China
Africa
Explore chapters and articles related to this topic
Container logistics and empty container repositioning
Published in Dong-Ping Song, Container Logistics and Maritime Transport, 2021
Simulation and optimisation have been well developed and applied in almost all subject fields. Simulation-based optimisation combines simulation with optimisation to optimise a set of decisions in a system in which the objective function is difficult to calculate exactly. The idea is to use an optimisation algorithm (e.g. gradient-based method, black-box search method, metaheuristic algorithms) to find the optimal decision variables in an iterative way (Fu 2015). Each iteration involves the simulation model to generate the responses of the current decision variables, i.e. evaluate the performance indicators of the system under the given the decision variables. The iterative procedure will continue searching the trial solutions and return the best solution after the termination criteria are met. An illustration of simulation-based optimisation approach is shown in Figure 4.13.
O
Published in Philip A. Laplante, Comprehensive Dictionary of Electrical Engineering, 2018
where h(x, y) is a point spread function which represents the image intensity response to a point source of light. See modulation transfer function, impulse response function, point spread function. optically inhomogeneous medium a medium whose refractive index randomly varies either in space, in time, or in both. It produces scattering of the light transmitted or reflected at it. optimal decision the best decision, from the point of view of given objectives and available information, that could be taken by the considered decision (control) unit; the term optimal decision is also used in a broader sense -- to denote the best decision that can be worked out by the considered decision mechanism, although this decision mechanism may itself be suboptimal. An example is model-based optimization as a decision mechanism at the upper layer of a two-layer controller for the steady-state process -- the results of this optimization will be referred to as the optimal decisions. optimal sensitivity minimal, in the sense of the H norm, value of the sensitivity function found using H infinity design (H ) techniques. For single-input-single-output systems, the H norm is simply the peak magnitude of the gain of the sensitivity function. In the multi-input-multioutput case, it is expressed by the maximal singular value of the sensitivity matrix function. To meet the design objectives, the optimal sensitivity is found for the sensitivity function multiplied by left and right weighting functions. It enables one to specify minimum bandwidth frequency, allowable tracking error at selected frequencies, the shape of
Decision Procedures
Published in Satya Prakash Yadav, Dharmendra Prasad Mahato, Nguyen Thi Dieu Linh, Distributed Artificial Intelligence, 2020
Optimization is a static as well as a dynamic problem. It includes both optimal control as well as decision theory. Apart from straightforward optimization, these areas have their own special features integrated with them. For instance, in statistical decision theory, the presence of uncertainty is important whereas control theory deals primarily with the dynamic systems. Optimization is used to select the optimal decisions from a set of alternatives when applied to decision-making. For that, it is required to assume the existence of a set of feasible alternatives and a well-defined objective function.
Using aggregated data under time pressure: a mechanism for coping with information overload
Published in Journal of Decision Systems, 2019
A number of strategies explaining the precise manner in which decision makers process information have been identified. These types of decision-making strategies are often broken up into optimal and heuristic or satisficing decision making (Simon, 1957). When using an optimal decision processing strategy, every possible decision outcome must be enumerated to identify and select the optimal solution. Specific strategies have been identified (e.g. additive decision strategy, additive difference decision strategy) where a decision maker places an importance value or weight on each of the attributes associated with a potential outcome (Payne, Bettman, & Johnson, 1988). The attribute weight is then multiplied by the value associated with that attribute to obtain an attribute score for a specific alternative. The importance weight/attribute pairs are added together for each attribute of an alternative generating an overall rating. Using these optimal strategies, the alternative with the highest rating is the ‘best’ decision.
Short- and long-term repeated game behaviours of two parallel supply chains based on government subsidy in the vehicle market
Published in International Journal of Production Research, 2020
Binshuo Bao, Junhai Ma, Mark Goh
In the non-cooperative game model, the duopoly manufacturers’ information is completely asymmetrical, and each of them makes the optimal decision independently to maximally occupy the market share. In the cooperative game model, the duopoly manufacturers completely share the market information with each other. In order to maximise the overall benefit, the optimal decision is made. In the cost-sharing contract game model, the market share of BEVs is gradually increasing due to the government subsidy and the green preference of consumers.
Optimal policy for production systems with two flexible resources and two products
Published in IISE Transactions, 2020
Jianjun Xu, Shaoxiang Chen, Gangshu (George) Cai
Based on Lemmas 2 and 3 and the above analysis, we are now ready to characterize the optimal resource allocation and production policy. Our method is based on the following ideas. First, the six switching curves and three hedging points are determined analytically by cost parameters, resource capacities, and utilization ratios. Some auxiliary curves, line segments, and critical points can be derived accordingly (see Figure 3). Following the two-step decisions we have introduced (even though they are made simultaneously), we decide the production quantities by using R1, depending on the location of the updated inventory levels to the three switching curves of and then decide the production quantities by using R2, depending on the location of the updated inventory levels to the three switching curves of Finally, the coordinate plane of can be segmented into eight regions, depending on the different product–resource combinations of the optimal policy; for example, Ri is the only resource used to produce Pi, R1 is the only resource used to produce both products, or R1 is the only resource used to produce P1 and P2 is not produced (see Figure 3). Detailed results are presented formally in Theorem 1. Based on our method, optimal decisions can be determined analytically. Mathematical definitions of the regions for this case and subsequent cases can be found in Appendix E. The borders of the regions are highlighted in bold (red). Dashed curves are additional curves that are not boundaries, but are important in determining the optimal decisions. Arrow signs are used to describe which resource is used to produce which products. For example, R1 -> P1 & P2 means R1 is used to produce both products. Other arrow signs can be interpreted accordingly.