Design Of Soft Sensing And Monitoring System For Matsutake Fermentation Process

Matsutake,known as the King of bacteria,has attracted much attention because of its anti-cancer and anti-tumor effects,and is now widely exported to Europe,Japan and other places.Market demand led to large-scale exploitation of matsutake,which seriously damaged the growth environment and survival space of wild matsutake.Therefore,it is urgent to find an alternative method of wild production of matsutake in order to ensure the effective utilization of the resources of matsutake.It was found that the mycelium and metabolite obtained by liquid deep fermentation were not only similar to wild fruiting bodies in nutritional value,but also shortened the growth cycle and protected the resources of Tricholoma matsutake.However,the internal mechanism of matsutake fermentation is complex and has high nonlinear and strong coupling.The key parameters(mycelium biomass,extracellular polysaccharide)which directly reflect the fermentation quality are difficult to be directly measured online by hardware sensors.If the traditional off-line test rules are used,there are some problems such as lag,pollution and so on,which seriously hinder the industrialization development of matsutake fermentation.A soft-sensing model of matsutake fermentation was constructed in this paper,and a monitoring system of matsutake fermentation process was designed by combining with embedded technology.The paper first introduces the research background and significance of Matsutake.By analyzing its fermentation process,the influencing factors and dominant variables are determined.Secondly,a soft sensor model based on least squares support vector machine(LS-SVM)is established based on the fermentation data.The problem of blindness and time consuming for selecting LS-SVM parameters is discussed.By changing the original fixed discovery probability into two-stage dynamic discovery probability,the cuckoo algorithm(CS)is improved and applied to the parameter optimization of LS-SVM.The simulation results show that the soft sensing model has high prediction accuracy and generalization ability.Then combining soft-sensing technology with embedded technology,transplanting soft-sensing prediction model into embedded system,using S3C6410 as processor,designing a monitoring system with the functions of data acquisition,soft-sensing processing,man-machine interface display,etc.The system has been applied to the industrial production of matsutake fermentation,and the feasibility of this system has been proved.