جستجوی مقالات مرتبط با کلیدواژه « stochastic optimization » در نشریات گروه « فنی و مهندسی »

Stochastic and robust optimizations have been considered as two different views of stochastic problems. While robust optimization takes optimization in the worst case, stochastic optimization regards no conservative view and merely focuses on expected value. However, a unilateral view of stochastic problems does not apply to most real problems. In this article, a hybrid robust and stochastic approach is proposed for optimization problems under uncertainty. Our major contribution is presenting different conservative levels in solving an optimization problem using a Hybrid Robust and Stochastic Optimization approach. To this end, we cluster uncertain parameters into different clusters using Latin Hypercube Sampling and k-Means clustering tools; having established various numbers of clusters of uncertain parameters, different clustering criteria and a Multi-Criteria Decision Making (MCDM) tool is employed to determine the optimal number of clusters of uncertain parameters. Then, a hybrid energy optimization model under uncertainty is applied to coordinate the scheduling of natural gas-fired electricity generation units and gas supply units (gas refinery) under natural gas and electricity demand uncertainty, with known probability distribution and uncertain parameters having different levels of conservatism. The results indicate that while no special trend is evident in the execution time as the number of clusters increases, the optimal value is decreased.

This study aims to develop a capacitated hub location-routing model to design a rapid transit network under uncertainty. The mathematical model is formulated by making decisions about the location of the hub and spoke (non-hub) nodes, the selection of the hub and spoke edges, the allocation of the spoke nodes to the hub nodes, the determination of the hub and spoke lines, the determination of the percentage of satisfied origin-destination demands, and the routing of satisfied demand flows through the lines. Capacity constraints are considered in the hub and spoke nodes and also the hub and spoke edges. Uncertainty is assumed for the demands and transportation costs, represented by a finite set of scenarios. The aim is to maximize the total expected profit, where transfers between the lines are penalized by including their costs in the objective function. The performance of the proposed model is evaluated by computational tests and some managerial insights are also provided through the analysis of the resulting networks under various parameter settings.

In this paper, the carrier selection problem is addressed with the purpose of avoiding shortage regarding vehicle necessity and creating a vehicle rental framework agreement. Achieving in-time delivery of relief supplies to the disaster inflicted areas is of utmost significance in terms of humanitarian relief. We propose a new two-stage stochastic model for determining the pre-disaster and post-disaster decisions in order to delivering relief items to the injured survivors. The pre-disaster phase focuses on determining amount of vehicle in the framework of contracts with suppliers, and deciding an appropriate coverage distance with regards to time and cost. The post-disaster phase aims to respond to the requests made by disaster inflicted areas swiftly and cost-effectively. The proposed MILP model considers a scenario-based approach to handle the uncertainty of demand. The L-shaped algorithm is used to solve this model. A real case study is presented with the aim of demonstrating the efficiency of the model. Moreover, numerical analyses are practiced to illustrate the importance and impact of the cost and the number of vehicle rental contracts in the studied problem. Finally, managerial insights have been presented to assist the relief organization management in making appropriate and efficient decisions.

As global trade flourishes, terminals endeavor to get higher income while adapting to an expanded intricacy concerning terminal administration tasks. Perhaps the most common issues such terminals encounter is the Berth Allocation Problem (BAP), which involves allotting vessels to a bunch of berths and time allotments while at the same time limiting goals, for example, total stay time or total assignment cost. Complex formats of actual terminals present spatial constraints that restrict the mooring and departure of vessels. In spite of the fact that significant research has been carried out with regard to the BAP, these real-world limitations have not been considered in an overall manner. In this paper, a stochastic Multi-Period Berth Allocation-Scheduling Problem in Different Terminals with Irregular Layouts (SBAP) considering multi-Period modes, generalized precedence relations are developed. To solve the (SBAP), a solution approach based on Stochastic Chance Constraint Programming (SCCP) and a solution approach based on Two-Stage Stochastic Linear Programming with Recourse (TSSLPR) is proposed. A mathematical model is solved to show the applicability of the suggested model and solution approach.

In deregulated electricity markets, the electricity consumer should distribute his required electricity optimally between different markets including spots markets with instantaneous price and bilateral contract markets. The present study is aimed to design a model for selecting the optimal electricity market portfolio, so the purchase costs can be minimized by considering a risk level. For this purpose, an optimization approach based on random planning was proposed to minimize costs and reduce power supply risk. Conditional value at risk was used as an appropriate and well-known factor for reducing unfavorable situations in decision-making under uncertain conditions. For simulations, the real information of Iran in 2018 was used as much as possible. Due to the small number of industrial subscribers, the whole population was studied. A genetic algorithm has been used to solve this optimization problem. In addition, MATLAB software was used for implementing the proposed model. The efficiency of the proposed model was proved by analyzing different sensitivities and the best components of the risk-averse decision-making purchasing portfolio in β=5 included from the energy exchange, then from the energy pool, and finally from bilateral contracts.

