共查询到20条相似文献，搜索用时 15 毫秒
1.
Quantification of Odours from Piggery Effluent Ponds using an Electronic Nose and an Artificial Neural Network 总被引：1，自引：0，他引：1
An Artificial Neural Network (ANN) and an electronic nose, AromaScan, were used to predict the piggery odour concentrations emanating from an effluent pond and to develop a confident, rapid, and cost-effective technique for odour measurement. Odour samples from five different piggery effluent ponds were analysed using the AromaScan and dynamic dilution olfactometry. The resulting sensor data were used to train the artificial neural network to correlate the responses to the odour concentrations measured by olfactometry. Effectiveness was evaluated through simulation with various pre-processing techniques and network architectures. The simulation results have shown that a two-layer back-propagation neural network, which has a tan-sigmoid transfer function in the hidden layer and a linear transfer function in the output layer, could be trained to predict piggery odour concentrations with high value of the correlation coefficient R of 0·984 under the best network performance. The results from the application of scaling and principal component analysis suggest that these techniques are necessary not only to avoid the failure of the network caused by saturation but also to enhance performance. An early stopping technique was shown to provide benefits to the network performance in terms of a decrease in computation time and overfitting. It was found that the optimal number of hidden neurons for the network was 20. Odour concentration of unknown samples were able to be predicted with significant accuracy. 相似文献
2.
3.
In this study, an artificial neural networks (ANN) model was developed to predict the removal of a polycyclic aromatic hydrocarbon (PAH), namely, naphthalene from marine oily wastewater by using UV irradiation. The removal rate was used as model output and simulated as a function of five independent input variables, including fluence rate, salinity, temperature, initial concentration and reaction time. The configuration of the ANN model was optimized as a three-layer feed-forward Levenberg–Marquardt backpropagation network with log-sigmoid and linear transfer functions at the hidden (12 hidden neurons) and output layers, respectively. By considering goodness-of-fit and cross validated predictability, the ANN model was trained to provide good overall agreement with experimental results with a slope of 0.97 and a correlation of determination (R 2) of 0.943. Sensitivity analysis revealed that fluence rate and temperature were the most influential variables, followed by reaction time, salinity and initial concentration. The findings of this study showed that neural network modeling could effectively predict the behavior of the photo-induced PAH degradation process. 相似文献
4.
基于BP神经网络的陕北黄土高原侵蚀产沙影响因子显著性研究 总被引：1，自引：0，他引：1
黄土高原地区的侵蚀产沙与影响因子的关系特征一直是研究热点。以陕北黄土高原23个小流域为实验样区,采用BP神经网络方法,将6个影响侵蚀产沙的因子作为输入变量,侵蚀产沙模数作为输出变量,通过输入变量与隐含层之间、隐含层与输出变量之间的权重矩阵关系,构建关系模型。实验结果显示,该方法可以有效地区分不同因子对侵蚀输沙模数影响的显著性;6个因子对侵蚀产沙影响的显著性由高到低依次为:岩土抗蚀性>蚕食度>沟谷密度>年均降雨量>NDVI指数>粉砂黏土含量。最后,随机选择3个小流域作为检验样本,采用BP神经网络进行预测,验证了该模型的有效性。该研究可望完善小流域侵蚀产沙分析方法。 相似文献
5.
海洋微生物酶反应器智能控制系统的研制 总被引：3，自引：1，他引：2
生物发酵过程具有严重非线性、高度时变性、高阶多变量和大型不确定的特点,为了实现生物发酵过程的自动化控制,提高生物技术产品的生产水平,使得发酵过程的参数检测、操作监视、自动控制等智能化,分析了影响发酵过程的主要因素以及各变量之间的耦合关系,综合应用传统的PID控制方法、模糊神经网络技术,构建了一种多变量模糊神经控制系统的前馈解耦算法并将其应用在发酵过程的思想,同时采用了冷凝、回收、利用发酵尾气技术,解决了尾气排放、罐体泄漏染菌的问题。模糊神经控制器和解耦部分独立设计,在模糊控制器中引入神经网络,解耦网络采用一层隐层,利用简化的学习算法,根据系统输出误差,在线调整网络权值,从而实现动态解耦而无需辨识被控对象的模型。该方法结构简单且计算量小,经实际应用结果表明这种控制算法具有很好的控制效果。 相似文献
6.
C Zondervan S Muresan H G de Jonge E U van Velzen C Wilkinson H H Nijhuis T Leguijt 《Journal of agricultural and food chemistry》1999,47(11):4746-4749
An electronic nose has been used to classify blockmilk products subjected to various heating processes based on their volatile composition. Multivariate analyses of electronic nose and GC/MS data are highly comparable with respect to relative changes in aroma profile going from raw to final product. Predictive properties of various neural networks based on the raw sensor output were moderate to good. 相似文献
7.
