基于主成分分析的BP神经网络估算藏北高寒草地覆盖度变化

藏北高寒牧区草地是中国高寒草地分布面积最大的地区。为了及时准确地获得该区域草地覆盖度的变化趋势,本研究利用多年气象数据、社会统计数据、GIMMS、MODIS两种归一化植被指数（NDVI）数据作为参数,构建 BP神经网络模型,估算2010—2014年藏北高寒草地年际变化趋势,并用主成分分析方法优化参数来改进模型。结果表明,① BP神经网络模型及其改进模型对藏北高寒草地覆盖度年际变化趋势与遥感值的相关系数为0.16、0.47,表明通过主成分分析优化参数后的BP神经网络模型具有较好的模拟效果。 ②两种BP神经网络估算的植被指数值与NDVI值平均误差率分别为2.36%、2.20%。均有较高的模拟精度。③从神经网络训练步数上看,BP神经网络结果训练收敛步长为5000,基于主成分分析的BP神经网络模型训练收敛步长为454,表明后者提高了计算效率,体现出良好的收敛性。因此,无论从年际变化趋势拟合程度、植被指数估算值精度、还是从计算效率来看,改进的BP神经网络模型对于估算藏北高寒草地覆盖度变化更加行之有效。     

山西日光温室逐日极端气温预测模型研究

利用山西省忻州市日光温室的室内小气候观测数据及气象站资料,用BP神经网络及逐步回归法建立以多种输入变量的不同天气条件下的日光温室内最高温度、最低温度的模型。结果表明,利用BP神经网络及逐步回归法建立的模型R2均在0.96以上,RMSE与AE大部分在2℃之下。利用逐步回归方法在模拟日光温室内晴天最高、最低温度和寡照的最高温度精度较高,利用BP神经网络模型在多云的最高、最低温度与寡照的最低温度模拟的精度较高。选择精度更好的模型对日光温室的极端气温做准确的预测,可为山西省设施农业的管理和调控及小气候预报提供支持。     

The Fuzzy Inputing And Outputing Method in Vibration Fault Diagnosis of Steam Turbine-Generator Set

A fuzzy neural network diagnosis model is established on the basis of the vibration failure features of steam turbine-gernerator set, two kinds of fuzzy inputing method are discussed. At last, the performance of the fuzzy neural network is compared with that of the conventional BP network. The results show that the method presented is suitable for identifying the vibration failure of steam turbine-gernerator set, and it is more efficient in deal with the uncertain data than BP network diagnosis.     

Nonlinear Forecast System for River Water Pollution Based on GIS

Based on the analysis of the water pollution spatial distribution characters of Yangtze River in Chongqing,a new method based on the integration of BP neural network and genetic arithmetic(GA) is proposed.For some shortcomings existed in the standard BP neural network,this method has ultimately overcome these shortcomings by combining the GA with BP artificial neural network through altering stimulating function,adding momentum factor to power value for BP algorithm and introducing genetic arithmetic to searching for the knots of the hidden layer,momentum factor and learning level.Using this method can easily overcome the difficulty of measuring the water prediction model's parameters.GIS is used as a tool for data management and spatial analysis,and the prediction result of the model for the water pollution spatial distribution characters of Yangtze River in Chongqing is visualized and explored with the precision of more than 78%.     

离差最大化结合BP神经网络评价烟叶化学品质

为解决烟叶化学品质现有组合评价方法的不足，探讨离差最大化组合评价法与BP神经网络相结合进行评价。选取4种典型单一评价法，首先采用改进熵权法和AHP法确定指标权重，然后根据离差最大化原理计算组合评价值，最后利用BP神经网络对组合评价值进行反演。结果表明，离差最大化组合评价值与单一评价法相关性较其他组合评价法更高，平均相关系数为0.9822;BP神经网络对组合评价值有较高的预测准确性与稳定性，预测值与实际值相对误差不超过3%，决定系数大于0.9900。说明，离差最大化组合评价法对单一评价法的组合效果更好，BP神经网络提高了组合评价的便捷性。     

A new pattern recognition method of gas sensor array detection

BP neural network based gas sensor array detection pattern recognition has some disadvantages, such as slow convergence and local minimum problem. A modified immune neural network model which combines BP algorithm and immune algorithm is proposed to enhance global search capability and improve the performance of the neural network model. Orthogonal test is adopted to design the study samples of neural network. This ensures the accuracy of neural network while reducing the number of samples. The simulation results show that the proposed pattern recognition method solves the cross sensitivity of gas sensor effectively, overcomes the disadvantages of traditional BP neural network and improves the learning speed and detection accuracy.     

