Thermal power plant start-up scheduling with evolutionary computation by using an enforcement operator

Power plant start-up scheduling is aimed mainly at minimizing the start-up time of both boiler and turbine, while limiting turbine rotor stresses to acceptable values. This problem has a number of local optima. In order to find the optimal start-up schedule efficiently, we apply population-based evolutionary optimization techniques-genetic algorithms (GA)-with our proposed “enforcement operator”. The purpose of this enforcement operator is to limit the search space to those promising areas where the optimal solution is supposed to exist. As shown in this paper, aided with this enforcement operator, the search efficiency improves significantly as compared to a GA-based search without such an operator. In addition, an optimal solution can be achieved, which reduces the start-up time by approximately 10%, or 20 minutes, relative to conventional methods