烟苗移栽时间评估模型研究与应用  

Research and Application of a Model for Evaluating the Transplantation Time of Tobacco Seedlings

作　　者：李群岭[1] 李力 林小兴 梁桂广 周效峰[1] 韦育新[1] 胡亚杰[1] 冯烨君 成鑫 冉渝澳 孙佳照 LI Qunling;LI Li;LIN Xiaoxing;LIANG Guiguang;ZHOU Xiaofeng;WEI Yuxin;HU Yajie;FENG Xejun;CHENG Xin;RAN Yu’ao;SUN Jiazhao(China Tobacco Guangxi Industrial Co.Ltd,Nanning 530000,China;School of Plant Protection,Southwest University,Chongqing 400715,China)

机构地区：[1]广西中烟工业有限责任公司,南宁530000 [2]西南大学植物保护学院,重庆400715

出　　处：《植物医学》2023年第6期54-63,共10页Plant Health and Medicine

基　　金：广西中烟工业有限责任公司项目(0633-224042118J00).

摘　　要：为了使移栽后的烟苗获得最优生长,从而提高烟草种植的效率和产量.本研究对烟株移栽后的健康状况与移栽时间的关键数据进行分析.采用多元回归分析、随机森林分析、BP神经网络3种模型对烟苗移栽时间的关键因子进行研究,寻找烟苗最优移栽时间并分析影响最优移栽时间的关键因素.采用J2EE平台Spring应用框架对模型应用进行开发,对模型研究结果进行可视化表达.结果表明:多元回归分析模型精准度为70.58%,随机森林分析模型精准度为67.82%,BP神经网络分析模型精准度为88.95%.苗床生长时间、移栽前10~15 d大田降雨量、移栽时3 d内平均最低气温、移栽时3 d内平均最高气温、苗床期大于10℃活动积温是影响移栽时间的关键因素.通过采集重庆地区烟草育苗棚生长基地内烟苗移栽相关数据进行模型精准度验证,3个模型中BP神经网络分析模型精准度最高.本研究为烟苗移栽时间的判定提供合理科学依据,通过该应用可以提前预判烟株移栽时间,从而对生产进行指导,同时也为数字化烟草种植业提供了一个重要的应用实例.In order to achieve optimal growth of transplanted tobacco seedlings and improve the cultivation efficiency and yield of tobacco.This study analyzes the key data on the health status after transplantation and transplanting time of tobacco plants.The key factors affecting the transplanting time of tobacco seedlings were studied using three models:multiple regression analysis,random forest analysis,and BP neural network to find the optimal transplanting time for tobacco seedlings and analyze the key factors affecting the optimal transplanting time.Spring application framework of the J2EE platform was employed to develop the mode of application and visualize the research results of the model.The results showed that the accuracy of the multiple regression analysis model,the random forest analysis model and the BP neural network analysis model was 70.58%,67.82%,and 88.95%,respectively.The key factors affecting the transplanting time are the period time of seedling raising,the rainfall in the field 15~10 days before transplanting,the average lowest temperature and highest temperature within 3 days of transplanting,and the≥10℃accumulated temperature during the seedling raising.The accuracy of the model was verified by collecting relevant data of tobacco seedling transplanting from the seedling greenhouse of tobacco growth base in Chongqing.The BP neural network analysis model has the highest accuracy among the three models.The study provides a reasonable scientific basis for determining the transplanting time of tobacco seedlings.This application can be used to predict the transplanting time of tobacco plants in advance to guide the production.This study provides an important application example for the digital tobacco planting industry.

关 键 词：烟苗移栽 关键因素 模型分析 应用开发 移栽时间判定 

分 类 号：S572[农业科学—烟草工业]