Studying Of Soft Measurement Technology And Implementation Based On HAFBPLS In Hyaluronic Acid Production

With the development of modern science and technology,people have higher requirements for industrial products.This is mainly reflected in the following aspects :timeliness, energy saving, convenience, traceability,etc.For example, in the field of chemical pharmaceutical industries based on the fermentation and other fields,it has been unable to meet the requirements of process operation and control during industrial production only by obtaining the temperature, pressure, liquid level, PH value and so on. Due to the complex and nonlinear material transformation and energy transfer in industrial production, it is hard for measuring parts of the variables of the process.In order to obtain the parameters which can not obtained by some existing technology in the production, we not only need the high-precision sensors,but also need comprehensive analysis models to reflect the relationships between different parameters and to obtain the values of the corresponding parameters. At present, the conventional methods are divided into two categories: a part of variables can be measured by the special sensors, and the others are detected by hand.However, the later methods are both costly and time-consuming. So far, there is no better solution to predict the concentration of the yield of fermentation in the hyaluronic acid fermentation industry. In order to solve this problem, the hyaluronic acid fermentation production concentration prediction method and the corresponding soft measurement model were researched and established in this paper. The experimental results show that there is a good prediction effect. At the end,we introduces the software of “remote monitoring and control system for fermentation process”.The main works of this paper are as follows:(1)Analysis the modeling methods of soft measurement model and describe the advantages and disadvantages of the mainstream modeling methods, which mainly contain Neural Networks and Support Vector Machines. Choosing the appropriate modeling algorithm for the prediction of hyaluronic acid. We take the Least square support vector machine(SVM)as the modeling algorithm.According to the influence factors of hyaluronic acid fermentation,we build the most appropriate soft measurement model by calculating the influence factor of each variable on the target variable.We also detail the steps of modeling.(2)The paper propose HAFBPLS(Hyaluronic Acid Forecast Based on PSO and LS-SVM),which is used to predict the concentration of hyaluronic acid.We select The Particle Swarm Optimization(PSO) to optimize the parameter of soft sensor model. We analysis the reason of the particle swarm optimization algorithm for the optimization of the parameters of a soft measurement model, and then discusses the principle of particle swarm optimization algorithm and its implementation steps. This paper compare the soft measurement model with other mainstream optimization algorithms(e.g. genetic algorithm).The result of experiment show that PSO-LSSVM has high prediction accuracy.(3) Build the prediction module of hyaluronic acid based on the soft measurement of HAFBPLS, and optimizing the soft measurement model by the limited memory method.(4)Realizing the remote monitoring system of the fermentation process, mainly including data reading module, historical data analysis module, real-time data display module, the transparent quality acid prediction module,and successing in using on cooperative enterprises.