supercritical unit. The simulation results show that the method has better optimization effect and can be applied in real life.