雄先型核桃雄花疏除的二次回归与BP神经网络模型研究

雄先型核桃雄花疏除（去雄）是提高产量的重要管理措施,为提高核桃去雄的效率,建立二次回归与BP神经网络模型。分别以乙烯利、赤霉素和甲哌鎓为自变量和核桃雄花脱落率为响应指标,进行田间建模试验,建立了二次多项式回归方程和BP神经网络模型,并于翌年进行BP模型田间确认试验。试验数据分为训练集、确认集和试验集,中心组合（二次旋转回归试验设计）田间建模试验得到的20组数据随机划为训练集(17)和确认集(3)数据,试验集为翌年田间确认试验得到的数据,BP神经网络的拓扑结构为3-5-1。（1）BP神经网络对确认集样本的预测值误差分别为1.3550%、0.4291%、0.3538%;（2）BP神经网络的预测值与田间确认试验结果相差为2.04%,回归预测值与田间确认试验结果相差为3.12%;（3）BP神经网络预测比回归预测提高预测精度1.0%以上。将二次多项式逐步回归分析和BP神经网络方法有效的结合使用,既可明确各因子的作用效应亦可得到相对准确的预测结果。     

基于逐步回归与BP神经网络的日光温室温湿度预测模型对比分析

为构建较准确的日光温室温湿度预测模型,于2011-2013年冬季(1月、2月、12月)天津市宝坻区开展温室内外环境监测试验,并建立3种天气类型(晴、多云、阴)下3个时段(0-8时、8-17时、17-23时)逐步回归与BP神经网络温室内温湿度预测模型。结果表明:1)温室内气温逐步回归模型9种情况下模拟值与实际值的绝对误差小于3℃的平均准确率Rate(≤3℃)为88%,平均均方根误差(RMSE)为2℃;BP神经网络模型9种情况下模拟值与实际值的绝对误差小于3℃的平均准确率Rate(≤3℃)为94%,平均均方根误差(RMSE)为1.6℃。应用BP神经网络建立的气温预测模型相对更为准确稳定。2) 相对湿度逐步回归模型9种情况下模拟值与实际值的绝对误差小于6%的平均准确率Rate(≤6%)为81%,平均均方根误差(RMSE)为5.7%;BP神经网络模型9种情况下模拟值与实际值的绝对误差小于6%的平均准确率Rate(≤6%)为80%,平均均方根误差(RMSE)为6.7%。两类模型均不适宜预测8-17时日光温室相对湿度,而17-23时与0-8时应用逐步回归建立的湿度预测模型相对更准确稳定。     

基于随机森林法的棉花叶片叶绿素含量估算

为了高效和无损地估算棉花叶片的叶绿素含量,本研究测定了棉花光谱反射率及叶绿素含量(soilandplant analyzerdevelopment,SPAD)值,对光谱数据进行包络线去除处理、立方根转换和倒数转换,以SPAD值与反射光谱之间的相关性为基础,通过随机森林法筛选出对棉花叶片SPAD值影响较大的特征波段,构建估算棉花叶片SPAD值的BP神经网络(back propagation artificial neural networks, BP ANN)、偏最小二乘回归(partial least squares regression,PLSR)两个模型。结果表明,在605～690nm范围内的反射率与SPAD值相关性达0.01显著水平,均呈负相关,相关系数最高值为-0.619。与原始光谱相比,经过变换后的棉花反射率与SPAD值相关性结果相差较大,其中去除包络线光谱在550～750 nm波段范围有效提高了相关性,相关性效果优于倒数转换数据和立方根转换数据。随机森林法能够有效评出对SPAD值影响较大的特征波段,进而提高模型估算精度。在两种模型中,基于去除包络线光谱建立的PLSR和BP神经网络模型的决定系数R~2分别为0.92、0.83,说明这两种模型的估算能力较好;两种模型RMSE分别为0.88、1.26, RE分别为1.30%、1.89%,表明PLSR模型的估算精度比BP神经网络模型高。从模型的验证效果来看,PLSR模型在估算棉花SPAD值方面有一定的优势和参考价值。     

