Research Of Expert System Knowledge Database In Soft-Sensor Modeling Of Biomass In Fermentation Process

Animalcule fermentation engineering is the foundation of bioengineering as well as modern biology technology and bioengineering industrialization. As one of the most important variables in fermentation process, biomass directly influences the control and optimization of fermentation process. However, with the lack of biological sensors, there are no effective measurement methods for biomass at present. Applying soft-sensor technology to estimate biomass on-line is a new effective approach to solve this problem. During the soft-sensor modeling of the biomass in fermentation process, constructing soft-sensor hybrid model combining mechanism model and identification model is key points to estimate biomass on-line. Expert system can use experts' knowledge and experience to ratiocinate, judge and make decision, which is an effective way to construct the soft-sensor hybrid model. Knowledge base stores special knowledge from field expert and it is the one important part of expert system. Therefore, researching expert system knowledge organization method and knowledge base is meaningful theoretically for structuring soft-sensor hybrid model.In this paper, on the basis of detailed analysis of the current methods of biomass soft-sensor and expert system application, expert system knowledge organization method is realized through analyzing the features of biomass soft-sensor model knowledge in fermentation process. Knowledge representation structure of expert system is founded in fermentation process; the knowledge representation form and knowledge organization layered structure of knowledge base in expert system are given; redundancy verification algorithm based on research of rules is given to solve redundancy problem. Furthermore, knowledge base of expert system is constructed combined with database based on .net technology.Experiment research shows that, reasoning by layered structure knowledge organization method can reduce step and reasoning time obviously in the same search condition comparing with reasoning by sequential structure knowledge organization method. Redundancy verification algorithm is a good approach to solve redundancy problem in rules lay of knowledge base of biomass soft-sensor expert system.