Optimizing parameters of semi-solid-state-extrusion process by genetic algorithm

The semi-solid state extrusion process is a kind of new process, which has a promised application future for producing tube and bar products. But there is a problem for the process in practice, that is, the process parameters are difficult to select and optimize. For solving the problem, the genetic algorithm has been adopted to optimize the parameters of the semi-solid state extrusion process. Some key problems, such as the coding principle, initial population generation, mapping of the object function to the fitness function, adjusting the value of fitness, reproduction number, crossover and mutation probability, had been investigated. Based on the results, the appropriate genetic parameters are determined and the optimized process parameters for semi-solid extrusion are obtained. The experimental effect for extruding AlCuSiMg alloy was excellent with the obtained parameters. The deforming load can be efficiently decreased. The calculated results are well coincided with the experimental ones and the relative error is lower than 3.9% and the absolute error is not larger than 3.4 MPa, which prove the optimizing method is feasible.