红兴隆管理局农业水文要素复杂性测度及其发展态势研究

发布时间：2017-12-31 15:25

本文关键词：红兴隆管理局农业水文要素复杂性测度及其发展态势研究　出处：《东北农业大学》2014年硕士论文　论文类型：学位论文

【摘要】：红兴隆管理局是我国重要的粮食储备基地和商品粮生产基地。如今,在国家粮食政策和经济利益双重影响下,管理局内水稻种植量迅猛增加,变旱为水面积迅速增长,局部超采现象时有发生,加之地表水、地下水利用存在较严重的不平衡、用水结构不合理等问题日益突出,实现生态可持续利用、经济利益最大化已经成为本区域研究重点。因此,从和谐发展理念出发,深入研究红兴隆管理局农业水文要素复杂性测度分析方法,揭示其复杂性空间变异规律,分析农业水文要素发展态势,进而实现农业水资源优化配置等,对促进区域经济发展及保障我国粮食产能安全具有重大意义。本文以红兴隆管理局下属12农场为例,运用符号动力学理论、分形理论正确诊断红兴隆管理局农业水文要素复杂性并排序,建立了复杂性视角下的复杂农业水文要素和农业水文要素复杂性预测模型。主要研究内容和结论如下： (1)采用等概率粗粒化LZC算法、多尺度半方差分维算法分别对红兴隆管理局各农场逐月地下水埋深序列、逐月降水序列进行复杂性诊断,对比分析诊断结果,筛选出最优测度方法。考虑到该方法粗粒化段数选取人为干扰因素大,引入自适应人工鱼群优化关键参数,以优化后的等概率粗粒化LZC算法诊断结果为最终评价结果：①逐月地下水埋深序列：红旗岭农场、二九一农场、曙光农场、北兴农场地下水埋深序列复杂性最高,江川农场、八五二农场、五九七农场、双鸭山农场复杂性一般,八五三农场、友谊农场、宝山农场、饶河农场复杂性最低；②逐月降水序列：八五三农场、八五二农场、红旗岭农场复杂性最高,双鸭山农场、江川(宝山)农场、饶河农场复杂性一般,友谊农场、北兴(曙光)农场、二九一农场、五九七农场复杂性最低。 (2)采用小波神经网络(WNN)、灰色小波神经网络(GWNN)算法构建复杂性视角下的红兴隆管理局复杂农业水文要素预测模型,结果表明：①灰色小波神经网络为红兴隆管理局各农场地下逐月水埋深序列、逐月降水序列最优预测模型；②友谊农场地下水埋深序列预测结果：2013、2015、2016年地下水埋深有上升趋势,2014年地下水埋深有下降趋势,纵向比较可知,地下水埋深序列有先上升后下降再上升趋势；③红旗岭农场地下水埋深序列预测结果：地下水埋深呈上升、下降波动变化,纵向比较可知,地下水埋深序列有先上升后下降再上升趋势；④597农场降水序列预测结果：降水序列呈上升、下降再上升趋势；纵向比较可知,降水序列整体呈上升下降趋势,降雨量峰值集中在6-9月之间；⑤853农场降水序列预测结果：降水序列下降趋势,只有2016年降水量略有回升；纵向比较可知,降水序列整体呈上升下降趋势,降雨量峰值集中在6-9月之间。 (3)构建基于复杂性诊断结果及最优预测模型结果的红兴隆管理局地下水埋深复杂性体系、降水复杂性体系,以地下水埋深序列最复杂的红旗岭农场、降水序列最复杂的853农场为例,分析红兴隆管理局农业水文要素复杂性动态变化规律,结果表明：①红旗岭农场地下水复杂性序列有整体下降趋势,2013年下半年复杂度值急剧下降,2014年复杂性先上升后下降,年内的复杂度值有一定的周期现象；②853农场降水复杂性序列有整体下降趋势,2012、2013年复杂度值急剧下降,2014年～2018年复杂性上升趋势,年纪内的复杂度值有一定的周期现象。
[Abstract]:Hong Xinglong administration is an important grain reserve base and commodity grain production base in China. Now, in the national food policy and economic benefits under the dual influence of authority, the rapid increase in the amount of rice planting, dry water area increased rapidly, partial overpumping have occurred, and the surface water, groundwater is not balanced the more serious water problems, such as unreasonable structure have become increasingly prominent, realize the sustainable utilization of the ecological and economic benefit maximization has become the research focus. Therefore, starting from the concept of harmonious development of science, bureau of agriculture and hydrological complexity measure analysis method in-depth study of Hongxinglong tube, reveal the complexity of spatial variability, analyzes the development trend of agricultural hydrological factors then, to achieve the optimal allocation of agricultural water resources, is of great significance to promote regional economic development and ensure grain production safety in China.
In this paper, Hong Xinglong administration under the 12 farm as an example, using symbolic dynamics theory, fractal theory, the correct diagnosis of Hong Xinglong administration of agricultural and hydrological complexity sorting, build the complex agricultural and agricultural hydrological hydrological complexity from the perspective of the forecasting model. The main research contents and conclusions are as follows:
