Systems Engineering - Theory & Practice最新文献

Order production priority is a key issue in production scheduling in make-to-order enterprises. Theory of constraints, linear planning, analysis of hierarchical program (AHP), and entropy weight provided the principles and methods of determining priority from various angles. In view of the deficiencies in the current theories and methods about the confirmation on order production priority, this study constructed a more systematic and comprehensive evaluation index system on it in the supply chain. Furthermore, a new emerging method on order production priority in the supply chain, based on radial basis function neural network model was put forward. After contrasting the effects among three different decision-making methods, evidences in this article indicated that the evaluation method ground on RBF neural network exceeds that on traditional AHP and BP neural network in the supply chain. Moreover, RBF neural network is excellent in producing accurate results within a short time span of training.

Starting from space-time relationship, an evolution model for techno-innovation network structure integrating time process with hierarchical decomposition is constructed, in which a qualitative analysis of inheritance and variation characteristics of the network structure in different observing scales by utilizing variations in time scales is made, and the theory basis of techno-innovation network structure evolution is put out in methodology. The findings are as follows: 1) With variations in observing scale, either for the evolution or for the hierarchy of techno-innovation network structure, there is a relation to some degree of as one falls the other rises between inheritance and variation characteristics. 2) The connections between hierarchies of techno-innovation network structure are in essence relations between the individuals and the collectivity, which are indefinite relations, and the variation is always in existence. 3) The theoretical model can be built through analysis of time series data, which is achieved by adopting rescaled range analysis (R/S analysis).

In a remanufacturing system considering random yield and random demand, the enterprise pays for the used product to end users to control the collection quantity. The collection price is determined before the random yield is realized, and the selling price is determined before the demand randomness is realized. The objective is to maximize the enterprise's utilization, which is the mean-variance function of the profit. An optimization model is addressed to derive the analytical solution of the optimal selling price, given the quantity of remanufactured products. Moreover, the uniqueness of the optimal collection price is proved with consideration of the risk attitude of the remanufacturer. The extensions to the model are developed and the numerical examples are provided to show the influence of the parameters on the system.

On the basis of the distinct characteristics of steel tube production process, the production scheduling problem of seamless steel tubes is described as flexible job-shop scheduling problem. Considering the parallel machines with capacity and speed constraint, maintenance of machines as well as intermediate inventory restriction, it is formulated as mixed-integer-programming(MIP) model to decide the flexible routes for every job and to optimize the sequence of jobs. The objective is not only to reduce delivery delay, but also to minimize idleness of machines and interruption in production. Given the problem is NP-hard (non-deterministic polynomial-time hard), modified genetic algorithm is proposed, whose effectiveness can be well verified in scheduling decision support system for the production of seamless steel tubes in Baoshan Iron & Steel Complex.

In the article, the deficiency of the existing smooth condition is pointed out. In order to overcome the deficiency, we classified the grey models, so that we can define the new smooth degree respectively according to the classification. The concept of the new smooth degree of the first-class grey model is put forward, and the essential characteristics of the sequence in an extreme condition such as the new smooth degree is 1 and 0 are given; furthermore, the new comparison criterion of the new smooth degree is introduced and the consistency of the new comparison criterion of the smooth degree with the prevision criterion is proved.

The article proposed the notion of strongest association rules (SAR), developed a matrix-based algorithm for mining SAR set. As the subset of the whole association rule set, SAR set includes much less rules with the special suitable form for personalized recommendation without information loss. With the SAR set mining algorithm, the transaction database is only scanned for once, the matrix scale becomes smaller and smaller, so that the mining efficiency is improved. Experiments with three data sets show that the number of rules in SAR set in average is only 26.2 percent of the total number of whole association rules, which mitigates the explosion of association rules.

Design for accelerated degradation tests (ADTs) in most literature concentrates on the case of just one stress, but the performance of many highly reliable products are affected by several stresses simultaneously, so it is necessary to study the optimal models for the ADTs with several stresses. This article first introduces a basic modeling framework of degradation path and then proposes an optimal model for ADTs with several stresses based on Wiener process. Under the constraint of total experimental cost, sample size, measurement times, and measurement frequency are obtained by minimizing the asymptotic variance of the estimated 100_p^th percentile of the product's lifetime distribution. Finally, a numerical example is presented to illustrate the proposed model.

An improved Hazard and Operability (HAZOP) modeling and reasoning method is proposed based on the fuzzy information fusion theory, in order to solve practical safety-related problems in industry, such as quantitative information loss and the difficulty of system safety decision-making during the traditional computer-aid HAZOP analysis. A directed graph knowledge model of systematical HAZOP analysis is first developed, taking advantage of different fuzzy quantitative methods corresponding to various variable nodes with different attributes in the system. Based on the D-S evidence theory, the quantitative deviated information between relevant nodes on the path of hazard propagation is used for the fuzzy information reasoning, which is to indicate the fusion reliability of hazard reasons and consequences of system, respectively, and provides an important foundation in safety decision-making. The results of the proposed method applied on the gas turbine compressor system in “Se-Ning-Lan” pipeline system present the possible hazard reasons and consequences according to the current running states with corresponding sorted fusion reliabilities, based on which the appropriate and effective safety maintenance plans are taken into action to avoid accidents. The comparison with traditional HAZOP analysis shows that the proposed method is able to solve the limitation and the uncertainty of traditional HAZOP qualitative analysis and improve the rationality of maintenance decision-making under the existence of multi hazard sources.

Countdown display unit at signalized intersection is one of the most commonly used, but also controversial traffic control equipments. From the perspective of drivers' psychology, the drivers' behavioral decision-making process and its influence factors were analyzed. Based on the dates surveyed, the Logistic model of drivers' behavioral decision-making at signalized intersection with countdown display unit was established, which was also compared with the model without countdown display unit. The sensitivity analysis results of these two models were given at last. The results indicated that the countdown display influenced the drivers' behavioral decision-making significantly. These fundamental works provided some knowledge to traffic management to accelerate the process in the improvement of countdown display setting and gave some advices to promote the safety of intersections.

According to the traditional lower-approximation sets, all of the credible rules cannot be induced in the ordered decision system. In order to solve this problem, the concepts of the increasing and decreasing descriptors are proposed, and the relationships between the support sets of these descriptors and the decision class are discussed. Furthermore, the reductions and relative reductions are investigated in terms of the decision classes of these two new descriptors, and the judgment theorems and discernibility functions with respect to these reductions are given. These methods provide a new theory and practical approach, which can generate all optimal credible decision rules from the ordered decision system.