| 引用本文: | 冯兆宇,崔天时,张志超,等.基于灰色神经网络与模糊控制的寒地水稻灌溉制度[J].灌溉排水学报,2018,37(4):71-79. |
| FENG Zhaoyu,CUI Tianshi,ZHANG Zhichao,et al.基于灰色神经网络与模糊控制的寒地水稻灌溉制度[J].灌溉排水学报,2018,37(4):71-79. |
|
| 摘要: |
| 【目的】精确判断寒地水稻的灌溉水量并建立适当的灌溉管理方式，保证作物正常生长需求，起到节水效果。【方法】根据寒地水稻特殊的生长环境和作物各生育期需水量，设计了基于灰色神经网络与模糊控制的寒地水稻灌溉制度，该智能灌溉制度通过建立微型气象站监测、传输稻田环境数据，并通过灰色神经网络预测理论预测出作物灌溉需水量和灌溉制度影响因子；以预测作物灌水量和作物最佳灌水量的差值及差值变化率作为模糊控制器的输入，灌溉时间长度为输出，驱动电磁阀，达到适时适量灌溉的目的。【结果】 MATLAB仿真结果表明，该灌溉控制方式比传统控制方式响应速度快、超调量小、控制效果好。在东北农业大学水稻试验田的试验结果表明该灌溉控制制度的节水率为11.59%，水稻产量和结实率也有所提高；在黑龙江省建三江农场的田间试验表明该灌溉制度的节水率高达13.54%。【结论】 该灌溉制度与传统控制方式相比具有很好的节水效果，能对作物各生育期灌溉需水量进行综合判断和管理，对提高水资源利用率、降低农业生产成本、实现精细农业有重要意义。 |
| 关键词: 微型气象站； 灰色神经网络； 模糊控制； 灌溉制度； 灌溉需水量 |
| DOI：10.13522/j.cnki.ggps.2017.0386 |
| 分类号: |
| 基金项目: |
|
| Calculating Irrigation Schedule for Rice in Temperate Regions Using a Combination of Grey Neural Network and Fuzzy Control |
|
FENG Zhaoyu, CUI Tianshi, ZHANG Zhichao, WANG Rui, LIU Chunli, WANG Lifeng
|
|
College of Electric and Information, Northeast Agricultural University, Harbin 150030, China
|
| Abstract: |
| 【Objective】 The purpose of this paper is to propose new models to help increase the water use efficiency in irrigation for rice in temperate regions. 【Method】 We used grey neural network and fuzzy control to design irrigation schedule for rice in temperate regions, considering both specific growth environment and demand of the rice for water during different growth stages. Meteorological data in rice field were measured and then transmitted from a micro meteorological station; the factors affecting the demand for irrigation strategy were estimated using the grey neural network model. The difference between the estimated demand for water and the optimal demand for water by the crop, as well as the change rate of their difference with time were taken as inputs in the fuzzy control; the duration of the irrigation was the model output. Field experiments were conducted to verify the model. 【Result】 The irrigation schedule calculated by the model had advantage of quick response, small overshoot and good controllability. The field experiment conducted in the rice field of Northeast Agricultural University showed 11.59% saving in water, and the experiment in Jian Sanjiang farm of Heilongjiang Province showed 13.54% saving in water, compared to the conventional irrigation schedule. 【Conclusion】 The proposed models for calculating irrigation schedule can increase water use efficiently and could play an important role in precision agriculture. |
| Key words: micro meteorological station; grey neural network; fuzzycontrol; irrigation schedule; rice in cold region |