基于数值预报产品的地表温度预报方法研究

为了开展地表温度预报业务,提高逐日地表温度预报准确率,利用2007—2012年的ECMWF和T213数值预报产品资料及抚顺市的逐日地表温度资料,采用逐步回归分析方法和BP神经网络模型分别构建抚顺市地表温度预报模型,并对模型的精度进行检验。结果表明,地表温度与ECMWF的高度场、海平面气压场、温度场和T213的散度场、高度场、海平面气压场、地面气压场、海平面K指数、水汽通量、相对湿度、温度场、地面气温和场涡度场均呈显著相关。对预报模型进行精度检验显示,地表平均温度和地表最低温度的预报效果较好,≤3℃预报准确率均达到79%以上。2种模型对比显示,BP神经网络预报模型总体上优于逐步回归预报模型;逐步回归预报模型较BP神经网络预报模型稳定。     

Method to Predict the Coke Rate Based on BP Neural Network

Coke rate is a very important technique index in the processing of metallurgical, and it is also an important goal that should be reached and controlled in practice.The blast furnace is a countercurrent heat and mass exchange reactor involving the solid, liquid and gaseous phases. Using computer encoded mathematical and statistical methods can not get the precise result. An improved 9-9-1 BP(Back propagation) neural network was trained and used in the prediction of the coke rate. The result indicates that the BP nets can predict coke rate accurately and the error between prediction and real coke rate less than 2%. And the use of a hybrid model in actual on-line intelligence control was also discussed.     

Analysis on the Control of Variable Speed Wind Turbine Using Neural Network Controller

It is very difficult to build the accurate mathematical model of the wind turbine generator system because of the uncertainty of air kinetics and the complexity of power electronics, especially when the wind speed changes abruptly or there is a disturbance. But the classical control needs the model. Using neural network controller to the wind turbine generator system can overcome these difficulties. The wind speed can be followed and the maximum power can be obtained under low wind speed by using the power coefficient curve BP neural network and the optimum pitch angle BP neural network. The maximum power can be kept and under the allowed range in the condition of high wind speed. The simulation model and result are given under the environment of MATLAB. The fluctuation of wind speed can be controlled and the disturbance can be cancelled by BP neural network controller.     

基于BP神经网络研究儿茶素在ADS-8树脂固定床中的洗脱过程

摘要：以信阳毛尖茶叶浸提液为原料,研究ADS-8树脂固定床吸附儿茶素后的洗脱过程,建立模型并优化工艺。基于BP神经网络建立洗脱模型,利用模型对因素进行仿真分析。模型误差为0.00108523,测试样本的试验值和模拟值的相关系数r=0.984,最佳工艺条件是温度20℃、流速1.0mL/min、乙醇浓度30%。基于BP神经网络建立的模型具有很强的逼近能力,为儿茶素在ADS-8树脂固定床中洗脱过程的预测、控制提供一定参考。     

THE NEURAL NETWORK CONROL TECHNOLOGY USED IN HYDRAULIC LOGIC VALVE SYSTEM

This paper is concerned with artifical neural network used in hydrauic logic valve control technique, the characters and structure of hydraulic logic valve, the main characters and topological construction of artifical neural network are analyzed in detail. The pre-feedback used in logic valve. is focused, the training method was derived and improved. Using this improved BP training method, the net work was trained and satisfactory experiment results were gained.     

Improvement and Application on Back Propagation Network Based on Partial Least-squares Algorithm

This paper proposes a novel BP network model based on nonlinear iterative partial least-squares algorithm which can fit nonlinear data. The novel BP network model can reduce iterative step number and advance learning effieieney. This paper pretreats data by nonlinear iterative partial least-squares algorithms. The weights initialization of input floor and output floor are set by applying the loading weights of dependent variable and cause variable, the member of hidden nodes are set by applying factor numbers of nonlinear iterative partial leastsquares algorithm, the connection co- efficient is set by applying the connection matrix B. Performances of the BP, PLS, and PLS-BP are analyzed and compared. The results show that the PLS-BP has better fitting and forecasting than BP and PLS.     

