嵌入式粒子群-遗传算法的水质COD检测特征波长优化算法

基于紫外-可见光谱法的水质测量中,光谱信号易受到系统噪声干扰、悬浮物散射干扰,且存在信息冗余、多重共线性等特征,导致水质COD测量中特征波长的选取产生较大偏差。因此,提出了基于嵌入式粒子群-遗传（EPSO_GA）算法的水质COD检测特征波长优化算法,以提高波长选择精度。为验证检测特征波长优化算法的可行性,采集了某高校池塘水样、生活污水和排水沟水样的光谱数据,利用EPSO_GA算法对预处理后的光谱数据选取特征波长。EPSO_GA算法采用实数编码方法实现了粒子群（PSO）优化算法和遗传（GA）优化算法的统一编码,在PSO算法中更新粒子时嵌入GA算法的选择、交叉、变异等操作,改善了这两种算法各自在光谱波长特征选取问题上的局限性。将EPSO_GA算法选取的特征波长结合偏最小二乘法（PLS）构建了EPSO_GA_PLS的水质COD预测模型,并且与传统的PSO算法、GA算法选取特征波长建立的PSO_PLS、GA_PLS和全光谱构建的PLS水质COD预测模型做了对比。结果表明：与PSO_PLS,GA_PLS和全光谱构建的PLS水质COD预测模型相比,EPSO_GA改善了PSO算法和GA算法在光谱特征波长选择中早熟和收敛速度慢的问题,降低了全光谱构建PLS水质COD预测模型的复杂度,提高了模型的预测精度。基于EPSO_GA算法建立的EPSO_GA_PLS水质COD预测模型,均方根误差降到了0.212 3,预测精度增加到0.999 3,可以快速定量检测水质COD,为紫外-可见光谱法测COD提供了更好的预测模型。     

改进SVR的内陆水体COD高光谱遥感反演

高光谱数据可以捕获内陆水体中不同浓度的化学需氧量（COD）引起的光谱变化,因此研究光谱反射率与COD浓度之间的关系对于COD的遥感估算至关重要。支持向量回归模型（SVR）具有适合小样本、泛化能力好的特点,基于SVR模型能够更加准确获得COD浓度和光谱数据之间的关系,但仍然存在参数选取困难和易陷入局部极值的问题。为了解决这个问题,将模拟退火-粒子群算法（SA-PSO）引入到支持向量回归机的参数优化过程中,提出了一种改进SVR（SA-PSO-SVR）的内陆水体COD高光谱遥感反演方法。以潍河流域为研究区域,通过野外测量获得了COD浓度和水表面光谱反射率。首先根据光谱反射率对COD的响应来确定敏感因子,把SA-PSO算法引入SVR的参数优化过程中建立了COD浓度与敏感因子之间的反演模型。最后利用珠海一号高光谱数据验证模型的准确性,进而获得了COD浓度的分布情况。通过光谱分析,可知该区域实测的水面光谱具有典型的二类水体特征,光谱曲线形状呈现明显的双峰特征,当浓度增加时,反射峰具有向短波长方向移动而反射谷向长波长方向移动的趋势。通过计算Pearson相关系数分析COD浓度和光谱之间的相关性,结果表明最佳的反演因子为518 nm/940.4 nm,623.6 nm/636.8 nm,729.2 nm/890.9 nm和752.3 nm/857.9 nm的四个波段比值组合;经过SA-PSO-SVR方法建立的COD估计模型的平均相对误差（MRE）和均方根误差（RMSE）分别为1.62％和2.99 mg·L-1（R2=0.86）,反演结果优于其他模型（SVR、BP神经网络和线性回归模型）。将实测水面光谱建立的最优模型应用于高光谱卫星影像上,RMSE和MRE分别为4.47 mg·L-1和11.87%。获得的潍河-峡山水库区域的COD反演结果显示：COD的整体浓度介于17~42 mg·L-1之间,韩信坝、峡山水库的东北部、渠河注入潍河的交汇处等区域的COD浓度高于其他水域。证实了SA-PSO-SVR是一种有效的COD高光谱反演方法,可供潍河流域水资源管理提供参考。     

A tabu search for Time-dependent Multi-zone Multi-trip Vehicle Routing Problem with Time Windows

We propose a tabu search meta-heuristic for the Time-dependent Multi-zone Multi-trip Vehicle Routing Problem with Time Windows. Two types of neighborhoods, corresponding to the two sets of decisions of the problem, together with a strategy controlling the selection of the neighborhood type for particular phases of the search, provide the means to set up and combine exploration and exploitation capabilities for the search. A diversification strategy, guided by an elite solution set and a frequency-based memory, is also used to drive the search to potentially unexplored good regions and, hopefully, enhance the solution quality. Extensive numerical experiments and comparisons with the literature show that the proposed tabu search yields very high quality solutions, improving those currently published.     

Municipal Water Demand Forecasting: Tools for Intervention Time Series

This article introduces some approaches to common issues arising in real cases of water demand prediction. Occurrences of negative data gathered by the network metering system and demand changes due to closure of valves or changes in consumer behavior are considered. Artificial neural networks (ANNs) have a principal role modeling both circumstances. First, we propose the use of ANNs as a tool to reconstruct any anomalous time series information. Next, we use what we call interrupted neural networks (I-NN) as an alternative to more classical intervention ARIMA models. Besides, the use of hybrid models that combine not only the modeling ability of ARIMA to cope with the time series linear part, but also to explain nonlinearities found in their residuals, is proposed. These models have shown promising results when tested on a real database and represent a boost to the use and the applicability of ANNs.     

