基于IPSO-BP模型的牧草产量预测方法研究

打开文本图片集

关键词牧草产量；预测模型；神经网络；气候条件；微粒群算法

中图分类号F326.12;TP183 文献标识码A文章编号 0517-6611（2026）02-0021-05doi：10.3969/j.issn.0517-6611.2026.02.003

AbstractTdresstheul-fctorcouplingallngsifageyieldpredictio，tisdyfocusothuoumunicipalaste research subject. By integrating data （including temperature，precipitation，sunshine duration，and 10cm ground temperature data） from six nationalbasiceteoroloicalsisfromO5to22，ndoulingwitthregion’saualforageyielddatandgowtstagelite quirementcharacteriticsoverthepastdecade，establisdaBnuralnetwork-basedfoundationalpredictionmodel.Teodelpaaetes wereoptimizedthroughtheitroductionofstandardpartieswarmoptization（PSO）andimprovedparticleswarmoptization（SO）alg rithms.SimulatoncomparativestudiesdemonsratehattheISOoptiidBeuralnetwrkacevessignificantpeformanehacent inpredictionacuracy，exhibitinglowerMeanAbsoluteErorcomparedwithoththebasicBPodelandPSO-BPmodel.Thegorit provedmodelalsodmonstratesstrongergeneralizationcapabilies，withISOshowinghigherglobaloptizationefiiencytanbasicO. Thesefindingsvalidateteaplicationvalueofteimprovedalgoritmfrfrageyieldpredictionudercomplexmeteorologicalcoditos.

Key wordsForage yield;Prediction model;Neural network;Climatic conditions;Particle swarm algorithm

畜牧业的发展离不开草产业的支撑[1]。（剩余8632字）