Parameter-correlation prediction method for injection molding process based on multi-gene genetic programming
In this article,since the selection and setting of parameters in the traditional injection molding process depend on manual experience,and there is a nonlinear and strong coupling relationship between the process parameters and the output pa-rameters,a parameter-correlation prediction method for the injection molding process is proposed based on multi-gene genetic pro-gramming.This method generates an initial multi-gene tree population according to the specific data,and then performs iterative selection,high-level and low-level tree crossover,as well as two types of tree mutation operations,in order to identify the most suitable model for the data;besides,efforts are made to evolve an algebraic equation that can clearly describe the relationship be-tween the process parameters and the output parameters,so as to determine the key parameters for output control.When setting the selection index,since the model equation's complexity will affect the efficiency in utilization,the complexity dimension is in-troduced as an evaluation index for multi-objective selection.With the injection molding process as an example,the experiments show that this method is effective.In addition,the parameter and sensitivity analysis has verified the model's robustness and re-vealed the hidden nonlinear relationship between various parameters.This study provides basis for parameter setting and optimiza-tion of the subsequent injection molding process.
parameter-correlation predictionmulti-gene genetic programmingmulti-objective selectioninjection molding process
万方数据
维普
