基于GMDH建模的蒸汽发生器水位重构

蒸汽发生器水位指示仪表出现虚假指示或丧失指示的情况时有发生,而目前又没有很好的方法实现蒸汽发生器水位的重新标定,主要靠经验来进行判断,所以当事故或故障发生时严重影响操纵员对核动力装置运行情况的判断。自组织理论模型（GMDH）是建立复杂非线性大系统数学模型十分灵活而通用的方法,在处理复杂非线性对象中能得到很好的效果。本文以主蒸汽管道破口事故下重构蒸汽发生器水位为例,提出了用GMDH重构蒸汽发生器水位的方法,并与仿真结果进行对比。结果表明,GMDH对蒸汽发生器水位重构的相对误差小、精度高,满足实际需要,能为船用核动力装置的安全运行做出指导。     

Application of the FCM-based neuro-fuzzy inference system and genetic algorithm-polynomial neural network approaches to modelling the thermal conductivity of alumina–water nanofluids

By using an FCM-based neuro-fuzzy inference system and genetic algorithm-polynomial neural network as well as experimental data, two models were established in order to predict the thermal conductivity ratio of alumina (Al₂O₃)–water nanofluids. In these models, the target parameter was the thermal conductivity ratio, and the nanoparticle volume concentration, temperature and Al₂O₃ nanoparticle size were considered as the input (design) parameters. The empirical data were divided into train and test sections for developing the models. Therefore, they were instructed by 80% of the experimental data and the remaining data (20%) were considered for benchmarking. The results, which were obtained by the proposed FCM-based neuro-fuzzy inference system (FCM-ANFIS) and genetic algorithm-polynomial neural network (GA-PNN) models, were provided and discussed in detail.     

水资源需求的驱动力分析及其预测

根据水资源需求系统的特点和所涉及的范围,以水资源需求势能概念为基础,分析了水资源需求驱动力的组成,并引进自组织数据挖掘的方法来构建驱动力合成模型,以广东省东江流域为例,对模型进行了验证分析。研究结果表明:水资源需求的驱动力主要由自然条件力、社会经济力和水资源管理政策力三力合成;基于自组织数据挖掘的水资源需求量预测模型可有效地模拟和预测出水资源需求变化过程,这为研究水资源需求预测提供了一种新的方法。     

基于GMDH因果关系的软件缺陷预测模型

软件缺陷预测是软件可靠性研究的一个重要方向。基于自组织数据挖掘(GMDH)网络与因果关系检验理论提出了一种软件缺陷预测模型,借鉴Granger检验思想,利用GMDH网络选择与软件失效具有因果关系的度量指标,建立软件缺陷预测模型。该方法从复杂系统建模角度研究软件度量指标与软件缺陷之间的因果关系,可以检验多变量之间在非线性意义上的因果关系。最后基于两组真实软件失效数据集,将所提出的方法与基于Granger因果检验的软件缺陷预测模型进行比较分析。结果表明,基于GMDH因果关系的软件缺陷预测模型比Granger因果检验方法具有更为显著的预测效果。     

基于最优组合的一类非线性建模方法及其在预测中的应用

１引言专门人才预测大系统不仅涉及社会、经济、科技、教育等诸方面,而且受地理、历史背景和民族特点的约束．因此,如何利用现有条件定量地预测未来专门人才的需求仍是一个极待解决的问题．２最优组合准则及其探讨用基本数据处理组合法（ＧｒｏｕｐＭｅｔｈｏｄｏｆＤａ...     

Modeling of convection heat transfer of supercritical carbon dioxide in a vertical tube at low Reynolds numbers using artificial neural network

Today, many researches have been directed on heat transfer of supercritical fluids; however, since the analysis of heat transfer in these fluids founded by a mathematical model based on the effective parameters is complicated, so in this paper, a group method of data handling (GMDH) type artificial neural network are used for calculating local heat transfer coefficient h_x of supercritical carbon dioxide in a vertical tube with 2 mm diameter at low Reynolds numbers (Re < 2500) by empirical results obtained by Jiang et al. [1].At first, we considered h_x as target parameter and G, Re, Bo^?, x⁺ and q_w as input parameters. Then, we divided empirical data into train and test sections in order to accomplish modeling. We instructed GMDH type neural network by 80% of the empirical data. 20% of primary data which had been considered for testing the appropriateness of the modeling were entered into the GMDH network. Results were compared by two statistical criterions (R² and RMSE) with empirical ones. The results obtained by using GMDH type neural network are in excellent agreement with the experimental results.     

Proton exchange membrane fuel cell performance prediction using artificial neural network

Polarization curves remain one of the parameters used to check the performance of fuels in terms of efficiency and durability. This investigation explores the application of artificial neutral network (ANN) to determine the voltage and current from a proton exchange membrane fuel cell having membrane area of 11.46 cm². Performance predictability for the group method of data handling (GMDH) as well as feed forward back propagation (FFBP) neutral networks were employed for the estimation of the current and voltage obtained from the Proton Exchange Membrane Fuel cell under investigation. The investigation presented models with good predictions even though GMDH neural network performed better than the FFBP neural network. The study therefore proposes the GMDH neural network as the best model for predicting the performance of a Proton Exchange Membrane Fuel cell. It was further deduced that an increase in reactant flow rate has direct effect on the performance of the fuel cell but this is directly proportional to the total irreversibilities in the cell hence to operate fuel cell economically, it is imperative that the hydrogen flow is made lower compare to the oxygen flow rate. This in effect will reduce the pumping power required for the flow of the fuel hence reducing the net loss in the cell.     

GMDH-PTSV 算法及其在发酵过程建模中的应用

本文提出了GMDH-PTSV(具有主成份双向变量选择器的GMDH)建模算法。在此算法中,变量的选择是根据对系统的测量数据,而不是根据人的经验构造出的(0-1)相关矩阵。通过仿真,对GMDH-PTSV算法和其它三种算法进行了比较。最后,应用GMDH-PTSV算法对一种生物发酵过程建立了模型。     

自组织数据挖掘方法研究综述

从对自组织数据挖掘方法(SODM)本身的改进和SODM与其他建模方法结合两方面介绍了SODM的研究进展.对SODM本身改进的方面包括模型表示、选择准则和搜索方法.分析SODM与神经网络、遗传算法、模糊法则归纳法以及模拟方法的结合.在此基础上,指出SODM改进方向应该在改善算法本身的基础上,将其从单纯的数据挖掘技术扩展为知识发现的全过程.