摘要
目的基于遗传算法与模拟退火算法相结合建立新的平滑指数模型,以亳州中药白芍为例对亳州中药材价格进行预测和实证分析。首先利用遗传算法对种群进行优化,通过基因的遗传、选择、交叉和变异逐渐产生近似最佳解,在利用模拟退火算法对其进行修正,得到最佳平滑参数值,建立新的指数平滑模型,运用建立的遗传模拟退火三次指数平滑模型对亳州白芍的价格进行预测。结果表明对于中药材价格随着时间的推移,价格具有非线性等因素,通过遗传模拟退火方法可以得到最佳的平滑参数值,为指数平滑预测的价格与实际价格相差较小,说明建立的遗传模拟退火三次指数平滑模型具有一定的准确性。遗传模拟退火三次指数平滑模型适合药材价格的预测,能够为市场和政府部门对市场的中药材价格调控起到的决策指导作用。
Based on the combination of genetic algorithm and simulated annealing algonthm, a new smoothing exponential model was established to forecast the price of Chinese medicine in Bozhou as an example. Firstly, the genetic algorithm was used to optimize the population, and the approximate optimal solution was generated by genetic inheritance, selection, crossover and mutation. The optimal smoothing parameter was obtained by using the simulated annealing algorithm to establish a new exponential smoothing model. The prediction of the price of white peony in Bozhou was carried out by using the genetic simulated annealing three - order exponential smoothing model. The results showed that the optimal temperature parameter could be obtained by genetic simulated annealing method for the price of Chinese herbal medicine, the price was nonlinear and so on. The difference between the price and the actual price of the exponential smoothing forecast was small, simulated annealing three - order exponential smoothing model had a certain accuracy. It could be concluded that the genetic simula- ted annealing three - order exponential smoothing model was suitable for the forecasting of the price of medicinal materials, which could serve the decision - making role of the market and the government departments on the market price regulation of Chinese herbal medicines.
出处
《佳木斯大学学报（自然科学版）》
CAS
2017年第5期857-860,874,共5页
Journal of Jiamusi University：Natural Science Edition
基金
安徽省教育厅自然科学重点基金项目(KJ2015A417)
安徽省教育厅自然科学重点基金项目(KJ2016A493)
亳州职业技术学院院级课题(BYK1511)
关键词
遗传算法
模拟退火
三次指数平滑
预测模型
中药材价格
genetic algorithm
simulated annealingChinese herbal medicine price
cubic exponential smoothing
prediction model
