Study On The Control Of Traditional Chinese Medicine’s Production Based On Genetic Neural Network

Granulation and coating are two key aspects in the production of the Chinesemedicine tablet. At present, the production of the traditional Chinese medicine industryhas many defects, such as the application of the advanced technology is not skilled, thelevel of automation and intelligent control is not high. These caused the traditionalChinese medicine had low active ingredients, impurities, inconsistent quality, and highproduction costs, waste and energy, which restricted the traditional Chinese medicineindustry standardization, modernization and internationalization. Therefore, usingadvanced unit to manufacture technology and equipment, applying online inspectionand other new methods and new technologies and optimizing of the process parametersare of great significance for the achievement of the upgrades and transformation of theproduction technology and equipments in the production of traditional Chinesemedicine, solving the problems of control of the quality in the production, ensuring thestability of the product quality, reducing the production costs, enhancing the safety oftraditional Chinese medicine production process and improving the technologicalcontent of the products. The paper is supported by national863planning issues: theresearch on the online detection technology of coating.This paper mainly introduces the following two aspects:1. the modeling in the process of Chinese granulation and coating: firstly, useneural network to establish a network structure model which takes process parametersas the input and granulation/coating product rate as the output. Then use the data samplewhich is collected with experience for experiment on the DPL-1multi-functionalgranulation and coating machine to train the network. And determine the mappingrelation between granulation/coating product rate and the process parametersrespectively. Finally, use genetic algorithm to optimize the input parameters of themodel to find the best process parameters which can make product rate the highest. Atthe same time, use the existing method for coating for Caoshan Hu Hanpian. Andcompare the confluences on the product rate by these two methods.2. the assessment criterion of the quality of granulation and coating: eachassessment criterion includes many contents. This paper takes coating thickness anduniformity into account when the coating quality is in the test. The product rate is calculated by controlling the assessment criterion strictly to verify the accuracy and thepracticality of the model.Finally, verify the feasibility of the method by simulation experiments. Theexperimental results show that the model in the control of the production of Chinesemedicine tablet based on genetic neural network can reduce the production costs,improve the product rate. And the product rate is higher by using the proposed methodin this paper for coating. So it can be used to guide the actual production.