The optimization algorithm of unit combination for fossil fuel power plant is studied in this paper. To resolve the problem of general-purpose algorithm of unit combinatorial optimization, the combinatorial problem of M units is abstracted into an algorithm problem which is to take out a number from each of M arrays and the sum of these M numbers is equal to a given value. Based on this, a new enumerative algorithm of unit combination optimization is proposed, which is general-purpose and able to fit any varying number of units. The complexity of this algorithm is analyzed. The recursive backtracking method is used to search for every possible combination. The correctness and completeness of the algorithm are verified by experiment. All of the possible solutions are searched and the optimal solution can be found. The drawback of recursive backtracking method is longer time of calculating. So it is only fit in fewer units. The exploring research for applying shuffled frog-leaping algorithm to resolve the unit combinatorial optimization is carried out. A shuffled frog-leaping algorithm (SFLA) of unit combination optimization is proposed, which is also general-purpose and able to fit any varying number of units. The complexity of this algorithm is analyzed. The algorithm has good global searching and local replacement tactics. The experimental results of the two algorithms proposed are compared and analyzed. There are drawbacks for using shuffled frog-leaping algorithm to resolve the unit combinatorial optimization, such as poor stability, the optimal solution is not guaranteed, etc. When the number of unit is fewer, it takes more calculating time for shuffled frog-leaping algorithm than that of recursive backtracking method. But when the number of unit is larger, the calculating time for shuffled frog-leaping algorithm is shorter than that of recursive backtracking method. |