多气体的SVM数据融合定性识别方法

针对基于神经网络的多气体定性识别方法中存在的过学习和泛化能力差的问题,提出了一种基于支持向量机(SVM)与多传感器数据融合的多气体定性识别方法。该方法采用结构化风险最小化准则的多类分类支持向量机对由多个气体传感器、温度和湿度传感器组成的传感阵列的数据进行融合,克服了传统方法的缺陷,消除了环境温度与湿度等因素的影响,实现了100%的定性识别率,实验结果证明了该方法的有效性。     

气体传感器阵列温度漂移抑制的研究

半导体气体传感器存在漂移问题,温度变化对漂移的影响尤为明显.在气体传感器阵列中,可以加入温度、湿度等传感器,监测其工作环境.实验系统采用恒温箱设定一组温度,制备气体样本20例(两种浓度样本各10例),采集传感器对样本的响应;通过人工神经网络来识别样本;当有误判发生时,在原网络中引入温度传感器的响应值,消除了误判,在一定程度上抑制了漂移,改善了网络性能,验证了该温度漂移抑制方法的可行性.     

用气体传感器特性数据评价酒精和醋酸的干扰程度

针对我国烹调中大量使用料酒和醋造成报警器误报的问题,以半导体传感器为例,论述了使用EXCEL一对气体传感器的气体浓度特性数据进行处理,利用其特性趋势线方程公式可以计算出使用该传感器制作成的报警器受酒精或醋酸等的干扰程度.     

基于L-M算法的反向传播网络的湿度传感器输出误差补偿研究

针对湿度传感器的输出非线性问题,提出了基于L-M算法建立BP神经网络进行补偿校正,实现电阻型湿度传感器的输入与输出非线性补偿,并与共轭梯度算法、拟牛顿算法所建立的神经网路模型进行对比,重点比较了模型迭代性能、标准偏差;最后发现当神经网络用L-M算法进行训练模拟时在迭代性能、标准偏差等方面具有更优异的表现,更适合湿度传感器的非线性特性的补偿校正。     

基于多传感器数据融合的智能小车避障的研究

针对智能小车避障问题,提出了一种将模糊逻辑和神经网络相结合的融合方法—Takagi-Sugeno(T-S)模糊神经网络方法。基于此方法的数据融合算法应用在智能小车避障运动中,采用多只超声波传感器和红外线传感器探测障碍物的距离和方向,采集的各种数据利用T-S模糊神经网络进行融合。通过实验仿真表明:此方法能够使智能小车对障碍物的灵活避障和导航行进。     

基于PSO-BP神经网络的湿度传感器温度补偿

针对自动气象站采用的HMP45D型温湿一体化传感器在实际应用过程中易受温度影响的问题,提出了基于粒子群优化算法(PSO)的BP神经网络温度补偿模型,利用粒子群优化算法对BP神经网络的初始权值阈值进行全局寻优,将粒子群优化算法优化好的权值阈值赋给BP神经网络,对BP神经网络进行训练。根据不同温度条件下测得的多组湿度传感器数据,通过建立模型,实现温度补偿,与传统BP神经网络补偿结果进行比较。实验表明,与传统BP神经网络模型相比,利用PSO-BP神经网络模型进行温度补偿后所得的误差绝对值之和降低了10.3887%RH,PSO-BP神经网络可以克服传统BP神经网络易陷入局部极值的局限,补偿精度更高,能更加有效地补偿温度对湿度传感器的影响。     

A time-efficient convolutional neural network model in human activity recognition

Activity recognition represents the task of classifying data derived from different sensor types into one of predefined activity classes. The most popular and beneficial sensors in the area of action recognition are inertial sensors such as accelerometer and gyroscope. Convolutional neural network (CNN) as one of the best deep learning methods has recently attracted much attention to the problem of activity recognition, where 1D kernels capture local dependency over time in a series of observations measured at inertial sensors (3-axis accelerometers and gyroscopes) while in 2D kernels apart from time dependency, dependency between signals from different axes of same sensor and also over different sensors will be considered. Most convolutional neural networks used for recognition task are built using convolution and pooling layers followed by a few number of fully connected layers but large and deep neural networks have high computational costs. In this paper, we propose a new architecture that consists solely of convolutional layers and find that with removing the pooling layers and instead adding strides to convolution layers, the computational time will decrease notably while the model performance will not change or in some cases will even improve. Also both 1D and 2D convolutional neural networks with and without pooling layer will be investigated and their performance will be compared with each other and also with some other hand-crafted feature based methods. The third point that will be discussed in this paper is the impact of applying fast fourier transform (FFT) to inputs before training learning algorithm. It will be shown that this preprocessing will enhance the model performance. Experiments on benchmark datasets demonstrate the high performance of proposed 2D CNN model with no pooling layers.

