局域法邻近点选取对供水量预测精度的影响

混沌局域法预测模型适用于非线性、非平稳的城市日供水量预测,而邻近相点个数的选取对该模型预测精度有直接影响。传统方法通常以嵌入维m作为参考值,凭经验选取m+1个邻近相点,且仅使用欧式距离法计算当前相点距离,无法反映相点的运动趋势,易引入伪邻近相点,导致预测精度的降低。鉴于此,将演化追踪法引入城市日供水量预测,通过挖掘邻近相点的历史演化规律对参考样本进行优选,以提高预测精度。最后,采用实际日供水量数据验证所提出方法,结果表明,运用演化追踪法优选邻近相点能显著提高日供水量预测精度,预测平均绝对误差由2.501%降低到1.683%。     

Application of modified Deep Belief Network in Forecasting Cotton Diseases and Insect Pests

[Objective] The occurrence and development of cotton diseases and insect pests are mainly related to environmental information. Because this environmental information is various, complex and unstable, the study on the prediction methods of cotton diseases and insect pests is a certain challenge. This study aims to establish a forecasting model for the timely and accurate prediction of cotton diseases and insect pests. [Method] A forecasting model of cotton diseases and insect pests is proposed based on environmental information and a modified Deep Belief Network (DBN) that is constructed by a three-layer restricted Boltzmann machine (RBM) and a supervised back-propagation (BP) network. In the method, the RBM is used to transform the original environmental information vectors into a new feature space related to the diseases and pests; the BP network is trained to classify and forecast the features generated by the last RBM layer and two rules of dynamic learning and comparison and dispersion are adopted to accelerate the training process of RBM. The proposed model was validated on a dataset of cotton bollworm, aphids, spider, cotton Verticillium wilt, and Fusarium wilt in a recent six-year period. [Result] Compared with the traditional prediction models of cotton diseases and insect pests, the proposed model can deeply explore the extensive correlation between the occurrence of cotton diseases and pests and environmental information. The results show that the proposed model has a higher accuracy compared with the classical predictive models, and the average forecasting accuracy is above 83%. [Conclusion] The proposed method is an effective crop disease and pest forecasting method that can provide a technical support for preventing cotton disease and insect pests.     

河南省粮食产量影响因素和预测方法

粮食生产是国民经济重要的组成部分,粮食产量对于保证我国的粮食安全具有重要意义。旨在提高粮食产量预测的科学性和准确性,在分析现有预测方法的基础上,文中将灰色理论和神经网络有机地结合在一起,通过灰色理论关联度分析,在众多影响粮食产量的因素中确定出主要的、客观的因素指标,通过这些指标利用人工神经网络具有的非线性建模和极高的拟合精度特点,应用到粮食产量预测中去。结果表明：人工神经网络预测的最大误差1.21%,平均误差0.63%。预测精度较高。为粮食产量预测提供了一种科学的、有效的预测方法。     

A Fuzzy Regression Method for Long-Term and Medium-Term Load Forecasting

In view of the observation data fuzziness and load pattern fuzziness,a new fuzzy regression prediction method was presented for long-term and medium-term load forecasting. With the established fuzzy regression model, the future load value can be forecasted based on the fuzzy historical observation data. The validity of the proposed method was verified with the numerical example of a practical system.     

不同插值方法对精细化预报产品在青岛地区的检验比较

青岛地区农业产业园位置分散在内陆、沿海和山地等多种地形区域,气温差异大,为提高现代农业精细化服务水平,笔者基于国家气象中心下发的全国区域5 km格点气象要素预报产品,采用最近邻域、双线性插值、反距离权重3种插值方法对气温预报产品在青岛地区的适用性进行对比检验。结果表明:内陆地区双线性插值法准确率最高,且平均绝对误差和均方根误差最小。沿海地区采用最近邻域法和反距离权重法均有较好的效果。对于山地,综合考虑准确率、平均绝对误差和均方根误差,较适用反距离权重插值方法。最高气温预报准确率内陆地区高于沿海和山地地区,且在8—10月为准确率最高时期,3—6月准确率相对较低;最低气温则在沿海地区准确率最高,且夏季的准确率明显高于冬季,其在7月准确率最高而1月准确率最低。研究结果为下一步将预报产品插值生成分辨率更细、准确率更高的精细化气象服务产品奠定基础。     

Application of Combination Grey Model in Power Load Forecasting

Based on the gray forecast theory, this paper studies the principle and deficiency in power load forecasting by the basic grey model and other improved models, and introduces a new method -the combination grey model to forecast the long-medium power load. Based on an example, the basic grey model, other improved models and combination grey model are used to forecast power load and results of all models are analyzed and compared. The calculation results show that forecasting power load by grey theory is credible and simple. For this type of complex problems such as forecasting the long-medium power load, the combination grey model is specially useful because of it's high precision and facility. The method can be used as one of the tools of forecasting the long-medium power load.     

