Soft measurement of pichia pastoris fermentation based on ISSA-XGBoost
Aiming at the problems that in fermentation process of Pichia pastoris,it is difficult to detect in concentration of bacteria online,and offline measurement has the problems of incomplete datasets caused by easy infection of bacteria,a soft measurement modeling method based on improved sparrow search algorithm(ISSA) optimized extreme gradient boosting (XGBoost )is proposed.Firstly,the principal component analysis (PCA ) algorithm is used to analyze on sample data to reduce the noise and redundancy.Secondly,adaptive hyperparameters and hybrid mutation strategy are introduced to standard sparrow algorithm(SSA)for enhancing the ability of the algorithm to jump out of local extremum and global search.Finally,the ISSA-XGBoost soft measurement model of cell concentration is constructed and compared with XGBoost and SSA-XGBoost models.The simulation experimental result shows that the root mean square error(RMSE)and mean relative error(MRE)of the ISSA-XGBoost model are lower than those of the XGBoost and SSA-XGBoost models,and the determination coefficient R2 of the ISSA-XGBoost model is closer to 1,indicating that the prediction precision is significantly better than that before the improvement.The demand for real-time measurement of bacteria concentration in Pichia pastoris fermentation process can be satisfied.
Pichia pastorissparrow algorithmextreme gradient boosting(XGBoost)soft measurement model
万方数据
维普
