Soft-sensing method for key parameters of photosynthetic bacteria fermentation process based on ba-lssvm

Abstract

The invention discloses a BA-LSSVM (Bat Algorithm-Least Square Support Vector Machines) based soft measurement method for key parameters of a photosynthetic bacteria fermentation process. According tothe invention, smart calculation is performed based on a hardware platform, measurement instruments and computer system software and real time online estimation is performed based on real time process data acquired through the measurement instruments. According to the invention, first, a proper auxiliary variable is selected according to a relevancy value. Then history tank batch data is collected and a fermentation data set divided into a training data set and a testing data set. An LSSVM model is designed and a BA is utilized for optimizing core parameters and punishment parameters of the LSSVM and the optimal BA-LSSVM model is acquired. Finally, the optimized model is utilized for predication of the key parameters. Therefore, real time online predication of the key variables of the photosynthetic bacteria fermentation process is realized. By utilizing the BA based LSSVM, defects in rate of convergence and local search capability are made up, optimization control of photosynthetic bacteria fermentation is facilitated and product yield and quality are improved.

Description

technical field [0001] The invention is used to solve the problem of online estimation of key parameters that are difficult to measure on-line in real time with physical sensors during the fermentation process of photosynthetic bacteria, and belongs to the technical field of soft measurement. Background technique [0002] At present, how to control the fermentation conditions of photosynthetic bacteria and improve product yield and quality as much as possible is the focus of research. The fermentation process is complex and changeable, and the factors that affect the micro-growth of photosynthetic bacteria (temperature, light intensity, pH value, bacterial concentration and other parameters) have a great influence on each other, and the internal relationship is complex. Traditional sensors can perform real-time online measurement of physical quantities such as temperature, pressure, and pH. However, there is no real-time online measurement equipment for bacterial concentrat...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the problem that it is difficult to use physical sensors for real-time on-line measurement or real-time measurement of expensive key parameters (such as cell concentration) in the fermentation process of photosynthetic bacteria, the present invention provides a method based on bat algorithm to optimize the least squares support vector machine (BA-LSSVM ) soft-sensing method for the key parameters of the photosynthetic bacteria fermentation process, the data of physical parameters such as temperature and pH value are obtained through conventional sensors, and the real-time online estimation of key parameters is realized by using the soft-sensing method

Images