     

Models for Prediction of Physiological Responses of Holstein Dairy Cows

The goal of the present study was to evaluate techniques for modeling the physiological responses, rectal temperature, and respiratory rate of black and white Holstein dairy cows. Data from the literature (792 data points) and obtained experimentally (5884 data points) were used to fit and validate the models. Each datum included dry bulb air temperature, relative humidity, rectal temperature, and respiratory rate. Two models based on artificial intelligence—artificial neural networks and neurofuzzy networks—and one based on regression were evaluated for each response variable. The adjusted models predict rectal temperature and respiratory rate as a function of dry-bulb air temperature and relative humidity. These models were compared using statistical indices. The model based on artificial neural networks showed the best performance, followed by the models based on neurofuzzy networks and regression; the last two performed similarly.     

Data-driven soft sensor development based on deep learning technique

In industrial process control, some product qualities and key variables are always difficult to measure online due to technical or economic limitations. As an effective solution, data-driven soft sensors provide stable and reliable online estimation of these variables based on historical measurements of easy-to-measure process variables. Deep learning, as a novel training strategy for deep neural networks, has recently become a popular data-driven approach in the area of machine learning. In the present study, the deep learning technique is employed to build soft sensors and applied to an industrial case to estimate the heavy diesel 95% cut point of a crude distillation unit (CDU). The comparison of modeling results demonstrates that the deep learning technique is especially suitable for soft sensor modeling because of the following advantages over traditional methods. First, with a complex multi-layer structure, the deep neural network is able to contain richer information and yield improved representation ability compared with traditional data-driven models. Second, deep neural networks are established as latent variable models that help to describe highly correlated process variables. Third, the deep learning is semi-supervised so that all available process data can be utilized. Fourth, the deep learning technique is particularly efficient dealing with massive data in practice.     

特种装备内部环境气氛传感系统设计与实现

为长期监测特种装备内部环境气氛的变化规律,研制了一种基于STM32微处理器的嵌入式气氛传感系统.核心敏感元件选用SHT15数字式温湿度传感器、LOX02数字式氧气传感器和自主研制的MEMS氢气传感器.设计了一种独特的低功耗供电方案,有效降低了系统的休眠功耗.测试结果表明:系统能够准确测量环境的温度、湿度、气体压力、氧气浓度和氢气浓度等重要指标,体积小、功耗低,对特种装备可靠性评估具有重要意义.     

多波段微波辐射计反演大气温湿廓线性能分析

为了为星载、机载以及地基微波大气温湿廓线探测仪通道的设置、大气参数反演指标的论证、反演算法的开发以及反演产品的质量评定提供参考依据,基于快速辐射传输模式(RTTOV10)和大气参数廓线库,建立了基于神经网络的微波大气温湿廓线反演性能分析方法,分析了反演方法、通道选择、亮温观测误差和地表比辐射率等因素对大气温湿廓线反演性能的影响。模拟试验分析表明:1神经网络反演算法显著优于线性统计回归反演算法,特别是对亮温观测噪声的敏感性相对较弱;2183.31GHz附近的水汽探测通道能够为大气温度廓线反演提供一定的信息;118.75GHz附近的温度探测通道对整个大气的温度反演均有明显影响,在200hPa附近误差的影响量达0.4K;350~60GHz和118.75GHz附近的温度探测通道对基于183.31GHz附近通道的湿度廓线反演具有重要影响,而且存在一定的互补性;4微波亮温观测误差以及地表比辐射率假定对大气温湿廓线反演有着显著影响。     

基于硫化物系的湿敏元件的研究

研究了Zn、Cd、Sb等系列硫化物薄膜湿敏元件的感湿特性 ,发现金属硫化物元件具有较好的湿敏特性和稳定性 ,对元件的频率特性、湿滞特性、响应 -恢复时间、阻抗、以及稳定性进行了测试 ,并对其敏感机理进行了探讨     

基于神经网络的不可靠数据恢复研究

智能系统中单点或少数传感器采集的数据在某一段时间出现不可靠问题,在装备有许多传感器的智能系统中普遍存在,即使在由先进的传感器构成的桥梁结构健康监测系统中,80%以上的虚假报警也是由于测量数据的不可靠性造成的.传统上对于不可靠数据的处理主要应用线性回归法、平均法等方法进行恢复,然而,大多数测量数据在时域上表现为非线性特征,传统方法恢复的数据在精度上是很难达到要求.以桥梁挠度数据作为研究对象,利用原始数据对挠度测量点进行了关联分析,并依据RBF神经网络强大的函数逼近能力,提出了一种基于神经网络模型来恢复不可靠测量数据的方法,并在仿真实验中,通过对比实验(该方法的均方误差为2E-9,线性回归法均方误差为0.6974)证实了该方法在理论和实践上的精确性和可行性.     

