摘要
马铃薯是我国重要的粮食作物之一,营养丰富,用途广泛,是一种谷物、蔬菜和水果功能兼具的食物,其蛋白质含量远高于其他块茎类食物,且富含优质的氨基酸。马铃薯生育期短,在湖北平原、丘陵地区冬种春收适宜发展早熟品种,对于填补全国南北方鲜薯市场供应空档期具有重要意义。因此,马铃薯产量的高效预测对于制定生长期间的种植管理措施及相关决策具有重要意义。为此,针对传统BP神经网络在产量预测中存在精度低、鲁棒性差等问题,利用鲸鱼算法(Whale optimization algorithm,WOA)对BP神经网络模型进行优化。同时,基于湖北地区2009-2021年间田间物联网获取的气象因子(大气湿度、大气温度、降雨量)、田间水热因子及马铃薯产量,采用BP神经网络模型、GA-BP神经网络模型(遗传算法优化)及WOA-BP神经网络模型对所选地区马铃薯产量进行预测。研究结果表明:WOA-BP神经网络模型精度明显高于GA-BP神经网络模型及BP神经网络模型,R2达到0.9764,预测值与试验值之间拟合程度较高,表明基于WOA-BP神经网络模型可以更加科学、合理、准确地进行马铃薯产量预测。
Potato is one of China's important food crops,rich in nutrients,widely used,is a grain,vegetables and fruit function of both food,its protein content is much higher than other tubers,rich in high-quality amino acid structure,potato fertility period is short,in the Hubei plain,hilly areas in winter and spring harvest suitable for the development of early varieties,for filling the supply gap between the country's southern and northern fresh potato market has an important Significance.Therefore,efficient prediction of potato yield is important for making planting management measures and related decisions during the growth period.In this study,the BP neural network model was optimized using the Whale optimization algorithm(WOA)to address the problems of low accuracy and poor robustness of traditional BP neural networks in yield prediction.In this study,based on the meteorological factors(atmospheric humidity,atmospheric temperature,rainfall),field water and heat factors and potato yield obtained from field IOTs in Hubei region during 2009-2021,BP neural network model,GA-BP neural network model(optimized by genetic algorithm)and WOA-BP neural network model were used to predict potato yield in the selected areas.The results showed that the accuracy of the WOA-BP neural network model was significantly higher than that of the GA-BP neural network model and BP neural network model,with R2 reaching 0.9764 and a better fit between the predicted and experimental values.Therefore,the WOA-BP neural network model can be used to predict potato yield more scientifically,rationally and accurately,which is an important guidance for potato growth and adjustment of field planting management measures.
作者
赵丙秀
董宁
Zhao Bingxiu;Dong Ning(Wuhan Vocational College of Software and Engineering,Wuhan 430205,China)
出处
《农机化研究》
北大核心
2024年第3期47-51,共5页
Journal of Agricultural Mechanization Research
基金
教育部高校学生司第一期供需对接就业育人项目(20220105995)。
关键词
马铃薯
神经网络模型
产量预测
鲸鱼优化算法
potato
neural network model
yield prediction
whale optimization algorithm