Volume 3, Issue 6

Domino effect is a fairly common phenomenon in process industry accidents, which makes many process industry accidents serious and the consequent losses enhanced. Domino effect of the major accidents in chemical cluster is emphasized. Many researchers have studied domino effect in chemical clusters from different perspectives. In the review, we summarize the research from three aspects: The statistical analysis of domino accidents in chemical process industry, the evaluation of domino accidents and the prevention of domino accidents in chemical clusters by game theory. From the analysis, we can find the characteristic of domino accidents such as the time and the location, the origin and causes of domino accidents. The methods of assessing domino effects such as quantitative risk assessment (QRA), Bayesian networks (BN) and Monte Carlo simulation (MCS) are analyzed. The prevention of domino accidents in chemical clusters using game theory is seldom, and there is still much space for improvement in enterprises’ efforts to prevent risk of domino accidents.

This paper compares differences between single regional input-output (SRIO) model and global interregional IO model (GIRIO) used in the area of carbon emission embodied in exports under global supply chain, and decomposes carbon emissions and carbon intensity in exports according to export path based on KWW (2014). Empirical analysis in China’s exports in 1995, 2000, 2005 and 2009 shows: 1) GIRIO model yielded about 2.11% less domestic emissions absorbed abroad, 0.8% more domestic emissions in exports, 1.5 times more foreign emissions in exports in 2009, compared to SRIO model. 2) USA and EU absorb most carbon emissions in China’s exports, but with declining share. Increasing domestic emissions firstly exported to developing countries but finally returned home and emissions from developing countries in China’s exports show their contribution from developing countries. Discrepancies between bilateral carbon emissions flows changes and direct trade flows changes suggest indirect carbon emissions transfer by third parties are neglected in SRIO model. 3) Despite dramatic decrease, China still bears 1.52 times more carbon emissions per GDP gained than foreign countries in fragmentation of China’s exports production in 2009, indicating different roles in production fragmentation.

This paper studies a dual-channel supply chain in which a manufacturer sells products to a retailer as well as to customers who are sensitive to both channel price and the retail service. Three game models (Manufacturer Stackelberg, Retailer Stackelberg and Vertical Nash) are built according to members’ different bargaining power in a dual channels system. The authors show that customers can receive lower channel price and higher retail service level when channel members have equal bargaining power, however, when the retailer occupies the market leadership, consumers always receive the least welfare because of the higher channel price and lower retail service. Interestingly, the retailer can take advantage of market leadership to make more profits, while the manufacturer is more willing to give up its power and act as a Stackelberg follower. Furthermore, Manufacturer Stackelberg and Vertical Nash is a strictly dominated strategy for the retailer and the manufacturer respectively.

This paper considers the uniform parallel machine scheduling problem with controllable delivery times, which assumes that the delivery times of jobs are linear decreasing functions of the consumed resource. It aims to minimize the maximum completion time under the constraint that the total resource consumption does not exceed a given limit. For this NP-hard problem, we propose a resource allocation algorithm, named RAA, according to the feasible solution of the uniform parallel machine scheduling problem with fixed delivery times. It proves that RAA algorithm can obtain the optimal resource allocation scheme for any given scheduling scheme in O ( n log n )time. Some algorithms based on heuristic algorithm LDT, heuristic algorithm LPDT and simulated annealing are proposed to solve the uniform parallel machine scheduling problem with controllable delivery times. The accuracy and efficiency of the proposed algorithms are tested based on those data with problem sizes varying from 40 to 200 jobs and 2 to 8 machines. The computational results indicate that the SA approach is promising and capable of solving large-scale problems in a reasonable time.

The conventional data envelopment analysis (DEA) measures the relative efficiency of decision making units (DMUs) consuming multiple inputs to produce multiple outputs under the assumption that all the data are exact. In the real world, however, it is possible to obtain interval data rather than exact data because of various limitations, such as statistical errors and incomplete information, et al. To overcome those limitations, researchers have proposed kinds of approaches dealing with interval DEA, which either use traditional DEA models by transforming interval data into exact data or get an efficiency interval by using the bound of interval data. In contrast to the traditional approaches above, the paper deals with interval DEA by combining traditional DEA models with error propagation and entropy, uses idea of the modified cross efficiency to get the ultimate cross efficiency of DMUs in the form of error distribution and ranks DMUs using the calculated ultimate cross efficiency by directional distance index. At last we illustrate the feasibility and effectiveness of the proposed method by applying it to measure energy efficiency of regions in China considering environmental factors.

A data-centric approach is proposed to facilitate the design and analysis of challenging complex systems and address the problems of currently existing model-based systems engineering (MBSE) methodologies. More specifically, based on three core steps of current MBSE methodologies, a high-level data meta-model, depicting the semantic relationships of high-level data concepts, is first presented to guide the data modeling for systems engineering (SE). Next, with respect to the six high-level data concepts, the data elements are collected as the modeling primitives to construct static and/or executable models, which can also act as a common and consistent data dictionary for SE. Then, the mapping associations amongst core data elements are established to associate the model elements in different steps and achieve the requirement traceability matrix. Finally, the feasibility of the proposed approach is demonstrated with an illustrative example.

This paper is concerned with testing for the varying coefficient spatial autoregressive models. Based on the profile likelihood estimation procedure, a profile generalized likelihood ratio test procedure is proposed to test spatial lag effects, and a residual-based bootstrap procedure is used to derive the p-value of the test. Some simulations are conducted to assess the performance of the test and the results are satisfactory.

In terms of distance function and spatial continuity in Voronoi diagram, a generic generating method of Voronoi diagram, named statistical Voronoi diagram, is proposed in this paper based upon statistics with mean vector and covariance matrix. Besides, in order to make good on the discreteness of spatial Voronoi cell, the cross Voronoi cell accomplished the discrete ranges in its continuous domain. In the light of Mahalanobis distance, not only ordinary Voronoi and weighted Voronoi are implemented, but also the theory of Voronoi diagram is improved further. Last but not least, through Gaussian distribution on spatial data, the validation and soundness of this method are proofed by empirical results.