用不确定性度量作战效能的评估方法

把作战过程分解成多个节点,用不确定性自信息量度量各节点完成任务的程度,并改造了信息熵函数,得出一种适合求解作战效能的评估方法.最后,针对一实例进行效能评估建模、验模和灵敏度分析.     

Data‐driven variation source identification for manufacturing process using the eigenspace comparison method

Variation reduction of manufacturing processes is an essential objective of process quality improvement. It is highly desirable to develop a methodology of variation source identification that helps quickly identify the variation sources, hence leading to quality improvement and cost reduction in manufacturing systems. This paper presents a variation source identification method based on the analysis of the covariance matrix of process quality measurements. The identification procedure utilizes the fact that the eigenspace of the quality measurement covariance matrix can be decomposed into a subspace due to variation sources and a subspace purely due to system noise. The former subspaces for different samples will be the same if the same variation sources dominate. A testing procedure is presented, which can determine the closeness of the subspaces under sampling uncertainty. A case study is conducted to illustrate the effectiveness of this methodology. © 2006 Wiley Periodicals, Inc. Naval Research Logistics, 2006.     

基于ICA的多姿态人脸表示

将独立成分分析(ICA)应用于多姿态人脸识别。对比分析了ICA和主成分分析(PCA)两种人脸识别方法的差异,并重点研究了多姿态人脸的独立成分(IC)表示。在基于权向量幅值的方法基础上,引入了基于比例因子的IC核选择的新方法。实验表明,新方法有利于提高识别的准确率和识别的效率。     

基于LMD频域近似熵频谱感知算法研究

针对现有频域近似熵频谱感知技术在低信噪比条件下抗噪声性能和检测性能有待提升的问题,提出了一种基于LMD频域近似熵的频谱感知算法。(1)算法筛选出3个PF分量累加求和,使得算法提取局部调频包络特征信息得到最优,进一步排除噪声不确定度的影响。(2)算法对累加PF分量进行频域变换后求其近似熵,增强算法对频域信息的嗅探能力,提升算法检测性能。Monte Carlo仿真结果表明,在噪声不确定度为0dB,采样点数为8 000的情况下,当信噪比大于-19 d B时,可以对2ASK信号达到100%的检测概率,与现有频域近似熵算法相比,检测性能约有17 d B的提升。     

Bullwhip and reverse bullwhip effects under the rationing game

When an unreliable supplier serves multiple retailers, the retailers may compete with each other by inflating their order quantities in order to obtain their desired allocation from the supplier, a behavior known as the rationing game. We introduce capacity information sharing and a capacity reservation mechanism in the rationing game and show that a Nash equilibrium always exists. Moreover, we provide conditions guaranteeing the existence of the reverse bullwhip effect upstream, a consequence of the disruption caused by the supplier. In contrast, we also provide conditions under which the bullwhip effect does not exist. In addition, we show that a smaller unit reservation payment leads to more bullwhip and reverse bullwhip effects, while a large unit underage cost results in a more severe bullwhip effect. © 2017 Wiley Periodicals, Inc. Naval Research Logistics 64: 203–216, 2017     

基于改进差分进化算法的内弹道参数优化

火炮内弹道参数直接影响其计算结果,内弹道参数优化设计是内弹道分析计算的重要环节。基于经典内弹道计算模型,以火药内弹道计算中最为敏感的燃速指数及其燃速系数为设计参量,以最大膛压及弹丸初速的计算值与测试值一致性为目标,对差分进化算法群体的初始化、变异因子及变异模式进行了改进,对某型火炮的内弹道参数进行了优化计算。优化设计结果表明改进后的差分进化算法满足内弹道设计工程实践要求,为内弹道参数优化设计提供了一种有效的算法。     

A note on “resource flexibility with responsive pricing”

We revisit the capacity investment decision problem studied in the article “Resource Flexibility with Responsive Pricing” by Chod and Rudi [Operations Research 53, (2005) 532–548]. A monopolist firm producing two dependent (substitutable or complementary) products needs to determine the capacity of one flexible resource under demand risk so as to maximize its expected profit. Product demands are linear functions of the prices of both products, and the market potentials are random and correlated. We perform a comparative statics analysis on how demand variability and correlation impact the optimal capacity and the resulting expected profit. In particular, C&R study this problem under the following assumptions/approximations: (i) demand intercepts follow a bivariate Normal distribution; (ii) demand uncertainty is of an additive form; (iii) and under approximate expressions for the optimal capacity and optimal expected profit. We revisit Propositions 2, 3, 4, 5, and 10 of C&R without these assumptions and approximations, and show that these results continue to hold (i) for the exact expressions for the optimal expected profit and optimal capacity, and (ii) under any arbitrary continuous distribution of demand intercepts. However, we also show that the additive demand uncertainty is a critical assumption for the C&R results to hold. In particular, we provide a case of multiplicative uncertainty under which the C&R results (Propositions 2 and 3) fail. © 2010 Wiley Periodicals, Inc. Naval Research Logistics 2010     

Inventory cost impact of order processing priorities based on demand uncertainty

We evaluate an approach to decrease inventory costs at retail inventory locations that share a production facility. The retail locations sell the same product but differ in the variance of retail demand. Inventory policies at retail locations generate replenishment orders for the production facility. The production facility carries no finished goods inventory. Thus, production lead time for an order is the sojourn time in a single server queueing system. This lead time affects inventory costs at retail locations. We examine the impact of moving from a First Come First Served (FCFS) production rule for orders arriving at the production facility to a rule in which we provide non‐preemptive priority (PR) to orders from retail locations with higher demand uncertainty. We provide three approximations for the ratio of inventory costs under PR and FCFS and use them to identify conditions under which PR decreases retail inventory costs over FCFS. We then use a Direct Approach to establish conditions when PR decreases retail inventory costs over FCFS. We extend the results to orders from locations that differ in the mean and variance of demand uncertainty. The analysis suggests that tailoring lead times to product demand characteristics may decrease system inventory costs. © 2002 Wiley Periodicals, Inc. Naval Research Logistics 49: 376–390, 2002; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/nav.10016     

低谐波注入的大功率三相整流电路

给出了三相桥式PWM整流电路主电路 ,分析了其实现单位输入功率因数和不对电网注入谐波的工作原理。在此基础上 ,提出了一种新型大功率三相PWM整流电路系统结构 ,其主电路采用了双模块并联的方式。与此同时 ,还探讨了其控制方法和相应的控制系统以及IGBT的驱动保护电路等。最后对所研制的容量为 10 0kW的实验样机进行了实验研究 ,结果证实了它的良好性能     

Dynamic pricing and inventory control with learning

Optimal operating policies and corresponding managerial insight are developed for the decision problem of coordinating supply and demand when (i) both supply and demand can be influenced by the decision maker and (ii) learning is pursued. In particular, we determine optimal stocking and pricing policies over time when a given market parameter of the demand process, though fixed, initially is unknown. Because of the initially unknown market parameter, the decision maker begins the problem horizon with a subjective probability distribution associated with demand. Learning occurs as the firm monitors the market's response to its decisions and then updates its characterization of the demand function. Of primary interest is the effect of censored data since a firm's observations often are restricted to sales. We find that the first‐period optimal selling price increases with the length of the problem horizon. However, for a given problem horizon, prices can rise or fall over time, depending on how the scale parameter influences demand. Further results include the characterization of the optimal stocking quantity decision and a computationally viable algorithm. © 2002 Wiley Periodicals, Inc. Naval Research Logistics 49: 303–325, 2002; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/nav.10013