PFC软件作为一款成熟的离散元分析软件,由于在处理连续与非连续介质方面的出色表现,得到了广泛的应用。但是PFC软件所需要的细观参数均需要采用室内试验数据通过试错法反复调试才能获得,效率低、盲目性高,严重影响后续试验数据,因此急需一种新的细观参数校准方法。本文以玉米秸秆颗粒的单轴蠕变试验为基础,结合离散元软件PFC 2D,通过正交试验多因素方差分析方法分析了Burgers模型宏细观参数之间的影响关系,从而证明宏细观参数之间存在着复杂关系,不宜采用通过回归分析获得宏细观参数之间的关系式的方式标定细观参数,适合利用BP神经网络进行参数标定,利用创建的BP神经网络对细观参数进行标定,根据测试组的标定结果分析得出Burgers模型各细观参数的标定精度均在92%以上,且误差较为稳定,而且训练好的神经网络相关系数R>0.96,从而证明BP神经网络的细观参数标定性能较为可靠。将玉米秸秆单轴蠕变试验的宏观参数带入训练好的BP神经网络中进行细观参数标定,比对模拟蠕变试验与物理蠕变试验发现,两者的蠕变曲线基本一致,应变量的最大误差为2%,证明了BP神经网络具有良好的参数标定能力,可为PFC参数标定提供一定的参考价值。 相似文献
8.
遗传RBF神经网络在卷烟香气质量评定中的应用 总被引：4，自引：1，他引：4
卷烟香气质量感官评定结果的准确性往往难以保证，因此研究准确、客观的评定方法是必要的。运用遗传RBF神经网络研究了基于气敏传感器阵列的卷烟香气质量评定方法。实例表明，该方法是可行的，所给出的遗传学习算法是有效的。为进一步开展卷烟香气质量客观评定方法的研究奠定了基础。 相似文献
9.
为指导节水灌溉策略的制定,利用基于多值神经元的复数神经网络(multilayer neural network with multi-valued neurons,MLMVN)方法,建立了土壤墒情多步预测模型。首先,利用均值法替换样本中的异常值并对缺失值进行补充,并由数据分析知土壤墒情数据为非平稳的非线性时间序列。然后,根据土壤墒情与环境因素(降雨量、气温和风速)的相关性分析结果选择降雨量为关键环境因素。最后将土壤墒情、降雨量及目标土壤墒情复数化,作为网络输入和期望输出建立MLMVN预测模型。结果表明,网络结构为240-15-1200-1时单步预测精度为0.883,采用循环预测法进行步长为72的多步预测,平均预测精度为0.853,比实数域误差反向传播神经网络BP提高了9.1%。研究表明,MLMVN模型多步预测误差累计小,预测结果可作为该地区节水灌溉策略制定的理论依据。 相似文献
10.
基于监测数据和BP神经网络的食品安全预警模型 总被引：10，自引：3，他引：7
以中国实际食品安全监测数据为样本,研究基于BP神经网络的食品安全预警方法。首先对食品安全日常监测数据进行筛选简化,选择其中与食品安全最为密切的167种检测项目,以此检测项目为指标按月度划分建立数据样本。然后建立以167种检测项为输入层,包含2个隐层,以化学污染、农药残留、兽药残留、重金属、微生物致病菌5大类为输出层的食品安全预警神经网络模型,最后用所得数据样本进行训练和验证。结果表明,基于BP神经网络的食品安全预警方法能有效识别、记忆食品危险特征,能够对输入样本进行有效的预测,研究有助于丰富食品安全数据的处理方法,有助于完善相关预警技术手段。 相似文献
11.
将人工神经网络技术应用于鲜茶叶的分类,茶叶图像面积、周长、凸壳面积、凸壳周长、等二阶距椭圆长轴长度、短轴长度、椭圆偏心率等几何参数和R、G、B三个彩色空间分量的均值、标准偏差、平滑度和一致性等纹理参数可以作为茶叶分类的特征值。试验表明,BP网络用于茶叶分类能够取得较好的效果,分类判断的正确率达到90%。网络的隐藏层和输出层为多个神经元时,其可能达到的分类效果要略好于隐藏层和输出层只有单个神经元的网络,但前者训练出的网络会出现权值不能收敛到全局误差最小值的情况,其可靠性不如后者。 相似文献
12.
13.
14.
Classifications of fish production methods, wild or farm-raised salmon, by elemental profiles or C and N stable isotope ratios combined with various modeling approaches were determined. Elemental analysis (As, Ba, Be, Ca, Co, Cd, Cr, Cu, Fe, K, Mg, Mn, Na, Ni, P, Pb, Sr, Ti, and Zn) of wild and farm-raised salmon samples was performed using an inductively coupled plasma atomic emission spectroscopy. Isotopic and compositional analyses of carbon and nitrogen were performed using mass spectrometry as an alternative fingerprinting technique. Each salmon (king salmon, Oncorhynchus tshawytscha ; coho salmon, Oncorhynchus kisutch ; Atlantic salmon, Salmo salar ) was analyzed from two food production practices, wild and farm raised. Principal component analysis (PCA) and canonical discriminant analysis (CDA) were used for data exploration and visualization. Five classification modeling approaches were investigated: linear discriminate function, quadratic discriminant function, neural network, probabilistic neural network, and neural network bagging. Methods for evaluating model reliability included four strategies: resubstitution, cross-validation, and two very different test set scenarios. Generally speaking, the models performed well, with the percentage of samples classified correctly depending on the particular choice of model and evaluation method used. 相似文献
15.
