Soft-sensing method for key parameters in edible fungus fermentation production process

Abstract

The invention discloses a soft measurement method for key parameters in the fermentation production process of edible fungi, which is used to solve the problem of online estimation of key biochemical quantities that are difficult to measure in real time on-line in the fermentation of edible fungi. Firstly, through the analysis of the process mechanism of edible fungus fermentation process, select appropriate auxiliary variables and establish a training sample database according to the historical tank batch data; combine the leading variables and auxiliary variables of the current tank batch fermentation process to be predicted with the historical fermentation process leading variables and historical auxiliary variables The variables constitute the least squares support vector machine soft sensor training sample database and construct the corresponding soft sensor model, and then use the gray wolf algorithm to optimize the regularization parameter γ and kernel parameter σ in the soft sensor model 2 And establish a soft sensor model based on the gray wolf optimization least squares support vector machine, and finally obtain the predicted value of the corresponding key biochemical parameters; the present invention adopts the gray wolf optimization algorithm for simulating the behavior of gray wolves, which has simple structure, few parameter settings, and strong global search ability. And does not take advantage of gradient information.

Description

technical field [0001] The invention belongs to the technical field of soft measurement and soft instrument structure, and specifically relates to a method for solving the three key biochemical problems of bacteria concentration, substrate concentration and edible fungus product quality that are difficult to measure online and real-time with physical sensors during the fermentation production process of edible fungi. On-line Estimation Problems of Variables. Background technique [0002] With the rapid development of the edible fungi industry, my country's traditional small-scale lagging production mode can no longer meet the requirements of the current market. Therefore, the realization of edible fungus liquid submerged fermentation and automatic production will be the mainstream planting mode in my country in the future. However, the first problem encountered in the application of advanced automated production is that the quality of edible mushroom fruiting bodies and othe...

AI Technical Summary

Problems solved by technology

When establishing the LS-SVM soft sensor model of the fermentation process of edible fungi, the regularization parameter γ and the kernel parameter σ 2 directly affects the fitting performance and generalization ability of the soft sensor model, but in actual use, the regularization parameter γ and kernel parameter σ of the LS-SVM soft sensor model 2 The probability of falling into a local extremum is high, and the global optimization ability is poor, which affects its measurement accuracy

Images