灰色—支持向量机的空调负荷预测

对空调负荷进行准确预测不仅对空调优化控制的意义重大,而且也是实现空调经济运行与节能的关键所在。为了提高建筑空调负荷的预测精度,在分析灰色模型和支持向量机建模特点基础上提出了一种空调负荷组合预测算法。该方法综合了灰色建模计算过程简单以及支持向量机自学习和泛化能力强的优点,能够更加有效地利用样本数据的有效信息,提高模型预测精度。首先,通过灰色建模过程弱化了样本数据的随机因素。然后,对灰色模型输出进行归一化处理及数据重构,以作为支持向量机的输入。最后,通过支持向量机模型的预测得到最终预测结果。将本文所提出的方法应用于福州一栋办公建筑的逐时空调负荷预测中,并与灰色模型及支持向量机模型作比较,证明了组合模型的预测值与实际运行值拟合度最高,平均绝对误差比灰色模型和支持向量机模型分别降低了47.84%和17.39%。该组合预测模型具有较高的预测精度和更好的泛化能力,具有较强的可行性和实用性。     

人工鱼群—粒子群混合算法优化进港航班排序

针对空中交通管理中的进港航班排序问题, 提出了人工鱼群—粒子群混合算法(AFPSO)这一航班排序算法来优化进港航班排序, 使时段内进港航班队列总延误时间最少。算法结合了基本人工鱼群算法(AFSA)和基本粒子群算法(PSO)各自的优点, 先以AFSA在全局寻找满意的解域, 再以PSO算法在这些解域中进行快速的局部搜索获得精确解, 最终使算法提高收敛速度和搜索精度。仿真结果表明, 在单跑道和双跑道情况下, AFPSO算法使得航班队列总延误时间比FCFS调度方法减少了20. 9%和34. 4%, 比基本AFSA减少了3. 2%和3. 5%。算法得到的满意解能够为自动化空中交通管理提供实时支持。     

基于二次型性能指标的混合动力汽车实时优化控制

为解决混合动力系统实时优化控制问题,本文提出了一种基于二次型性能指标最优的混合动力汽车功率分配优化方案.通过合理的假设和近似,建立了混合动力系统的线性模型,并利用二次型最优控制理论将混合动力最优控制问题转化为二次型最优调节问题进行求解,得到了一个结构简单的实时优化控制算法.5种道路工况下的仿真结果表明,本文提出的控制方法在未来道路工况未知的情况下能够实现混合动力系统的实时优化控制,且节油率与离线计算以燃油消耗最小为性能指标的全局最优控制的节油率相近.     

使用新混合模糊聚类算法的模糊系统建模方法

为了进一步提高模糊系统建立模型的精度,提出一种新的模糊系统算法ANFIS-HC-QPSO：采用一种混合型模糊聚类算法来对模糊系统的输入空间进行划分,每一个聚类通过高斯函数的拟合产生一个隶属度函数,即完成ANFIS系统的前件参数--隶属度函数参数的初始识别,通过具有量子行为的粒子群算法QPSO与最小二乘法优化前件参数,直至达到停机条件,最终得到ANFIS的前件及后件参数,从而得到满意的模糊系统模型。实验表明,AN-FIS-HC-QPSO算法与传统算法相比,能在只需较少模糊规则的前提下就使模糊系统达到更高的精度。     

β-环糊精-氧化石墨烯超分子杂化体的构筑及应用进展

β-环糊精是由7个D-吡喃葡萄糖单元通过α-1,4-糖苷键键连成环的超分子主体分子，“内疏水、外亲水”的独特结构赋予了其优异的分子识别能力；氧化石墨烯类材料凭借其优良特性成为近几年的研究热点。由β-环糊精和氧化石墨烯构筑的超分子杂化体在兼具二者特有性能的基础上又有新功能的引入。本文综述了β-环糊精-氧化石墨烯超分子杂化体的构筑方式，按二者间的连接方式，分别为共价键和非共价键两种连接方式，其中通过共价键连接是目前最主要的构筑方式；此外对β-环糊精-氧化石墨烯超分子杂化体的特征和表征进行了简述。同时对β-环糊精-氧化石墨烯超分子杂化体在水污染处理、电化学检测、药物控释和催化等领域的应用进展进行了综述。最后对该超分子杂化体在构筑和应用上的发展趋势进行了展望。     

