基于神经网络算法的果树需水预测研究

刘勇; 何淑林; 刘慧敏; 金立强

引用本文:	刘勇,何淑林,刘慧敏,等.基于神经网络算法的果树需水预测研究[J].灌溉排水学报,0,():-.
	LIU Yong,HE Shu-lin,LIU Hui-min,et al.基于神经网络算法的果树需水预测研究[J].灌溉排水学报,0,():-.

【打印本页】【下载PDF全文】【查看/发表评论】【EndNote】【RefMan】【BibTex】

过刊浏览高级检索

本文已被：浏览 1135次下载 0次
分享到：微信更多字体:加大+\|默认\|缩小-
基于神经网络算法的果树需水预测研究
刘勇¹, 何淑林¹, 刘慧敏¹, 金立强²
1.黑龙江大学哈尔滨;2.黑龙江东部节水设备有限公司哈尔滨

摘要:

【目的】为了实现农业精准高效灌溉,构建了蒸腾量预测模型。【方法】首先,利用主成分分析的方法对采集的果园环境数据进行分析,筛选出影响果树蒸腾量的关键因子。然后,建立以长短时记忆神经网络为基础的预测模型来预测果树蒸腾量。为了提高预测的精度,在LSTM神经网络的基础上加入了注意力机制,构成Attention-LSTM预测模型。【结果】将本文的模型与其他模型的预测精度进行对比,经过仿真实验,证明本文提出的预测模型预测精度更高。【结论】本文提出的预测模型可以实现果园精准灌溉,提高水果产量,具有一定的实际意义。

关键词: 蒸腾量预测;LSTM神经网络;主成分分析;注意力机制

DOI：

分类号:TP301.6；S126

基金项目:中央地方引导项目(项目编号：SBZY2021E006);绥化市本级应用研究与开发项目

Study on Water Demand Prediction of Fruit Trees Based on Neural Network Algorithm

LIU Yong¹, HE Shu-lin¹, LIU Hui-min¹, JIN Li-qiang²

1.Heilongjiang University;2.Heilongjiang East Water Saving Equipment Co,Ltd

Abstract:

【Objective】In order to realize efficient and intelligent irrigation of orchard, transpiration prediction model was established in this paper.【Method】Some experiment were conducted. Firstly, Principal Component Analysis was used to analyze the collected orchard environmental data, and the key influencing factors were screened out. Then, a prediction model based on Long Short-Term Memory Networks was established to predict the transpiration of fruit trees. In order to improve the accuracy of prediction, Attention Mechanism is added to the LSTM neural network to form the attention-LSTM prediction model.【Result】By comparing the prediction accuracy of the model proposed in this paper with other models, the simulation experiment proves that the prediction model proposed in this paper has high prediction accuracy.【Conclusion】The prediction model proposed in this paper can achieve precision irrigation of orchard and improve fruit yield, which has a certain practical significance.

Key words: Prediction of transpiration; LSTM neural network; PCA; Attention mechanism