Method for improving concentration of glutamic acid fermented product based on interval optimization

Abstract

The invention discloses a method for improving concentration of a glutamic acid fermented product based on interval optimization; the method comprises the steps: acquiring historical production data of glutamic acid fermentation process and establishing a three-layer BP (Back Propagation) neural network model of the glutamic acid fermentation process; expanding three-layer BP neural network model of the glutamic acid fermentation process to be an interval neural network model of the glutamic acid fermentation process; using PH value, fermentation temperature and ventilation quantity during fermentation as variables, using the concentration of the glutamic acid fermented product as an optimization target, performing interval optimization for the interval neural network model of the glutamic acid fermentation process by means of an interval particle swarm optimization algorithm and obtaining the optimal fermentation temperature, the optimal ventilation quantity and the optimal PH value of the fermentation process for the maximum concentration of the glutamic acid fermented product; and fermenting glutamic acid according to the optimal fermentation temperature, the optimal ventilation quantity and the optimal PH value of the fermentation process.

Description

technical field [0001] The invention belongs to the field of area optimization control of complex industrial processes, and in particular relates to a method for increasing the concentration of glutamic acid fermentation products based on area optimization. Background technique [0002] In the optimal control of industrial processes, the optimal solution obtained by traditional optimization algorithms is represented as an optimal point in the solution space, that is, point optimization. But in the optimal control of complex industrial processes, area optimization is more suitable. The so-called area optimization means that the optimal solution is expressed as an area, not just a point. Area optimization is an extension of the concept of point optimization, because a "area" can be regarded as a collection of countless "points", and when the boundary of the area shrinks to a "point", it is point optimization. For a dynamic process, the collection of "point data" at each time...

AI Technical Summary

Problems solved by technology

[0003] The reasons for using area optimization are: (1) With the progress of point optimization, after entering a certain area, the effect of continuing to optimize the optimization objective function is no longer obvious, and this area can be used as the final optimization (2) Even if the unreliable factors caused by the process knowledge that the optimization object depends on are not considered, only the influence of noise and disturbance on the working conditions in engineering practice makes point optimization difficult to achieve; (3) especially for Some irreversible processes, once the input seriously deviates from the set value, the serious consequences caused by it are irreversible

Images