共查询到20条相似文献，搜索用时 0 毫秒
1.
《植物保护学报》2016,(3)
为建立更准确、稳定的病虫害预测预报模型,减少农作物病虫害损失、提高农作物产量与质量,运用主成分分析法从42个基础气象因子中整合形成8个新的自变量输入模型,采用试凑法对网络关键参数进行筛选,用2002—2011年数据进行网络训练,建立了以Morlet小波函数为传递函数的小波神经网络模型,并与以Sigmoid函数为传递函数的BP神经网络模型进行了比较。在小波神经网络训练过程中,有6年拟合精度在90%以上,平均拟合精度为89%,预测结果 MAPE值为4.1939,MSE值为5.9764;在BP神经网络的训练过程中,有4年拟合精度超过90%,平均拟合精度仅为81.07%,预测结果中MAPE值为6.4694,MSE值为8.2457。从训练结果看,小波神经网络更能准确描述麦蚜发生期的变化规律,其拟合能力较BP神经网络好;从预测精度和模型的稳定性来看,小波神经网络好于BP神经网络。 相似文献
2.
采用常规气象观测资料建立了兰州市太阳总辐射BP神经网络预测模型,利用神经网络释义图和连接权法剔除了模型中的冗余变量,用优化的BP网络模型预测了兰州市1996-2000年的太阳辐射,并用实测数据验证了该模型。结果表明:该方法增加了模型的透明度,提高了模型的可靠性和鲁棒性,模拟结果与实测值非常吻合,模拟值的各项误差指标值均很小,模拟值与实测值的拟合优度R2达到0.987,通过与其他经验模型的模拟结果进行对比,优化的BP网络模型的模拟效果最好,精度明显高于其它经验模型。因此,对于无太阳辐射观测的地区,优化后的BP网络模型是预测当地太阳辐射的一种有效方法。 相似文献
3.
BP神经网络和多元线性回归在粮食产量空间分布预测中的比较 总被引：2，自引：0，他引：2
根据2011年河北省92个平原县粮食产量数据,按农业生产过程和需求确定自然和经济两大类影响因素共8个指标。根据逐步多元线性回归进行回归建模,并运用BP神经网络建立影响因素和粮食产量的网络模型。对比两种模型的拟合效果,BP神经网络模型在精度、稳定性、泛化度和理论基础上都要优于线性回归模型,可用于时空混合的空间预测。最后根据可估相关因子的变化代入BP模型对2015年粮食产量进行空间预测。从多角度说明了BP神经网络在空间分布预测中的优势性。 相似文献
4.
引入遗传算法优化BP神经网络权重和阈值的方法建立黄土坡面产流入渗模型.模型以雨强、降雨历时、表层40 cm土壤前期含水量、坡度值为输入项,径流量、入渗量为输出项,用实测资料对网络进行模拟和预测.模拟结果平均误差6.32%和1.93%,预测结果平均误差为5.71%和1.92%.并与传统BP神经网络模型和定雨强Philip... 相似文献
5.
BP神经网络和支持向量机在积温插值中的应用 总被引：1，自引：0，他引：1
积温插值是分布式积温获取的重要途径。为了提高积温插值的精度,应用BP(Back Propagation Learning Algorithm)神经网络模型和支持向量机(Support Vector Machine,SVM)模型建立甘肃省及周边地区的积温插值模型。结果显示:1)从积温插值的空间分布来看,SVM模型比BP神经网络模型能够体现出更多的细节;2)SVM模型的插值精度总体上显著高于BP神经网络模型;3)在平均相对误差(MRE)最大的西区,相比BP神经网络模型的7.19%,SVM模型将误差降低到了5.47%;4)东区两种模型的MRE最小,BP神经网络模型为2.97%,SVM模型为2.03%;5)与分区建模前相比,分区后的插值精度有所提高,BP神经网络模型将MRE降低了0.04%,SVM模型降低了0.11%。 相似文献
6.