Convolutional neural network based early fire detection

The detection of manmade disasters particularly fire is valuable because it causes many damages in terms of human lives. Research on fire detection using wireless sensor network and video-based methods is a very hot research topic. However, the WSN based detection model need fire happens and a lot of smoke and fire for detection. Similarly, video-based models also have some drawbacks because conventional algorithms need feature vectors and high rule-based models for detection. In this paper, we proposed a fire detection method which is based on powerful machine learning and deep learning algorithms. We used both sensors data as well as images data for fire prevention. Our proposed model has three main deep neural networks i.e. a hybrid model which consists of Adaboost and many MLP neural networks, Adaboost-LBP model and finally convolutional neural network. We used Adaboost-MLP model to predict the fire. After the prediction, we proposed two neural networks i.e. Adaboost-LBP model and convolutional neural network for detection of fire using the videos and images taken from the cameras installed for the surveillance. Adaboost-LBP model is to generate the ROIs from the image where emergencies exist Our proposed model results are quite good, and the accuracy is almost 99%. The false alarming rate is very low and can be reduced more using further training.

     

基于多传感器信息融合的智能小车避障

针对智能小车在不确定坏境下自主避障的情况,采用超声波传感器和红外传感器相结合来感知外界环境信息。将传感器采集到的各种数据利用T-S模糊神经网络进行融合．通过实验仿真表明：此方法能够使智能小车对障碍物灵活避障．     

Prediction of indoor temperature and relative humidity using neural network models: model comparison

The use of neural networks grows great popularity in various building applications such as prediction of indoor temperature, heating load and ventilation rate. But few papers detail indoor relative humidity prediction which is an important indicator of indoor air quality, service life and energy efficiency of buildings. In this paper, the design of indoor temperature and relative humidity predictive neural networks in our test house was developed. The test house presented complicated physical features which are difficult to simulate with physical models. The work presented in this paper aimed to show the suitability of neural networks to perform predictions. Nonlinear AutoRegressive with eXternal input (NNARX) model and genetic algorithm were employed to construct networks and were detailed. The comparison between the two methods was also made. Applicability of some important mathematical validation criteria to practical reality was examined. Satisfactory results with correlation coefficients 0.998 and 0.997 for indoor temperature and relative humidity were obtained in the testing stage.     

一种结合粗糙集与神经网络的多传感器信息融合算法

近年来,神经网络被广泛应用于多传感器信息融合.但是当传感器数量庞大时,过高的输入神经网络的信息维数会导致神经网络训练速度下降,甚至不收敛.针对上述问题,对传统的基于神经网络的融合算法进行了改进,利用粗糙集的冗余数据约简算法,剔除部分传感器的输入,同时将剩余的传感器信息重新组合,形成维数较小的数据分别训练,从而避免了输入数据维数过高带来的问题,较之于传统算法,算法在训练阶段的迭代次数等时间性能以及融合阶段的准确性两个方面均有所提高.     

BP神经网络在巡逻机器人多传感器数据融合中的应用

针对传统危化品仓库固定式监控器中监控范围小、报警准确率低的特点,研究了一款危化品仓库巡逻机器人,采用以拉依达准则改善BP神经网络融合性能的多传感器数据融合方法,通过采集泄露危化品浓度、仓库内环境温度和湿度等数据,在对数据进行拉依达去噪、归一化后利用BP神经网络进行融合输出.样机试验结果表明,该方法可有效提升危化品仓库巡逻机器人对空间环境的把握度,大幅度提高报警的准确性和可靠性,同时具备良好的传感器扩展性.     

图卷积神经网络在点云语义分割中的研究综述

点云数据蕴含丰富的空间信息,可以通过激光雷达、3D传感器等设备大量采集,被广泛应用于自动驾驶、虚拟现实、城市规划和3D重建等领域。点云语义分割作为3D场景理解、识别和各种应用的基础而受到广泛关注。但不规则的点云数据无法直接作为传统卷积神经网络的输入,而图卷积神经网络可以利用图卷积算子直接对点云数据进行特征提取,使得图卷积神经网络已逐步成为点云语义分割领域的一个重要研究方向。基于此,对图卷积神经网络在3D点云语义分割应用中的研究进展进行综述,根据图卷积的类型对基于图卷积神经网络的点云语义分割方法进行分类,按照不同类别对比分析主流方法的模型架构及其特点,描述几个相关点云语义分割领域常用的公共数据集和评价指标,对点云语义分割方法进行总结和展望。     

Cokriging for cross-attribute fusion in sensor networks

Optimisation of the number of required measurement points and their location is an important research topic in sensor networks. Finding the optimal positions increases spatial coverage and reduces deployment costs. This paper presents an approach for the case that two attributes have to be measured with a different number of available sensors. The proposed cokriging method performs cross-attribute fusion in sensor networks by being based on the analysis of multi-variable spatial correlations. To the best of our knowledge, this scientific work is the first one considering kriging and cokriging interpolations as IF methods. The single-variable ordinary kriging and bi-variable methods were applied to experimental data. The combination of humidity and temperature data in a refrigerated container is used as exemplary case, humidity measurements are considered to be the expensive attribute to measure. The average estimation error for intermediate points was estimated as a function of the number of humidity sensors. When variability is high, data fusion using the bi-variable method produced results as accurate as the single-variable one, without the necessity of deploying a large number of humidity measuring points, by complementing the estimation with temperature measurements.