Abstract
This paper aims at the shortcomings of manual offline sampling and offline analysis of the traditional bio- fermentation technology, proposes an intelligent online monitoring system for the bio-fermentation, presents the overall system structure, detail hardware and software design, and then using actual examples to verify the effectiveness of the system design in this paper. The system is embedded with improved continuous hidden Markov measurement model, which can effectively monitor the biological fermentation parameters online and utilize the minimum classification error to optimize the CHMM to achieve the purpose of controlling the convergence error of the algorithm. The simulation results show that the mean variance of the proposed algorithm is 0.182, and the calculated mean variance of the neural network algorithm reaches 0.911. The simulation algorithm proposed in this paper is significantly higher in the calculation accuracy than that of the neural network algorithm. At the same time, by using reliability we can better identify the simulation results and avoid blind control.
