Price Prediction Of Chinese Herbal Medicine In Longxi Based On BP Neural Networks

In recent years,with the rapid development of the traditional Chinese medi-cine industry,the demand for Chinese medical herbs both in domestic and interna-tional markets has increased sharply.Chinese medical herbs have the dual nature of drugs and subsidiary agricultural products,the prices are easily affected by various factors,such as the national or local policy regulation,market demand,the planting area,natural environment and human factors,etc.Since these various factors cor-related each other,there is strong randomness and uncertainty for the price.Therefore,it is an urgent problem for the government,enterprises,investors and vast herbalists to know how to deal with the connection between various fac-tors and how to reduce or eliminate the negative effect that caused by price change as possible as far to stabilize the market of the traditional Chinese medicine indus-try.In view of this,the thesis used the BP neural network with nonlinear character-istics to predict the price of Chinese medicinal materials,so that provide references for the scientific decision-making of the government and efficient management de-cision-making of Chinese medicinal enterprises.First of all,the thesis made a comprehensive summary of the basic situation of the development of the Chinese medicinal materials market and analyzed the characteristics and causes of price changes.The analysis was mainly made from the three aspects:market supply level,demand for Chinese medical herb and char-acteristics of Chinese medical herbs market.Then,through the data analysis and collection of codonopsis pilosula in longxi region,a BP neural network price prediction model based on L-M improved algorithm was proposed.The monthly price data of the wholesale market in Longxi,the main producing area of radix codonopsis,were collected and sorted,which constituted the time series data of the price change of radix codonopsis.On this basis,the BP neural network prediction model based on the improved L-M algo-rithm is verified.The experiment results show that compared with the traditional BP neural network prediction model,the improved L-M algorithm-based predic-tion model is more accurate in predicting the price of Chinese medicinal materials.At the same time,the characteristic analysis and correlation prediction of a variety of Chinese medicinal materials were conducted.Finally,the demand analysis of the Chinese medicinal materials price predic-tion system is carried out,according to the analysis results of modular design and implementation,the Chinese medicinal materials price prediction model is applied to the Chinese medicinal materials price prediction system,a Chinese medicinal materials price prediction software system is realized.The correlation analysis and prediction of several kinds of Chinese medicinal materials were further carried out,and the price prediction of one or several kinds of Chinese medicinal materials under the influence of each other was obtained.