Multistage Soft Sensing Modeling And Analysis Of Photosynthetic Bacteria Fermentation Process

With the continuous improvement of China's comprehensive strength,the constraints on resources and environment are becoming more and more serious.In particular,the contradiction between supply and demand of water resources that depends on survival is very prominent,and more than half of the cities in the country have insufficient water supply.However,the overall pollution of rivers and lakes in China is serious,and the water quality of groundwater resources is deteriorating,which has seriously jeopardized people's health and hindered the development of economy and society.China's economy is entering a new stage of the new normal in an all-round way,and the prevention and control of water pollution has become the focus of concentration.Photosynthetic bacteria have many functions such as phosphorus removal,denitrification,desulfurization,hydrogen production,oxidation of sulfides,and also have good decomposition and conversion of heavy metals and toxic organic substances.These excellent functional properties are of great value in the treatment of various wastewaters,and the bacteria themselves are rich in a variety of high nutritional value substances,which has attracted people's attention.Due to the wide geographical area and large population in our country,the demand for high-quality photosynthetic bacteria is extremely high.In the actual production fermentation process,in order to obtain high yield and quality of photosynthetic bacteria,it is necessary to optimize the fermentation process according to the progress of fermentation,so as to provide a suitable environment for their bacterial growth.However,the fermentation process of photosynthetic bacteria is complex and variable,showing multivariable and nonlinear characteristics,especially the key variable(living cell concentration)of the reaction fermentation progress is difficult to measure online in real time,which is not conducive to real-time optimization control of the fermentation process and affects the improvement of fermentation quality.Based on the above problems,this topic introduces soft sensing technology and establishes a multi-stage soft sensing model to predict the living cell concentration in the fermentation process of photosynthetic bacteria.Firstly,the paper introduces the purpose and significance of research on photosynthetic bacteria,and points out that it is difficult to measure the living cell concentration in the fermentation process.Secondly,it analyzes various soft sensing modeling methods at home and abroad,and summarizes the corresponding advantages and disadvantages.Then the fermentation process of photosynthetic bacteria is analyzed,the specific liquid fermentation process is detailed,the influence of fermentation process parameters on fermentation and the detection methods of fermentation parameters is described,which is conducive to the selection of auxiliary variables and the effective collection of sample data,so as to prepare for the establishment of the model.Thirdly,the principle of soft sensing technology is introduced,the least squares vector machine(Least squares vector machine,LSSVM)is selected as the basic soft sensing model,and an improved bat algorithm(Improved bat algorithm,IBA)is adopted to optimize the model parameters of LSSVM.Under the condition of neglecting the variation characteristics of photosynthetic bacteria fermentation process,a single global soft sensing model of living cell concentration based on improved bat algorithm optimized least squares support vector machine(IBA-LSSVM)is established and verified by MATLAB simulation.Compared with the unoptimized BA-LSSVM model,the soft measurement method has better prediction accuracy and anti-interference ability.However,due to neglecting the fermentation characteristics of photosynthetic bacteria,there is still room for improvement in prediction accuracy.Finally,the fuzzy c-means algorithm(Fuzzy c-means algorithm,FCM)is used to cluster the sample data in this paper,and the fermentation process is divided into three stages.Then the IAB-LSSVM is used to establish corresponding local sub-models for these three stages,and then the sub-models are combined to obtain the final multi-stage soft-sensing model.Through simulation,compared with the single global IBA-LSSVM model,the multi-stage model has better stability and prediction performance,which can solve the online detection problem of living cell concentration in the fermentation process.The multi-stage soft sensing method based on IBA-LSSVM for photosynthetic bacteria fermentation in this paper has high prediction accuracy and stability,which provides a reference method for solving variables that are difficult to directly measure in the same type of industrial production process.