Abstract: In order to solve the structural optimization problem of long-span transmission tower, topology combination optimization (TCO) method and configuration combination optimization (CCO) method based on discrete variables are presented respectively. The adaptive genetic algorithm (AGA) is used to search optimization solution. Firstly the fitness function of AGA is obtained by using the penalty function. Then the change schemes of genetic coding are proposed and the adaptive crossover and mutation operators are acquired. Finally, four optimization methods are discussed using the AGA, including rod cross-section optimization (RCSO) method, shape combination optimization (SCO) method, TCO method and CCO method. The topology optimization rules are presented for the TCO method, and the layering optimization rules are presented for the CCO method. The simulation results demonstrate that the optimization results of the proposed TCO method and CCO method are obviously better than those of the RCSO method and SCO method. Complex change in the configuration of tower is permitted by The CCO method, thus the optimization result is the best in these methods.