Abstract: Based on CNC tracing grinding machine with double grinding carriages, which is the core equipment in the automobile engine crankshaft production line, a research on the optimization of crankshaft grinding process is made to shorten machining used time and reduce production cost through reasonable planning double grinding wheels' work steps and formulating maintenance strategies. To fully develop the advantage of double grinding carriages servo systems, regard the grinding process as a multiple travelling salesman problem, build the grinding process optimization model at which make average machining used time and production cost of single crankshaft as optimization target, use cycle time of production line and structure of grinding machine as restrain. This model still considers the time of dressing grinding wheels and replacing the scrap grinding wheels during long period grinding process in production line except the grinding time, the movement time and the auxiliary time, more according with actual situation. An additional optimization model which allows one of the grinding carriages return to safety position waiting is built as supplement when there is no feasible solution satisfying the condition that double grinding carriages always work simultaneously due to the structural constraints of grinding machine, making the technology has more practicability. The grinding process of a specific crankshaft is optimized under different sets of weights to obtain the corresponding optimal solutions. The correctness of the proposed model and the validity of the algorithm are verified by comparing the solutions before and after optimizing.

Key words: non-circular grinding, process optimization, multiple traveling salesman problem, crankshaft production line, improved genetic algorithm