Parameter controlling method in biofermentation process

Abstract

The invention discloses a regulating method of control parameters in biologic fermentation process, which includes steps of initially utilizing empirical data to train artificial neural networks in a lab, obtaining the weight of each node of the artificial neural network, then directly outputting control parameters of a PID controller on the site according to the measured production parameters via the artificial neural networks, and directly designing to realize the regulation according to the weight of the obtained nodes. The invention fully utilizes the empirical data accumulated by users in the actual production process, and directly maps the field production condition parameters into the PID control parameters via the system for the first time, thereby realizing self-adaptation in the control process. The artificial neural networks do not need to retrain, and the regulation can be realized directly according to the weight of each node obtained from lab statistic analysis. The invention has accurate and quick field parameter regulation, can self-adapt to various working conditions, and has fine control effect.

Description

Technical field [0001] The invention relates to a control parameter adjustment method in biological fermentation process control, is suitable for high-performance control of biological fermentation devices in bioengineering, and belongs to the technical field of bioengineering control. Background technique [0002] The biological fermentation process is a relatively complex production process, in which many indicators, such as temperature, air pressure, pH, etc., are crucial to the quality of the product. Although computer control technology has been widely used in fermentation industrial process control, its mainstream control mode is still based on traditional process control theory, and due to the complexity of the biological reaction process, its system control performance is not satisfactory. For manufacturers of complete sets of biological fermentation equipment, they need to provide products of different specifications and models according to different customer requir...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to overcome the deficiencies in performance degradation caused by the inaccuracy of the artificial neural network simulation mathematical model of the controlled object in the prior art, and at the same time to overcome the traditional artificial neural network training in real time when the system is running. The system is complicated to implement, and the training data cannot reflect the time-varying nature of the parameters in the biological fermentation process. A method for adjusting the control parameters in the biological fermentation process is provided. Real-time mapping control parameters of system working conditions

Images