风电机组关键位置载荷预测对风电机组安全、经济运行具有重要意义。通过建立SCADA数据与载荷间的近似关系对风电机组关键位置载荷进行预测。采用BP神经网络建立SCADA数据和载荷的关系模型,利用SCADA数据与载荷间的相关性来筛选模型输入参量,采用试错法确定BP神经网络的层数与神经元数量。针对某2.5 MW风电机组的7处关键位置进行了载荷实测。研究表明,在不采用风速作为输入参量的情况下,模型的预测结果与实测结果具有良好的一致性,相对误差的均值在1.28%到15.6%之间,决定系数R2在0.951到0.882之间;与试错法选择输入参量相比,基于相关性计算的输入参量选择方法能够更高效地筛选出更多恰当的SCADA参量,从而进一步提高预测准确度。因此,基于BP神经网络建立SCADA数据与载荷的近似关系可作为风电机组关键位置载荷预测评估的有效手段。 相似文献
16.
基于BP神经网络模型的荔枝树叶面积测定方法 总被引：7，自引：1，他引：7
为了准确、快速地测定荔枝树叶面积，设计了一个BP神经网络模型，输入参数为叶片长度和叶片最大宽度，输出参数为叶面积。用LI-3000A型叶面积仪测量所得到的样本数据对网络进行训练，测试样本的网络输出与网络目标的相关系数达0.99609，网络模型是有效的。用训练后的网络模型对10组未参加建模的样本数据进行叶面积测定，误差平方和为1.2929，优于回归方程法的2.511。训练好的BP神经网络模型可以在不破坏叶片的情况下，简单、快速、经济地测定大量的荔枝树叶片面积。 相似文献
17.
M. Amini K. C. Abbaspour H. Khademi N. Fathianpour M. Afyuni & R. Schulin 《European Journal of Soil Science》2005,56(4):551-559
Design and analysis of land‐use management scenarios requires detailed soil data. When such data are needed on a large scale, pedotransfer functions (PTFs) could be used to estimate different soil properties. Because existing regression‐based PTFs for estimating cation exchange capacity (CEC) do not, in general, apply well to arid areas, this study was conducted (i) to evaluate the existing models and (ii) to develop neural network‐based PTFs for predicting CEC in Aridisols of Isfahan in central Iran. As most researches have found a significant correlation between CEC and soil organic matter content (OM) and clay content, we also used these two variables for modelling of CEC. We tested several published PTFs and developed two neural network algorithms using multilayer perceptron and general regression neural networks based on a set of 170 soil samples. The data set was divided into two subsets for calibration and testing of the models. In general, the neural network‐based models provided more reliable predictions than the regression‐based PTFs. 相似文献
18.
19.
针对作物生产碳排放预测较为困难的实际问题,提出基于BP神经网络算法的玉米生产碳排放预测模型。选择地处河西走廊石羊河下游的民勤绿洲246家农户,面对面调查玉米种植户农场内生产投入数据,将玉米生产投入数据作为神经网络输入层;查阅和梳理国内外相似区域玉米生产环节碳排放系数,运用碳足迹生命周期法计算得到的碳排放值作为神经网络输出层;基于BP人工神经网络算法,运用试凑法确定网络隐含层节点个数,建立河西绿洲玉米生产碳排放预测模型,选择多元线性回归模型、多元非线性回归模型,对该模型有效性进行评估。研究结果表明,3层且各层节点数9、10、1的神经网络结构能够准确预测河西绿洲玉米生产碳排放,其碳排放预测值为0.763 kg(CO_2-eq)·kg~(-1)(DM);9-10-1结构的神经网络预测模型的相关系数(R~2=0.984 7)高于多元线性和非线性回归模型,该神经网络结构模型的均方根误差(RMSE=0.069 1)、平均绝对误差(MAE=0.051 3)均低于其他模型,BP神经网络算法预测性能明显优于其他预测模型。该研究为准确预测农业生产碳排放提供了新思路和可操作方法。 相似文献
20.
Arvisenet G Le Bail P Voilley A Cayot N 《Journal of agricultural and food chemistry》2002,50(24):7088-7093
In food matrices, where starch is often used as a gelling or texturing agent, the occurrence of amylose-aroma complexes and their effect on the release of aroma compounds are difficult to determine. Indeed, thick or gelled systems are known to reduce the diffusion rate of flavor molecules, resulting in an increase of retention. Moreover, interactions between aroma compounds and matrix components might increase the retention of aroma compounds. The complexing behavior of three aroma compounds with amylose was studied by DSC and X-ray diffraction to determine the relative importance of these two factors. Their interaction properties were different: two of them formed complexes, and the third did not. These aroma compounds were added in food matrices containing different starches that induced different textures. Their retention was studied by static headspace analysis. The retention of aroma compounds appeared to depend on the amylose/amylopectin ratio of starch, both from the formation of complexes and by a viscosity effect. 相似文献