一种新型蓄电池充电控制策略研究

在风光互补发电系统中,蓄电池作为风光互补发电系统的储能设备,在整个发电系统中起着非常重要的作用,而蓄电池充电控制方法的优劣,是影响其使用寿命的一个重要因素。文章对风光互补发电系统中蓄电池的常用充电控制方法进行了分析,提出了一种新的控制策略,即将蓄电池充电电压、充电电流采样值分别与蓄电池事先设定好的恒压过充电压、恒流充电限流值和浮充电流值进行实时比较,来判断采取哪种合理的充电方式,对蓄电池起到尽可能的保护,从而延长了风光互补发电系统中蓄电池使用寿命。     

智能混合控制器在转炉干法除尘中的应用

转炉干法除尘蒸发冷却器出口烟气温度大多数采用常规串级控制方式,但转炉冶炼前期控制效果并不理想.对转炉冶炼过程及蒸发冷却器出口烟气温度的控制难点进行研究,找出常规串级控制方式的缺点,并设计出一种智能混合控制器.采用智能模糊控制器和常规控制器并联的控制方式,对蒸发冷却器出口烟气温度进行控制,控制效果理想.实际应用表明,同常规串级控制器相比,智能混合控制器具有响应快、被控对象稳定等优点.     

Hybrid systems in seawater desalination—practical design aspects, present status and development perspectives

Ever since seawater desalination has been applied on an industrial scale, and particular in the countries of the Arabian Gulf, the application of desalination processes in dual-purpose facilities—water and power—as a hybrid configuration has been discussed in many feasibility investigations and also planning concepts. It is above all the combination of reverse osmosis with thermal processes that has found increasing interest with the aim of ensuring, as economically as possible, uniform water supplies under the specific, greatly varying load conditions in the Gulf countries. Such design concepts for hybrid configurations encompass straightforward structures with a low degree of coupling between membrane and thermal desalination processes, but range up to very complex configurations with strong interconnections on both the water side and thermally, as well as with several desalination processes connected in series or in parallel. Classical hybrid concepts in which the permeate from an RO desalination component is mixed with distillate from thermal desalination have already been implemented in Saudi Arabian dual-purpose plants, like Jeddah and Yanbu-Medina. Although hybrid systems of greater complexity have been addressed in many design studies and publications, up to now none has been brought to fruition. Coming into consideration asthe design basis for determining the capacity shares of the various desalination processes operated in a hybrid configuration are: arrangement of thermal cycle of the power plant component; water/power ratio of the dual-purpose seawater desalination and power plant; provision of undiminished water production of the desalination plant as electricity generation varies; provision of a specified drinking water quality with regard to composition and salt content; combination of all these aspects. Also gaining in importance are concerns of environmental pollution and sustainable development when selecting seawater desalination and power plant configurations, as well as their optimization when considering desalination and electricity generation as a whole. In the practical design of hybrid membrane and thermal systems, aspects come to light, though, that restrict linking of the two systems and joint utilization of facilities, as conceived in studies and conceptual design investigations. This applies both for common utilization of intakes and the use of heated up cooling water from thermal processes as a feed stream for the RO part of the desalination process. Additionally, requirements of drinking water composition, particularly chloride content, TDS and compliance with a specific residual content of boron, influence specifically the design of the membrane process part and its share in the total desalination capacity. Such practical aspects have greatly influenced the design and configuration of the Fujairah hybrid plant for which, from a total desalination capacity of 100 MIGD (454,600 m³/d), the share of 37.5 MIGD (170,500 m³/d) makes its seawater RO plant the biggest currently being constructed anywhere in the world. From the findings of the engineering of this plant and the idea that, by increasing interconnection between the two processes on the water side, it is possible to advance a hybrid configuration of this type with regard to cost optimization in the membrane installation, but also by joint utilization of the intake equipment, perspectives result for applied research efforts over the near and long terms, for example: long-term behavior of membranes at elevated temperatures; tendency for biofouling in membrane process with common utilization of cooling water and brine; influences of such interconnections on the overall availability of the facility. But also for the operation and maintenance organization of such large facilities, consequences can be foreseen for the future development of hybrid plants, particularly for operation management and organisation of the interplay of the different power plant and desalination systems, monitoring of SWRO membrane replacement and cleaning, as well as controlling water quality.     

用人工神经网络预测混杂复合材料混杂效应

本研究建立一个二输入单输出的BP人工神经网络模型，并用QBASIC语言编制了相应的软件。利用该神经网络模型对混杂效应与混杂比及分散度系数间关系进行了预测。研究结果表明，网络经过61223次的迭代，预潮值误差为0．12％，具有很高的预测精度，可用于混杂复合材料混杂效应的预测。