Multivarialbe Boiler-Turbine H_infty Control System Genetic Algorithm Weighting Functions $W_1$(s), $W_2$(s), and design parameter $gamma$ that are given by Glover-Doyle algorithm, to optimally follow the output of reference model. The first method to do this is that the gains of weighting functions $W_1$(s), $W_2$(s), and design parameter are optimized simultaneously by genetic algorithm with the tournament method that can search more diversely, in the search domain which guarantees the robust stability of system. And the second method is that not only by genetic algorithm with the roulette-wheel method that can search more fast, in that search domain. The boiler-turbine H_infty control system designed by theabove second method has not only the robust stability to a modeling error but also the the better command tracking preformance than those of the H_infty control system designed by trial-and-error method and the above first method. Also, this boiler-turbine H_infty control system has the better performance than that of the LQG/LTR contro lsystem. The effectiveness of this boiler-turbineH_infty control system is verified by computer simulation.