Modeling a stochastic multi-objective supplier quota allocation problem with price-dependent ordering

One of the interesting subjects in supply chain management is supply management, which generally relates to the activities regarding suppliers such as empowerment, evaluation, partnerships and so on. A major objective of supplier evaluation involves buyers determining the optimal quota allocated to each supplier when placing an order. In this paper, we propose a multi-objective model in which purchasing cost, rejected units, and late delivered units are minimized, while the obtained total score from the supplier evaluation process is maximized. We assume that the buyer obtains multiple products from a number of predetermined suppliers. The buyer faces a stochastic demand with a probability distribution of Poisson regarding each product type. A major assumption is that the supplier prices are linearly dependent on the order size of each product. Since demand is stochastic, the buyer may incur holding and stockout costs in addition to the regular purchasing cost. We use the well-known L-1 metric method to solve the supplier evaluation problem by utilizing two meta-heuristic algorithms to solve the corresponding mathematical problems.     

基于重新排序的退化工件最小化总延误时间问题

考虑了错位限制下的含有退化工件的重新排序问题,即工件的实际加工时间看作是工件开工时间的线性函数．重新排序就是在原始工件已经按照某种规则使目标函数达到最优时有一新工件集到达,新工件的安排使得原始工件重新排序进而产生错位．研究了最大序列错位和总序列错位限制下的退化工件最小化总延误时间问题,其最优排序的结构性质是使得原始工件集和新工件集中的工件是按加工率αj非减的序列排列,基于此通过分阶段排序和动态规划方法给出了两个问题的多项式时间的最优算法．     

A rapid analytical method for predicting the oxygen demand of wastewater

In this study, an investigation was undertaken to determine whether the predictive accuracy of an indirect, multiwavelength spectroscopic technique for rapidly determining oxygen demand (OD) values is affected by the use of unfiltered and turbid samples, as well as by the use of absorbance values measured below 200 nm. The rapid OD technique was developed that uses UV–Vis spectroscopy and artificial neural networks (ANNs) to indirectly determine chemical oxygen demand (COD) levels. It was found that the most accurate results were obtained when a spectral range of 190–350 nm was provided as data input to the ANN, and when using unfiltered samples below a turbidity range of 150 NTU. This is because high correlations of above 0.90 were obtained with the data using the standard COD method. This indicates that samples can be measured directly without the additional need for preprocessing by filtering. Samples with turbidity values higher than 150 NTU were found to produce poor correlations with the standard COD method, which made them unsuitable for accurate, real-time, on-line monitoring of OD levels.     

SPE-HPLC purification of endocrine-disrupting compounds from human serum for assessment of xenoestrogenic activity

Assessment of xenoestrogenic activity in human serum samples requires the removal of endogenous sex hormones to assure that the activity measured originates from xenobiotic compounds only. Serum samples representing high, medium and lower accumulation of persistent organic pollutants (POPs) were extracted using solid-phase extraction (SPE) followed by normal-phase high-performance liquid chromatography (NP-HPLC) for separation of POPs from endogenous hormones. The recovery of polychlorinated biphenyl (PCB) congeners in spiked serum samples was up to 86 %, making the extraction method suitable for the study. MVLN cells, stably transfected with an estrogen receptor (ER) luciferase reporter vector (estrogen response element chemically activated luciferase expression, ERE-CALUX), were exposed to the reconstituted SPE-HPLC extracts for determination of the integrated estrogenic activity. The effects of PCBs were analyzed by direct in vitro exposure of PCBs (nos. 138, 153, 180) and by ex vivo analysis of SPE-HPLC extracts from serum spiked with the PCBs. Similar effects on ER transactivation were observed for the direct in vitro and the ex vivo analysis experiments. The ER transactivation responses determined for actual serum samples were in the linear range of the dose-response curve. 17β-Estradiol titrations showed that the xenoestrogenic effects were mediated via ER. Moreover, our SPE-HPLC-ERE-CALUX assay was demonstrated to elicit high interlaboratory correlation. In the present study the combination of SPE-HPLC purification and the ex vivo estrogenic responses measured by ERE-CALUX was validated and considered to be a valuable tool to assess the combined ER effect of lipophilic serum POPs where additive/synergistic and agonistic/antagonistic effects are integrated giving an overall estimate of exposure and bioactivity.      

房地产行业供需模型及调控模型的研究

首先,以房地产市场的非均衡计量经济模型为基础,得到房地产住房供需模型;然后根据"在有确定调控结果的前提下找出具体调控策略"的思想,建立了与房价、住房需求、住房供给有关的倒向常微分方程作为房地产调控模型,为政府提供倒向调控依据,并根据房地产的可持续发展模型检验调控策略的成效.     

基于对故障规律统计分析的汽车备件需求预测

汽车备件的需求与汽车故障紧密相关,文章介绍了一种在对汽车故障进行统计分析并确定其分布规律的基础上预测备件需求的方法,预测中需要结合整车保有量的历史数据以及故障与备件的对应表。用统计的方法对某型客车的故障信息进行分析,认为故障的规律可用四种典型的分布进行描述。实例验证了这种方法的准确性高于传统方法,并且在计算机的辅助下可以方便操作。