烟气脱硫技术的模糊评价

应用模糊数学原理，将素质决策与数量决策相结合，以三种典型烟气脱硫技术为例，对所涉及的定性和定量指标进行处理，经模糊线性加权变换，评估和优选出适合我国国情，具有较好的环境，社会，经济效益的烟气脱硫技术，比较表明，模糊决策方法在脱硫技术的评估和优选中优于单一的素质决策或数量决策方法，是一种有前途的科学的定量化决策方法。     

羧酸及其衍生物急性毒性的QSAR研究

基于定量结构-活性相关性(QSAR)原理,研究了27种羧酸及其衍生化合物结构与其急性毒性LC50之间的内在定量关系。应用遗传算法从大量结构参数中优化筛选出与LC50最为密切相关的五个参数作为分子描述符,得出影响羧酸及其衍生物急性毒性的主要结构特征为分子的大小及其空间效应等。分别采用支持向量机(SVM)方法和多元线性回归(MLR)方法建立了相应的QSAR预测模型,并对所建模型分别进行了内部验证和外部验证。结果表明,两种模型均具有较高的稳定性、预测能力及泛化性能。其中,支持向量机模型对训练集和预测集样本的预测平均绝对误差分别为0.149和0.211,优于多元线性回归方法所得结果。     

住宅小区绿色生态性评价研究

通过构建住宅小区绿色生态性评价指标体系,综合运用层次分析法和模糊数学理论等系统工程理论和方法,系统评价了住宅小区的绿色生态性,为其研究提供新的途径。     

A framework for resilience assessment in process systems using a fuzzy hybrid MCDM model

Resilience engineering (RE) has recently emerged as a novel safety management paradigm in socio-technical organizations. It is believed that RE is more compatible with the characteristics of complex socio-technical systems. The multicriteria nature and the presence of both qualitative and quantitative latent factors make RE substantially more complex especially in quantifying and modeling aspects. To address this issue, the present research aims to develop a fuzzy hybrid multicriteria decision-making (MCDM) model for quantifying and evaluating resilience using the fuzzy Analytic hierarchy process (F-AHP) and fuzzy VIKOR (F-VIKOR) techniques. Initially, an evaluation framework including six resilience indicators and 43 sub-indicators was established. Afterward, the F-AHP method was used to determine the weight of the resilience indicators, while the F-VIKOR method was employed to rank the resilience performance of the different operational units. To present the model capability, we evaluated the resilience of a gas refinery as a typical instance of socio-technical systems. The findings revealed the performance level of resilience indicators in all units of the studied refinery and their ranking based on the computation of the index value (Qi). With respect to the Qi values, the best and worst performance of units from the resilience perspective was specified. Results indicate that the proposed model can serve as an effective evaluation approach in complicated systems and can be used to effectively design strategies to improve system safety performance. To the best of our knowledge, this is the first study that evaluates the resilience using the VIKOR and AHP in a fuzzy environment in the process industry.     

Correction model for CO detection in the coal combustion loss process in mines based on GWO-SVM

Accurate detection of CO gas is crucial to the prevention of coal combustion. Tuneable diode laser absorption spectroscopy (TDLAS) is a reliable method for CO detection during coal combustion. The influences of temperature and pressure cause changes in the line strength and linewidth of the index gases’ absorption spectra, leading to sizable measurement errors. To correct the distortion of the CO absorption spectrum caused by temperature and pressure fluctuation, a compensation model based on the grey wolf optimizer–support vector machine (GWO–SVM) was proposed. The results were compared with those of the single SVM, the back propagation neural network (BPNN), and multiple regression analysis (MRA). MRA was revealed to result in the lowest accuracy, which indicated that MRA is not ideal for compensation in TDLAS. The hyperparameter selection of the SVM had the disadvantages of randomness and blindness, which led to instability and large errors. The BPNN achieved better correction in the training stage, but severe overfitting occurred in the testing stage. The modified results revealed that the GWO–SVM model had higher accuracy and stability than the other models. It effectively inhibited the effects of temperature and pressure on the measured concentration and greatly improved the measurement accuracy. The equipment is thus suitable for CO gas detection with the aim to preventing coal combustion loss, and it can be further applied to loss prevention in other process fields.     

