Abstract:

To obtain the accurate solutions of the steady state of aircraft, a hybrid genetic algorithm was proposed according to the analysis of the basic characteristics and constraints of the steady state. Based on the concept of “individual learning potentiality”, this algorithm logistically integrates the Lamarckian learning and Baldwinian learning mechanisms. It rationally distributes the number of local search among the population to make the advantage of the learning into full play, meanwhile inhibit its disadvantage. The algorithm can solve any preassigned steady state on the basis of the state variables rather than the real form of the aerodynamic model. Simulation results validate the proposed algorithm that combines the two learning mechanisms.

Key words: artificial intelligence, genetic algorithm, steady flight state, individual learning potentiality, Lamarckian learning, Baldwinina learning