Soft Sensing And Predictive Control In Marine Biological Enzyme Fermentation Process Based On The BLS-SVM

Marine microorganism is an important part of "blue agriculture". It possesses a lot of characters such as pressure resistance, alkali resistance, cold resistance and the characteristics of species diversity, which makes the enzyme preparation it produces have prominent characteristics and advantages. But the biochemical reaction process of biological enzyme has the features of time-varying, nonlinear, multivariate, strong coupling, which makes the internal reaction mechanism difficult to build. In addition,the key biological parameters are limited by hardware detection elements which causes real-time online measurement and optimization of control to be difficult,during the reaction. Due to this, it restricts the auto industrial development of marine biological enzyme. Therefore, being able to test key parameters and optimize control during the biochemical reaction of marine biological enzyme as well as in fermentation automatic control system, has far-reaching significance to the improvement of the biological fermentation equipment integrated automation level and economic benefits.This involves in marine biological enzyme fermentation technology and the study of fermentation equipment which cannot go online detection of biochemical reactions in the process of key biological parameters(such as cell concentration,residual sugar concentration, activity) and real-time control and other issues, which proposes the generalized predictive control based on soft measurement model. This project being based on the ARM9 fermentation control system, we design the software and hardware control system and develop the real-time monitoring interface of the host computer. The concrete work is as follows:Firstly, in order to solve the problem that crucial biological variables(such as biomass concentration, glucose concentration and enzyme activity) in are difficult to hardware devices detect, a modeling method based on the least squares support vector machine(LS-SVM)within the Bayesian evidence framework is proposed. The LS-SVM model which can predict biological parameters accurately in the process of fermentation is established. Comparison of the conventional LS-SVM model in test simulation, results show that the LS-SVM based on Bayesian inference has features of high prediction precision, good generalization ability, and rapid calculation.Secondly, ways on how to set precise models have been offered as the nonlinear characteristics of coupling in marine biological enzyme fermentation process.BLS-SVM nonlinear control model is set in marine biological enzyme fermentation process, as well as the piecewise linear precision model, with the adoption of sampling point and linearization, which possesses some advantages like simple structure and high accuracy.Thirdly, in order to solve the problem that the traditional partial filling material are difficult to achieve optimal control. On the basis of the generalized predictive control theory, how to control the continuous feeding in the process and how to optimize the control online has been raised, as well as design methods on predictive control based on the differential evolution optimization to calculate the optimal control sequence. Experiment and simulation results show that the generalized predictive control system for BLS-SVM model of based on the imperial differential evolution optimization has high control accuracy, good stability and can meet the optimal control requirements of batch feeding in the biological fermentation process.Finally, using embedded processor ARM9 as lower machine and PC as upper machine, the digital surveillance system is designed for the soft sensing of crucial parameters in fermentation process and the decoupling control of the marine biological enzyme fermentation process. In the hardware design of the control system,the monitoring system platform of Marine biological enzyme was designed by choosing reasonable hardware structures. Software design included the design of database module, serial communication module and monitoring interface module. The system run in the liquid state fermentation tank, and monitoring interface shows the excellent stability of the system during operation, and the effect is good.