A practical assessment of risk-averse approaches in production lot-sizing problems

ABSTRACT

This paper presents an empirical assessment of four state-of-the-art risk-averse approaches to deal with the capacitated lot-sizing problem under stochastic demand. We analyse two mean-risk models based on the semideviation and on the conditional value-at-risk risk measures, and alternate first and second-order stochastic dominance approaches. The extensive computational experiments based on different instances characteristics and on a case-study suggest that CVaR exhibits a good trade-off between risk and performance, followed by the semideviation and first-order stochastic dominance approach. For all approaches, enforcing risk-aversion helps to reduce the cost-standard deviation substantially, which is usually accomplished via increasing production rates. Overall, we can say that very risk-averse decision-makers would be willing to pay an increased price to have a much less risky solution given by CVaR. In less risk-averse settings, though, semideviation and first-order stochastic dominance can be appealing alternatives to provide significantly more stable production planning costs with a marginal increase of the expected costs.     

Visualization of temporal differences between dominant perceptions in temporal dominance of sensations (TDS) and temporal check-all-that-apply (TCATA) perceptions using dominance-highlighted TCATA (dTCATA) curves

The temporal dominance of sensations (TDS) method measures dynamic changes of panelists’ attention to the sensory attributes of products. The temporal check-all-that-apply (TCATA) method measures all sensory attributes perceived at each moment of an evaluation. However, unlike in TDS, significant levels cannot be calculated in TCATA. This study proposes the use of dominance-highlighted TCATA (dTCATA) curves, which are highlighted TCATA curves that show significant time periods for the TDS data of different panels. Twelve R&D panelists evaluated five commercial corn soups using the TCATA method. Then, 125 consumer panelists evaluated the same products using the TDS method. The dTCATA curves showed TCATA curves for all attributes for each product evaluated by the R&D panel highlighted with the dominance rates identified by the consumer panel in the TDS evaluation. For example, for product 1, some attributes (sweet, viscosity) showed relatively high citation proportions in the TCATA evaluations of the R&D panel and significant dominance rates in the TDS evaluations of the consumer panel. In contrast, consommé flavor showed relatively low citation proportions in TCATA but significant dominance rates in TDS. By merging TDS and TCATA data, we could compare consumers’ dominant sensations with the evaluations of R&D panelists. This comparison could provide useful insights to product developers. In some cases, we observed attributes with significant dominance rates that were under-identified by the R&D panel in TCATA. This could suggest that most of the R&D panel may not have perceived these attributes; therefore, during product development, these attributes should be carefully considered.     

纤维面层表面非稳态过滤效率的研究

在介绍纤维过滤和颗粒过滤的基本理论基础上，根据表面过滤的“尘滤尘”现象，应用颗过滤理论建立了纤维层表面非稳态过滤效率的数学模型。对比研究表明该模型能较深刻地揭示了非稳态过滤过程随过滤进间的变化规律。从而对深化纤维过滤器的净化机理的认识，提高过滤效率，延长滤料使用寿命，加强清灰管理等方面有理论及应用意义。     

Aerobic granular sludge in a sequencing batch reactor

In a laboratory scale sequencing batch reactor (SBR) granules were cultured under aerobic conditions. To enhance the growth of granular sludge the SBR was operated with very short sedimentation and draw phases resulting in the washout of slow settling biomass. Fast settling granules were retained in the reactor and thus had an advantage over flocs with a slower settling velocity. After 40 days of operation granules were the dominant form of microbial aggregates in the reactor, even though some pin-point flocs remained in the system. Granules taken from the reactor were stored for weeks without disintegrating. After about 130 days of operation the granule quality and COD-removal worsened. The reasons for that are yet to be investigated.     

钙质砂的耦合变形机制与本构关系探讨

针对低应力下钙质砂的颗粒破碎现象具有特殊性，分析了围压对钙质土颗粒的破碎的影响，揭示其变形过程的力学机制；然后考虑钙质土的颗粒破碎与滑移两种变形机制的耦合作用，并分别采用损伤模型和边界模型予以描述。理论计算与试验结果吻合较好。     

高效液相色谱法测定消炎利胆冲剂中芍药苷的含量

采用超声法提取消炎利胆冲剂中的芍药苷,建立了高效液相色谱法测定芍药苷含量的方法。采用Agilent ZORBAX SB-C18色谱柱(250mm×4.6mm,5μm),柱温26℃,流动相为0.2%磷酸溶液-甲醇(58∶42,体积比),流速1.0mL·min-1,检测波长230nm。结果表明:在2.016~336μg·mL-1的范围内芍药苷浓度与峰面积呈良好的线性关系,平均回收率为100.42%,RSD为0.97%。该方法简便、准确、重复性好,可用于消炎利胆冲剂中有效成分芍药苷的含量测定。     

基于人工免疫的机车二系调簧方法

为进一步提高电力机车二系支承载荷调整优化算法的性能,解决已有算法在加垫数量控制上的欠缺,并避免冗余计算,提出了一种基于人工免疫算法的两级结构机车二系载荷调整方法:免疫优势克隆选择多目标算法。针对调簧问题的偏好结构特点,设计具有调簧寻优阶段和可控加垫量寻优阶段的两级寻优算法,分别定义了两个阶段的寻优目标和可行域,进一步将调簧问题的先验知识作为免疫优势引入算法设计中,并结合调簧问题的具体特点以及人工免疫知识设计了特定的抗体编码形式、变异策略和免疫优势获得算法。对不同车型的多次应用试验统计结果表明:该算法不仅具有良好的优化载荷分布特性,还能稳定地控制调簧产生的加垫数量,并且能够避免冗余计算,极大地提高了调簧效率和实际应用价值。