Because of its implicit parallelism and global searching ability,Genetic Algorithm(GA) becomes the widely used Algorithm in resolving Job Shop Scheduling Problems(JSP).However,with JSP's specialty,the satisfying resolution can hardly receive from simple GA(SGA),of which the most obvious flaw is it always converges at a local extremum instead of the global one.This paper describes the optimization of GA,by way of improving the initialized population,reduce the selecting pressure and add SA operator into GA,t...