分析了塔里木盆地南缘和田绿洲洛浦灌区地下水位较高的成因,研究了灌区地下水位年内的动态变化规律,利用小波分析的多分辨率功能和人工神经网络的非线性逼近功能,建立了基于小波变换和BP神经网络的地下水埋深动态预测小波神经网络模型。结果表明:灌区地下水位变化具有周期性、季节性的特点,并且可以被小波神经网络模型准确地模拟,模型预测洛浦灌区在未来几年内地下水位会持续上升,年平均升幅为0.1 m左右,因此当地应加强地下水的科学管理。 相似文献
7.
晚稻稻瘟病BP神经网络分区预报 总被引：2，自引：0，他引：2
应用相关分析方法分析了浙江省19个县1988～1999年晚稻稻瘟病发病与有关环境因子的关系,筛选了8个气象因子用于晚稻稻瘟病发生程度长期预报。根据各预报因子与稻瘟病发病程度相关性,采用邻接二维图论聚类分析法,将19个点(县)划分为4个生态区。每个生态区内运用BP神经网络技术建立模型,并进行拟合和试报。1997～1999年试报验证,在划分稻瘟病生态区的基础上,应用BP神经网络模型对稻瘟病进行长期预测预报是可行的,3年试报成功率分别是78.95％、84.21％和78.95％。文中还对该方法与过去常用的预报方法的试报结果作了比较。 相似文献
8.
几种方法在粮食总产量预测中的对比 总被引：4，自引：2，他引：2
根据吉林省1978~2007年的农业数据,选定了14个影响农业生产的因素作为研究对象,分别采用了主成分分析法、BP神经网络法、灰色预测法和逐步回归分析法4种分析预测方法,通过SPSS和Matlab工具将原始数据进行处理,得到4种不同的预测模型,进而基于这4种模型对吉林省的粮食产量进行预测,并将各种预测产量和实际产量进行拟合分析。研究结果表明,拟合性最好的是BP神经网络法,其拟合确定性系数为0.899;其次是主成分分析法(拟合确定性系数为0.834)和逐步回归法(拟合确定性系数为0.787);拟合效果最差的是灰色预测法(确定性系数为0.744)。粮食总产量估算精度最高的是BP神经网络法,达到93.67%;其次是主成分分析法,为90.45%。 相似文献
9.
10.
区域农作制优先序是区域农业结构调整的主要依据。本研究通过对西北旱作农区农业自然资源、农田生态环境和农村经济发展水平等现状的调研,总结出西北旱作农区主要的7种农作制度。并应用BP神经网络模型,以随机技术模拟生成的评价指标序列与其所属的评价等级值进行网络训练。网络训练后,将不同农作制度的评价指标为网络输入,通过计算可以得到相应的农作制优先序的评价等级值。结果表明,西北旱作农区农作制的优先序为:粮草果畜复合制粮棉(油)轮作制特色产业(苹果)型农作制抗旱节水型农作制小杂粮、草畜主导型农作制两粮一肥型农作制粮油豆草肥田制。本方法只要给定相应的评价指标值,通过BP神经网络模型的计算,可直接得出农作制度优先序的评价等级值,可以用于农作制度优先序的评价。 相似文献
11.
Lesions of tea (Camellia sinensis) leaves are detrimental to the growth of tea crops. Their adverse effects include further disease of tea leaves and a direct reduction in yield and profit. Therefore, early detection and on-site monitoring of tea leaf lesions are necessary for effective management to control infections and prevent further yield loss. In this study, 1,822 images of tea leaves with lesions caused by three diseases (brown blight, Colletotrichum camelliae; blister blight, Exobasidium vexans; and algal leaf spot, Cephaleuros virescens) and four pests (leaf miner, Tropicomyia theae; tea thrip, Scirtothrips dorsalis; tea leaf roller, Homona magnanima; and tea mosquito bug, Helopeltis fasciaticollis) were collected from northern and central Taiwan. A faster region-based convolutional neural network (Faster R-CNN) was then trained to detect the locations of the lesions on the leaves and to identify the causes of the lesions. The trained Faster R-CNN detector achieved a precision of 77.5%, recall of 70.6%, an F1 score of 73.91%, and a mean average precision of 66.02%. An overall accuracy of 89.4% was obtained for identification of the seven classes of tea diseases and pests. The developed detector could assist tea farmers in identifying the causes of lesions in real time. 相似文献
12.