Short-Term Electric Load Forecasting Based Trends Combination Model

In short term load forecasting based on ANN,weather is one of the important factors which impacts on load greatly. In order to capture the effect of weather on load, this paper presents a novel thought based on ANN and trends combination short term load forecasting. Decompose the underlying relationships between load and weather variables into three main trends of weekly, daily and hourly. Three separated ANNs capture each trend. Another ANN to arrive at the final forecast combines the forecasts yielded by individual ANNs. The performances of the proposed model and the traditional model are compared on the basis of one week ahead hourly forecasts. Results indicate that the proposed ANN based model can achieve greater forecasting accuracy than the traditional ANN based model.     

Data Analysis Method for Neural Network Based on Short-term Load Forecasting

In neural network based short-term load forecasting, complexity and redundancy of input data have a negative effect on network training efficiency and forecasting precision. Focusing on solving this problem, a multiple method of data processing is developed. Firstly a method called input variable contribution analysis is applied, which divides input variables into primary variables and minor variables according to their contribution to network output. Minor variables are tossed out. Then principal component analysis is applied to primary variables to eliminate linear correlation among them, thus reduce the variable dimension. Based on this method, the main components are gotten, and then simplified network structure is designed. The result shows that after data processing, the training time is reduced noticeably and forecasting precision is enhanced.     

大型商场建筑夏季冷负荷动态预测模型

夏季建筑冷负荷的正确预测是实现大型复杂中央空调优化运行、节能降耗的关键。笔者探讨了商场建筑冷负荷的主要影响因素,确定了建筑动态冷负荷预测模型的输入,提出了夏季基于新风机组供电频率的商场顾客率间接测量方法,解决了商场内顾客量难以检测的难题。还提出了AFC-HCMAC神经网络预测模型算法,实现了大型商场建筑冷负荷的动态预测。仿真结果表明:顾客率在商场冷负荷预测中占有重要地位,在冷负荷预测模型中增加商场顾客率可显著提高预测精度;AFC-HCMAC神经网络预测算法与传统的HCMAC神经网络算法比较,可有效降低神经网络节点数,提高预测精度。     

自适应加权最小二乘支持向量机的空调负荷预测

为了提高建筑空调负荷的预测精度,在分析空调负荷主要影响因素的基础上提出了一种基于自适应加权最小二乘支持向量机(AWLS-SVM)的建筑空调负荷预测方法。该方法根据预测误差的统计特性,采用基于改进正态分布加权规则,自适应地赋予每个建模样本不同的权值,以克服异常样本点对模型性能的影响。建模过程中采用粒子群优化(PSO)算法对模型参数进行优化,以进一步提高模型预测精度。基于DeST模拟数据将AWLS-SVM方法应用于南方地区某办公建筑的逐时空调负荷预测中,并与径向基神经网络(RBFNN)模型、LS-SVM模型及WLS-SVM模型作比较,其平均预测绝对误差分别降低了51.84 %、13.95 %和3.24 %,并进一步基于实际空调负荷数据将该方法应用于另一办公建筑的逐日空调负荷预测中。预测结果表明：AWLS-SVM预测的累积负荷误差为4.56 MW,亦优于其他3类模型,证明了AWLS-SVM具有较高的预测精度和较好的泛化能力,是建筑空调负荷预测的一种有效方法。     

Processing Fault Data Method in Load Forecasting Based on Wavelet Analysis

The reliability and reality of load historical data is the foundation of load forecasting.But,the impact load in running power system,and the disturb data in collecting load data through the SCADA may cause much fault data in load historical data. Focusing on solving this problem, a method through adjusting amplitade of its wavele modulus maxima and processing the wavelet decomposed detail signal by soft threshold based on wavelet analysis and singularity theory, then fault date can be eliminated,so that,the real historical imformation and regulation data can be gained by load forecasting.     

The Application of Fuzzy Theory to Power Load Forecasting

A new method was presented for power load forecasting.Based on the fuzzyclustering technique,the historical samples of power load and its relative environmental factors wereclassified into several typical categories,the fuzzy numbers and sets were then used to describe thepatterns of load variation and the features of the environmental factors for every class. Finally,byunderstanding the state of future environmental factors, the future power load can be predictedthrough determining the category of load variation pottern. The validity of the proposed method wasverified with a practical medium-term load forecasting.     

Application of Attribute Reduct Method in Spatial Load Forecasting

Spatial load forecasting is a process distributing the total forecasted load to all partitioned area, and involving more spatial information and more factor influencing application of the future small area, which need a great deal of memory space and longer operation time. Rough set is new method of data analysis. It need not be provided with any advanced information except data set. But attribute reduct is its main algorithms. Division matrix approach on rough set used to reduce the attribute related to land - use decision in order to remove redundancy attribute and then the rules of small area land - use decision is distilled. The method obtained better effect and enhanced the total load forecasting efficiency.     