(1) using equiprobable LZC algorithm, multi-scale fractal algorithm of semi variance of Hong Xinglong administration of each farm monthly groundwater depth series, monthly precipitation sequence comparative analysis of complexity of diagnosis, diagnosis, screening out the optimal measurement method. Considering the method of coarse grain number selection much disturbed factors the introduction of adaptive artificial fish swarm optimization, the key parameters to optimize the equiprobable LZC algorithm diagnosis results as the final evaluation results: Monthly groundwater depth series: Hongqiling farm, 291 farm, dawn farm, Beixing farm groundwater depth series highest complexity, Jiangchuan farms, 852 farms, 597 farms, Shuangyashan the 853 farm, farm complex, friendship farm, Baoshan farm, farm Raohe complexity is lowest; the monthly precipitation sequence: 853 farms, 852 farms, Hongqiling The highest complexity farm farm in Shuangyashan, Jiangchuan, (Baoshan) farm, farm Raohe complex, friendship farm, North Hing (Shu Guang) farms, 291 farms, 597 farms in the lowest complexity.
(2) using wavelet neural network (WNN), grey wavelet neural network (GWNN) algorithm to construct the complexity from the perspective of Hong Xinglong administration of agricultural complex hydrological forecasting model, the results show that: 1. The grey wavelet neural network depth sequence of Hong Xinglong administration of each farm underground water monthly, monthly precipitation sequence optimal prediction model; prediction the friendship farm groundwater depth series: 201320152016 years of groundwater depth has increased, in 2014 the groundwater depth decline, longitudinal comparison, groundwater depth series have increased after the first drop and then increased; the prediction results the Hongqiling farm groundwater depth series: groundwater depth increased, decreased fluctuations. The longitudinal comparison shows that groundwater depth series have increased after the first drop and then rise again; the prediction results of 597 farm precipitation sequence: precipitation increased, under the Fall and then rise again; the longitudinal comparison, the overall upward trend of precipitation, rainfall concentrated in the peak between 6-9 months; the prediction results of 853 farms: precipitation precipitation decreased, only the precipitation in 2016 rose slightly; the longitudinal comparison, the overall upward trend of precipitation, rainfall concentrated in the peak between 6-9 months.
(3) the construction complexity of diagnosis and optimal prediction model based on the results of Hong Xinglong administration system of underground water depth, complexity of precipitation system, Hongqiling farm to groundwater depth series the most complex, the most complex precipitation sequences of 853 farm as an example, analysis of red prosperous changes, complexity of dynamic Agricultural Bureau hydrological management results show: the Hongqiling farm groundwater sequence complexity have declined overall trend, the second half of 2013 the complexity value fell sharply in 2014 the complexity increased after the first drop, the complexity of value during the year is periodic phenomenon; the 853 farm precipitation sequence complexity overall downward trend, 20122013 years complexity value fell sharply from 2014 to 2018. The complexity of a rising trend, the complexity of the value in the period of age phenomenon.

【学位授予单位】：东北农业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F323.213;S271

【参考文献】