Yam (Dioscorea spp.) is a tropical tuber crop produced for food and medicinal purposes. Yams are infested by a broad taxonomic diversity of insect pests. We reviewed a total of 73 insect species associated with Dioscorea species in different parts of the world. Yam is infested by 48 species when the crop is in the field, and after harvest (i.e. in storage) the tubers are attacked by 27 species. The majority of these insects belong to the order Coleoptera (35 species) followed by the orders Hemiptera (15), Lepidoptera (13), Isoptera (5), Hymenoptera (2), Diptera (1), and Thysanoptera (1). Yam scales, mealybugs, and a few beetles cause significant losses to tubers both in the field and in storage. We review various methods of management for these insect pests. An integrated approach towards managing these pests, both in the field and in storage, is essential. Post-harvest losses in storage of yam can be reduced, partly, by using biocontrol agents for mealybugs and scales. Alternative strategies for the management of insect pests of tubers in storage are discussed. 相似文献
13.
Ernesto Gianoli Isaías Ramos Armando Alfaro-Tapia Yolvi Valdéz Erik R. Echegaray Erick Yábar 《国际虫害防治杂志》2013,59(4):283-289
Abstract In the Urubamba Valley, Peruvian Andes, we evaluated the abundance of insect pests and natural enemies in experimental plots where maize was grown either as monoculture, intercropped with beans, or intercropped with beans plus associated weeds. We also assessed the consequences of the cropping system for maize yield. The main insect pests recorded were Diabrotica spp. (Coleoptera: Chrysomelidae), Carpophilus sp. (Coleoptera: Nitidulidae) and Pagiocerus frontalis (Coleoptera: Scolytidae), and their overall abundance did not differ among crop diversity treatments. However, there was a significant adverse effect of crop diversity on the maximum abundance of both Carpophilus and Pagiocerus. The main beneficial arthropods were Paratriphleps sp. (Hemiptera: Anthocoridae), ladybirds and spiders, and their density did not differ among treatments. Maize yield did not decrease with increasing plant diversity. The reported benefits of intercropping, together with the associated efficiency in land use, make this traditional agricultural practice a valuable alternative to the use of pesticides, particularly for resource-poor Andean farmers. 相似文献
14.
N. Monaghan 《国际虫害防治杂志》2013,59(2):172-176
Abstract Sorghum halepense (L.) Pers. (Johnson grass) is an important weed in many countries. In Australia it invades crop lands, provides a harbour for diseases and insect pests of sorghum and maize, contaminates sorghum grown for seed and constitutes a safety hazard along roadsides. 相似文献
15.
A. K. Raheja 《国际虫害防治杂志》2013,59(2):229-233
Abstract A method for assessing the damage by insect pests to cowpeas is presented. A crop life table for cowpeas grown in northern Nigeria shows the estimated losses due to insect attack at various stages of growth of the plants. The potential loss in yield due to insects is over 90%, with 70% of this loss occuring during the flowering and pod formation stages. 相似文献
16.
为实现大田作物灌溉的精细化管理,研究了基于气象因素的生育期ET0预测模型。采用灰色理论对ET0与日均、日最高、最低温度,日均风速,相对湿度以及日照时数进行灰色关联度分析,结果表明ET0与温度（包括日均、最高、最低温度）及相对湿度的灰色关联度较高。在分析ET0与上述气象因素间的相关系数基础上,采用日均温度、日均风速以及日照时数作为模型的输入,ET0作为输出,建立了BP神经网络（BPNN）预测模型;采用日均温度、日均风速、日照时数及灰色关联度作为输入,建立了模糊最小二乘支持向量机（FLSSVM）预测模型。研究结果表明,BPNN模型的训练值决定系数为0.8643,平均相对误差6.29%,预测值决定系数为0.8099,平均相对误差7.83%;FLSSVM模型的训练值决定系数为0.9684,平均相对误差2.89%;预测值决定系数为0.9663,平均相对误差3.43%。BP神经网络与FLSSVM模型的精度均较高,可以用来预测ET0日值,这为大田作物的精细化灌溉管理提供理论与技术支持。 相似文献
17.
