Abstract: This paper establishes a multi object fuzzy reliability optimization model with fuzzy reliability optimization theory and a matter element method. By using im proved operators, simulating annealing and lozenge thinking, an improved genetic algorithm that adopts a real code method is obtained. In this algorithm, crosso ver rate and mutation rate can be adjusted automatically during evolution. Optim ization of planar gear transmission with multiple objects is achieved.

Key words: improved genetic algorithm, matter element, multi object, fuzzy reliability optimization