中国水稻稻纵卷叶螟区域气象等级预测模型

为了建立水稻不同区域、不同生育期的水稻稻纵卷叶螟预测模型,利用103个植保站2000-2014年稻纵卷叶螟虫情资料与气象资料,采用SPSS软件进行相关分析与主成分分析,在逐步回归的基础上建立水稻稻纵卷叶螟不同区域、不同生育期的发生发展气象等级与迁入气象等级预测模型。结果表明：以华南早稻为例,影响早稻移栽分蘖期发生发展的关键因子包括3月下旬累积降水量、4月上旬平均气温、4月下旬平均气温、5月上旬平均气温以及5月上旬最高气温＞30℃天数;迁飞的关键因子包括4月平均相对湿度、3月下旬平均气温、3月下旬平均相对湿度以及4月上旬平均日照时数。通过2000—2012年数据回代检验发现,不同水稻种植区移栽分蘖期和抽穗开花期发生发展气象等级平均准确率能达到80%以上,迁入气象等级在85%以上。通过2013—2014年外推预报时,发生发展气象等级平均准确率在80%以上,迁入气象等级在78%以上;当预测站点样本数较少时,预报的准确率普遍下降,西南一季稻下降明显。预测模型可从气象角度对中国水稻稻纵卷叶螟发生发展和迁入进行预测。     

灰色人工神经网络在稻瘟病发生预报中的应用

摘要：针对稻瘟病在水稻生长过程中存在的严重危害,笔者基于四川资中地区1998-2008年的稻瘟病发生资料,运用灰色人工神经网络的方法（GBP ）,建立了稻瘟病发生的预报模型,结果表明：灰色人工神经网络模型的平均相对误差为0.0946,远远优于GM(1,1)模型的1.8857。灰色人工神经网络模型可以拟合任意一种函数关系,且该模型信息利用率高,避免了系统数据辨识方法在序列累加时因正负抵消而产生信息失真的现象。灰色人工神经网络模型的拟合和预测精度较高,可以用于该地区稻瘟病发生的预测工作。     

用模糊神经网络提高洪峰预报精度的研究

在大量研究的基础上,提出了基于模糊理论的神经网络改进算法,用来提高对洪峰的预报精度。该方法在网络训练时引入模糊理论来确定网络误差修改的程度。引入的算法增大了大值输出样本和期望输出的误差,使得网络向着提高洪峰拟合精度的方向修改权重。应用表明,改进的模糊BP神经网络能够较好的反映洪水演进机理,提高了神经网络洪水预报模型对洪峰的预报精度,保证了洪峰预报的可靠性。     

Cellular automata spatial temporal data model and forecasting approach

The standard cellular automata(CA) model is expanded to meet requests of space time dynamic simulation and forecast under the platform of geographic information system(GIS). Taking power load forecasting of the electric power industry as the specific application, the relations between dynamic model of the land use and power load space are established. The data and attribute data interactive discrete in spatial temporal data management have been solved. The CA theory is practically used to simulate the process of urban land use dynamic development, to forecast future land use types of each small area, to establish spatial load forecasting model. It breaks through the localization of all kinds of forecasting methods of traditional space time separation power prediction. The effectiveness of the prediction method is verified by example.     

Methods of City Gas Load Forecasting and Their Analysis

Gas load forecast has great influence on the planning, operation and control of gas system and has obvious economic benefit. In this paper the methods of gas load forecasting and their characteristics are systematically introduced. Then, it is pointed out that gas load forecasting is an important manner in modem management of gas system.     

贵德县梨树始花期与气象因子的相关分析及预报模型

为准确预报贵德地区梨树的始花期提供方法支持。笔者利用2007—2014 年青海省贵德县气象局观测的梨树始花期资料和地面气象观测资料,对影响梨树始花期的气象因子进行分析。结果表明：影响梨树始花期的主要气象因子是气温稳定通过5.0℃到3 月底的积温值。利用逐步回归分析方法建立了基于主要气象因子的梨树始花期预报模型,用所建立的预测模型对2007—2014 年梨花始花期进行回测,准确率较高。对2015 年梨树始花期日期进行预测,预测值与实际观测值之间相差1 天,预报值基本吻合。所建立的预测模型能准确预测贵德县梨树始花期的预报。     

东北地区杨树烂皮病气象预报模型研究

为了研究东北地区杨树烂皮病气象预报预警技术,以2002-2008年东北地区杨树烂皮病发生程度为研究对象,利用相应气象资料,在相关分析、逐步回归和逐步判别分析方法的基础上,构建东北地区杨树烂皮病气象适宜度预报模型,并将气象适宜度指数划分为非常适宜、适宜、基本适宜、不适宜4个级别,以反映气象条件对杨树烂皮病发生发展的适宜程度。结果表明：东北地区杨树烂皮病发病面积与冬季气温日较差≥15℃的天数、春季和初夏平均气温呈正相关,冬季气温日较差越大、春季和初夏气温越高,越有利于杨树烂皮病发生发展;入冬冻害、春季温湿条件对东北地区杨树烂皮病发生发展起主导作用,气温、空气湿度与杨树烂皮病发生程度关系最为密切;影响东北地区杨树烂皮病发生发展的关键气象因子分别是3月下旬风速、3月空气相对湿度、4月温湿系数、3月和5月及4月上下旬气温、上年11-12月气温日较差≥15℃的天数。预报模型对2002-2008年的历史拟合和重发生年2009年预报取得了较好的效果,分级和分省外延预报平均准确率均在75%以上,3省预报准确率顺序为：黑龙江>吉林>辽宁。