Positive long-term effect of mulching on species and functional trait diversity in a nutrient-poor mountain meadow in Central Europe

Oligotrophic mountain meadows are threatened biodiversity hotspots throughout Europe. The traditional summer hay-making followed by autumn grazing is no longer economic and question is whether alternative managements can maintain both plant species and functional diversity typical of these habitats. In the Bohemian Forest Mts., we applied three treatments (mowing once a year - i.e., cutting and removing the biomass, mulching once a year - cutting and leaving the crushed biomass to decompose in situ, fallowing - no cutting) in order to assess temporal changes in meadow vegetation, plant trait composition and biomass production in a 13-year experiment. We recorded for each species twenty-five traits as to be most informative of plant strategies related to growth, resource acquisition and carbon-water economy. We compared different components of trait composition (community averages that mostly reflect traits of dominant species vs. the Rao index of functional diversity that reflects trait dissimilarity among species) and their impact on biomass production. We show that mulching promotes species and functional diversity by facilitating heliophilous forbs and legumes with more acquisitive strategies in resource use and release, e.g., higher foliar N and P content. This occurs at the expense of tall grasses (with resource-retentive strategies, e.g., high leaf dry matter content) which dominate the mown and fallow plots. The divergence in most quantitative traits indicates that niche complementarity is the dominant assembly process in mulched plots, which can prevent competitive exclusion and enable species coexistence. The divergent development was detected only after 5-6 years. This slow floristic and functional response is caused by acidity of soil and severe mountain climate that preclude rapid responses of vegetation to land-use changes. We conclude that mulching represents a good compromise maintaining both plant species and functional diversity as well as a relatively high biomass production. Mowing without grazing leads to gradual nutrient loss and thus reduces the productivity and diversity in these oligotrophic ecosystems. Fallowing causes gradual loss in diversity by increased grass competition and litter accumulation.     

单矢量水听器抗混响方法研究

混响是主动声纳的主要干扰。矢量水听器是较新型的水声测量设备,其接收的混响和目标信号之间存在相位差异。基于这些差异,探索了矢量声能流方法用于抗混响处理的可行性,在理论上得到了较高的空间处理增益和时间处理增益。计算结果和仿真结果表明,相对于常规的声压平方积分器,该方法具有很好的抗混响效果。     

Risk assessment of coupling links in hazardous chemicals maritime transportation system

The hazardous chemicals maritime transportation system (HCMTS) belongs to a typical complex tech-social system, which is comprised of multiple functions with interactions. To quantitatively assess the risks triggered by failure coupling links (CLs) between upstream and downstream functions, a hybrid method integrating the Functional Resonance Analysis Method (FRAM), fuzzy set theory, and risk matrix is proposed in the present study. A total of 10 essential functions and 16 CLs involved in HCMTS were identified by FRAM. The fuzzy set theory was adopted to aggregate the weights of risk consequence and likelihood for CLs. A risk matrix with a continuous scale was utilized to assess and rank the CLs’ risks. The results have been demonstrated by the comparisons, indicated that the risk indexes (RIs) of the CLs between the functions < Consignment of HCs> and <Undertaking the transportation>, <Consignment of HCs> and <Declaration of HCs>, <Loading HCs in port> and <Ship navigation at sea > are higher in the whole chain of HCMTS.     

基于分形理论的中国原油管道网络结构研究

为深入分析全国及各区域原油管网的结构与布局,揭示其空间结构特征,论文运用分形理论的三种维数:加权长度-半径维数、分枝-半径维数和关联维数,分别从密度、连通性及通达性研究了全国及各区域原油管网。研究结果表明:全国原油管网的密度和连通性均存在明显的东西分异特征,密度自东向西呈"线状-散点"的下降形态,连通性自东向西呈"散点-线状"的升高形态,而通达性则普遍较低;对各区域来说,东北、华北和华东等东部管网除连通性较低外,其他分形指标相对较高;西北管网的密度、通达性均不如东部,而且其新疆管网的连通性数值超过2,是不正常的连通结构;华南管网的多数分形指标均低于其他地区,其管网结构雏形尚未形成。     

SVM与ANN在湖泊富营养化评价中的对比研究

支持向量机是由Vapnik等提出的建立在统计学习理论基础上的一种新的机器学习方法,由于其使用结构风险最小化原则代替经验风险最小化原则,又由于其应用了核函数思想,它可以较好地解决非线性问题;人工神经网络(ANN)已经较成功解决模式识别和任意非线性函数回归问题,但是存在训练样本不足,并可能出现过拟合现象。SVM的结构风险最小化算法引起了科学界的关注,对传统基于经验风险最小化的神经网络算法提出了挑战,文章介绍了SVM和ANN的基本原理,并对二者在巢湖富营养化水平评价上做对比研究,结果表明,ANN比较容易陷入局部最优,支持向量机评价结果更加符合实际。