氯虫苯甲酰胺种子处理对花生地上及地下害虫的防治效果 总被引：1，自引：0，他引：1
为明确氯虫苯甲酰胺种子处理对花生地下害虫及地上刺吸式口器害虫的防治效果,筛选出合适的田间用药剂量,通过室内盆栽接虫试验和田间试验研究氯虫苯甲酰胺对花生地下害虫沟金针虫Pleonomus canaliculatus（Faldermann）、铜绿丽金龟Anomala corpulenta Motschulsky以及地上害虫花生蚜Aphis craccivora Koch的综合防治效果。结果表明,盆栽药剂拌种处理后15 d,氯虫苯甲酰胺有效成分用量为2.40、3.20 g/kg种子处理对沟金针虫的防治效果为77.42%和82.58%,对铜绿丽金龟的防治效果为77.12%和85.62%;田间拌种处理收获时,2.40、3.20 g/kg种子剂量的氯虫苯甲酰胺对花生田沟金针虫的防治效果为68.12%和76.33%,对铜绿丽金龟的防治效果为71.38%和80.71%;3.20 g/kg种子剂量处理对花生蚜的防治效果最高,为36.39%;该药剂拌种处理后对花生种子无不良影响,对出苗和生长有一定的促进作用,2.40、3.20 g/kg种子剂量处理后增产率为16.16%和18.44%。表明氯虫苯甲酰胺种子处理对花生地下害虫沟金针虫和铜绿丽金龟具有很高的防治效果和持效性,同时对花生蚜也有一定的兼治效果,有明显的防虫增产作用,可推广应用于防治花生田地下害虫。 相似文献
18.
Michael C. Dike 《国际虫害防治杂志》2013,59(3):191-192
Field experiments were carried out at Samaru, Nigeria in the 1993 and 1994 cropping seasons to investigate the effect of insect pests on the yield of bambara groundnut. There was significant reduction in pod and seed yields of bambara groundnut in both years due to insect pests. The application of a mixture of Karate e.c. at 25 g a.i./ha and dimethoate e.c. at 400 g a.i./ha, respectively, at fortnightly intervals from crop flowering significantly increased pod and seed yields of bambara groundnut, preventing pod losses of 67% and 38% and seed losses of 71% and 43% in 1993 and 1994, respectively. 相似文献
19.
Artificial infestation methods were employed to generate damage function graphs for guilds of three rice insect pests. In the vegetative stage, infestation of whorl maggot (Hydrellia philippina Ferino) and defoliators (a mixed population of Naranga aenescens Moore and Rivula atimeta [Swinhoe]) was applied as a combined or single pest attack. Infestation of yellow stemborer (Scirpophaga incertulas [Walker]) and leaffolder (Cnaphalocrocis medinalis [Guenée]) each were infested in one or two crop growth stages. Combinations of three stresses (four N application rates, solar radiation from the wet or dry season, and/or the presence or absence of weeds) were applied to each guild to develop more holistic economic injury levels (EILs). Slopes on the linear portion of damage curves of each pest guild became steeper with each additional stress applied, whether from multiple insect pest attack, low N, low solar radiation, or weeds. EILs can be developed by pest managers from the graphs of the damage functions for the variables quantified in this study. 相似文献
20.
Abstract Weeds are major components of agro-ecosystems and they affect the biology of pests and beneficial insects in several ways; provision of flowers, presence of neutral insects, modification of crop microclimate, production of chemical stimulus, alteration of colonisation background, etc. The more relevant beneficial crop–weed–insect interactions are discussed. The role of weeds in regulating pest populations is illustrated by studies on bean (Phaseolus vulgaris L.) cropping systems, in which chemical interactions between Empoasca kraemeri Ross &; Moore and two grass weeds were identified. It is proposed to continue and develop this kind of research in order to incorporate weeds into pest control strategies. 相似文献