基于LM-BP神经网络的耕地土壤养分等级划分模型──以皖南山区为例

研究旨在通过BP神经网络方法,构建起LM-BP网络结构(5-M-1)模型,达到对土壤养分等级划分的目的,为合理的土壤养分管理提供可靠依据。采用Levenberg-Marquardt (LM)训练算法,构建3层网络模型：一个输入层、一个隐含层、一个输出层,利用3层网络作为耕地土壤养分等级划分模型。利用土壤养分各级评价标准作为模型的训练样本和测试样本,以此来对BP神经网络进行训练和测试,并对歙县土壤养分进行综合评价。结果表明：LM-BP网络结构对测试样本输出的预测值和实际参考值是一致的。最终通过灰色关联模型和主成分分析方法对歙县土壤养分的综合评价结果与BP神经网络的模拟结果相对比,发现也是基本一致的。LM-BP网络结构应用于土壤养分等级划分中,得到了很好的预测效果,为智能算法应用于农业领域奠定了良好的基础。     

西藏林区森林火险等级中短期预报方法

利用2009/2010、2010/2011和2011/2012西藏林区防火期（11月—翌年4月）气象观测资料和T639数值预报资料,基于人工神经网络BP算法,建立了西藏林区森林火险等级1~7天预报模型,历史拟合率超过85%;通过对2012/2013防火期间的森林火险等级试报检验结果表明,前3天的平均绝对误差不超过0.5级,7天的平均绝对误差不超过0.6级;与直接利用数值预报模式气象要素预报结果相比,有效地纠正了数值模式要素预报的系统偏差,表明模型预报效果良好。该模型的建立提高了对西藏高原森林火险等级的预报准确性,为森林火险防御和消防调度提供了参考。     

Module Design of Refrigerant Themodynamic Properties Calculation for Simulation of Refrigeration and Air Condition System

This paper constructs a model of refrigerant thermodynamic properties calculation for simulation of refrigeration and air condition system based on the parametric equations of state of refrigerant. The Visual C++ 6.0 language is used to develop this model, which has a friendly user interface (UI). This program is easy to learn and use and it has reliably function. Compared with the data provided by the standard graphs and tables of refrigerant thermodynamic properties, the calculation results of this model have certain precision and meet the expectative requirements.     

Assembly fault rate analysis using grey relation and least squares support vector machines

To get the relationship between assembly fault rate and its attributes, least squares support vector machine (LSSVM)is introduced to quantitatively study assembly fault rate. Aiming at the drawbacks of assembly reliability evaluation method(AREM), the attributes of assembly-fault-rate-affecting 5M1E(Man, Machine, Material, Method, Measurement and Environment) factors obtained by AREM are improved, hence the LSSVM model with all attributes is established. To reduce the time of calculating the assembly fault rate and provide the priority for assembly reliability improvement, grey relation analysis is applied to extracting the main attributes, at the same time genetic algorithm(GA)is used for parameter optimization in LSSVM. The assembly fault rate analysis results show that the method using grey relation analysis and least square support vector machine is simpler and more accurate compared with other methods such as LSSVM model using all attributes and BP neural network using main attributes.     

基于无人机高光谱影像的玉米地上生物量反演

【目的】研究以玉米地上干生物量为研究对象,探讨基于无人机高光谱数据利用人工神经网络法反演生物量的可行性。【方法】在吉林省蔡家镇开展玉米氮肥梯度试验,并进行无人机高光谱数据和地上干生物量获取,共获数据30组。随机选22组数据用于建模,剩下8组用于模型的外部验证。分别基于光谱指数法和BP神经网络算法构建反演模型,比较分析各种方法反演玉米生物量的优劣。【结果】结果表明：和基于光谱指数构建的生物量反演模型相比,BP神经网络模型取得了更好的反演结果。其建模时决定系数为0.99均方根误差为0.08 t/ha,相对均方根误差为3.39%;外部验证时,决定系数为0.99,均方根误差为0.15 t/ha,相对均方根误差为8.56%。【结论】BP神经网络模型可有效提高无人机高光谱遥感反演玉米地上生物量的精度。