A model algorithm for defining the vertical profile of absolute humidity by ground measurements of humidity

In this study a double model algorithm is developed for the simulation of the vertical profiles of monthly values of absolute humidity. The thrust of the algorithm is that is depends on the use of ground measurements for humidity and is based upon the use of two model relations for altitudes above and below 1 Km respectively. Monthly absolute humidity profiles as estimated using the specially constructed double model algorithm compare well with those calculated with the use of radiosonde data for the time period 1980–1990 (Helliniko station, Southern Greece). In particular the model estimation is very satisfactory at altitudes above the Planetary Boundary Layer and rather satisfying at altitudes below 1 Km. This implies that the model algorithm may be useful in supporting climatological studies in areas where lack of relevant information on the vertical distribution of water vapour is recognised.     

Energy consumption estimation model for dual-motor electric vehicles based on multiple linear regression

ABSTRACT

The drive range of electric vehicle (EV) is one of the major limitations that impedes its universalism. A great deal of research has been devoted to drive range improvement of EV, an accurate and efficiency energy consumption estimation plays a crucial role in these researches. However, the majority of EV’s energy consumption estimation models are based on single motor EV, these models are not suitable for dual-motor EVs, which are composed of more complex transmission mechanisms and multiple operating modes. Thus, an energy consumption estimation model for dual-motor EV is proposed to estimate battery power. This article focuses on studying the operating modes and system efficiency in each operating mode. The limitation of working area of each mode ensures the vehicle dynamic performance, then PSO algorithm is adopted to optimize the torque (speed) distribution between two motors to improve the system efficiency in the coupled driving mode. Finally, the energy consumption estimation model is established by multiple linear regression (MLR). The result shows that the proposed model has a high precision in energy consumption estimation of dual-motor EV.     

Kinetic parameter estimation for decomposition of organic peroxides by means of DSC measurements

The thermal stability of organic peroxides (cumene hydroperoxide 80 wt% and dicumyl peroxide) was studied by means of calorimetric measurement (DSC, TA Q1000) in an isotherm mode and a dynamic mode. Analysis of power profiles released in the isothermal mode was combined with the analysis of the decomposed compounds by a gas chromatograph/mass spectrometer (GC/MS) to determine the reaction mechanisms corresponding to each of the two reactions. In this work, a methodology for estimating kinetic parameters was based on the comparison of the power profile (dynamic mode) given by the model to that obtained experimentally by changing the parameters values. Parameter estimation is achieved using the mixed estimation method where a genetic algorithm is combined with a locally convergent method.     

基于在线检测信息的储罐底板腐蚀状态智能评价方法

为了解决利用声发射在线检测技术对储罐底板腐蚀状态进行评价时,主要依赖检测人员经验的问题,使该项技术能更好地推广和应用,利用储罐底板在线检测的声发射信息和外观检查信息,并根据相关标准及专家经验,确定与储罐底板腐蚀状态相关的表征因素。采用遗传算法(GA)改进贝叶斯网络(BN)搜索方法,建立基于GA的BN智能评价方法。针对声发射在线检测信息和外观检查信息,分别建立基于标准用声发射因素、基于声发射因素和综合考虑声发射因素和外观检查因素的基于在线检测信息的储罐底板腐蚀状态评价模型。通过对测试样本的评价,对比声发射检测专家评价结果,其中基于在线检测信息的储罐底板腐蚀状态评价模型的准确率为96%,该模型能够对储罐底板腐蚀状态进行可靠的智能评价。     

基于遗传算法的支持向量机预测有机物自燃点的研究

根据定量构效关系(QSPR)原理,研究自燃点(AIT)与其分子结构间的内在定量关系。以265种有机化合物作为样本集,随机选择238种作为训练集,27种作为测试集,用遗传算法(GA)进行变量选择,分别建立多元线性回归(MLR)模型和支持向量机(SVM)模型研究有机物的自燃点与其分子结构间的关系。通过分析,发现造成模型预测效果不佳的原因是试验数据本身存在问题。通过对2个模型的比较,结果为GA-SVM模型明显优于GA-MLR模型,说明自燃点与其分子结构间具有很强的非线性关系。     

基于智能优化算法的河流水质模型参数的优化简

为探索河流水质模型参数新的求解方法,根据有限的实测数据,分别应用免疫进化优化算法和免疫进化优选的捕食搜索算法,对河流水质模型计算公式中的多参数进行优化。将优化得到的计算公式用于国内外若干河流的河段中DO浓度值的拟合,并与实测结果进行了比较。结果表明,将免疫进化优化算法或免疫进化优选的捕食搜索算法优化得到的水质模型参数精度不仅较高,而且相对稳定,从而为河流水质模型参数的优化提供了一种新方法。     

基于支持向量机的地下水质量综合评价

提出基于支持向量机（SVM）机器学习算法的地下水质量评价模型。首先给出了训练样本生成和数据规范化处理的具体方法,然后采用支持向量机的多分类算法构建模型,并使用k折交叉核实方法对参数进行验证优化。最后通过实证分析,并与单因子指数法、模糊综合评价法和BP神经网络法的评价结果对比分析可知,该方法简便易行,评价结果客观且准确度较高,具有很强的实用性。     

基于主成分和粒子群优化支持向量机的水质评价模型

水质的评价是治理水污染必不可少的工作。为了准确、快速地对水质进行评价,利用主成分分析法从水质监测常见的多个物化指标提取出主成分,然后将主成分作为支持向量分类机的输入,利用历史数据进行水质评价训练并用粒子群算法优化参数,构造出水质评价模型,将从物化指标中得出的主成分代入此模型即可得到水质类别。最后,选取水质监测点实测数据进行试验,结果表明,模型的水质评价结果准确且稳定。     

模拟退火算法在大气环境质量综合评价中的应用

为了建立一种对多项指标都能普遍使用的大气环境质量综合评价模式,利用模拟退火算法对幂函数加和型指数评价中参数进行优化,并用优化后的综合评价指数公式对大气环境进行综合评价。该公式应用于大量实例的分析评价结果与其他多种评价方法评价结果比较表明:该公式不受污染物种类和数目多少的限制,计算简便,具有可行性、实用性和通用性。