Abstract: In order to solve the problem of many time scales, a hierarchical greenhouse environment control system was designed. In the system from lower to higher there were three sub-systems: measuring and controlling layer, optimal process layer and supervising layer. Plant environment optimal control strategies in intelligent greenhouse were researched so as to get the best photosynthesis rate and at the same time cut down energy consumption. In the daytime the optimal goals were decided according to the maximal photosynthesis rate principle. In the nighttime on better plant growth conditions the optimal goals were decided by energy saving principle. The object function was constructed by forecasting greenhouse inside temperature in case of no energy used and by the theory of integrated temperature. The optimal target values were computed by genetic algorithm. Compared with the optimal result and recording data in real system, the method was reasonable and could achieve energy saving and the maximal photosynthesis rate in case of restrained light.