Stochastic programming is often used to solve optimization problems where parameters are uncertain. In this article, we have proposed a mathematical model for a three-stage transportation problem, where the parameters, namely transport costs, demand, unload capacity and external purchasing costs are uncertain. In order to remove the uncertainty, we have proposed a new transformation technique to reformulate the uncertain model deterministically with the help of Essen inequality. The obtained equivalent deterministic model is nonlinear. Furthermore, we have provided a theorem to ensure that the deterministic model gives a feasible solution. Finally, a numerical example, following uniform random variables, is presented to illustrate the model and methodology.

Metaheuristic optimization algorithms are a relatively new class of optimization algorithms that are widely used for difficult optimization problems in which classic methods cannot be applied and are considered as known and very broad methods for crucial optimization problems. In this study, a new metaheuristic optimization algorithm is presented, the main idea of which is inspired by models in kinematics. This algorithm obtains better results compared to other optimization algorithms in this field and is able to explore new paths in its search for desirable points. Hence, after introducing the projectiles optimization (PRO) algorithm, in the first experiment, it is evaluated by the determined test functions of the IEEE congress on evolutionary computation (CEC) and compared with the known and powerful algorithms of this field. In the second try out, the performance of the PRO algorithm is measured in two practical applications, one for the training of the multi-layer perceptron (MLP) neural networks and the other for pattern recognition by Gaussian mixture modeling (GMM). The results of these comparisons are presented in various tables and figures. Based on the presented results, the accuracy and performance of the PRO algorithm are much higher than other existing methods.

Abstract In this paper, impacts of various uncertainties such as random outages of generating units and transmission lines, forecasting errors of load demand and wind power, in the presence of Demand response (DR) programs on power generation scheduling are studied. The problem is modelled in the form of a two-stage stochastic unit commitment (UC) which by solving it, the optimal solutions of UC as well as DR are obtained. Generating units’ constraint, DR and transmission network limits are included. Here, DR program is considered as ancillary services (AS) operating reserve which is provided by demand response providers (DRPs. In order to implement the existent uncertainties, Monte Carlo (MC) simulation method is applied. In this respect, scenarios representing the stochastic parameters are generated based on Monte Carlo simulation method which uses the normal distribution of the uncertain parameters. Backward technique is used to reduce the number of scenarios. Then, scenario tree is obtained by combining the reduced scenarios of wind power and demand. The stochastic optimization problem is then modelled as a mixed-integer linear program (MILP). The proposed model is applied to two test systems. Simulation results show that the DR improves the system reliability and also reduces the total operating cost of system under uncertainties.

Here, issues connected with characteristic stochastic practices are considered. In the first part, the plausibility of covering the arrangements of an improvement issue on subjective subgraphs is studied. The impulse for this strategy is a state where an advancement issue must be settled as often as possible for discretionary illustrations. Then, a preprocessing stage is considered that would quicken ensuing inquiries by discovering a settled scattered subgraph covering the answer for an arbitrary subgraph with a high likelihood. This outcome is grown to the basic instance of matroids, in addition to advancement issues taking into account the briefest way and resource covering sets. Next, a stochastic improvement model is considered where an answer is sequentially finished by picking an accumulation of “points”. Our crucial idea is the profit of adaptivity, which is investigated for an extraordinary sort of an issue. For the stochastic knapsack issue, the industrious upper and lower cutoff points of the “adaptivity hole” between ideal adaptive and non-adaptive methodologies are checked. Also, an algorithm is described that accomplishes a close ideal estimate. Finally, complicational results are shown to verify the optimal adaptive approaches.

This paper considers a three-stage fixed charge transportation problem regarding stochastic demand and price. The objective of the problem is to maximize the profit for supplying demands. Three kinds of costs are presented here: variable costs that are related to amount of transportation cost between a source and a destination. Fixed charge exists whenever there is a transfer from a source to a destination, and finally, shortage cost that incurs when the manufacturer does not have enough products for supplying customer’s demand. The model is formulated as a mixed integer programming problem and is solved using a multicriteria scenario based solution approach to find the optimal solution. Mean, standard deviation, and coefficient of variation are compared as the acceptable criteria to decide about the best solution.

This paper proposes a stochastic framework for demand response (DR) aggregator to procure DR from customers and sell it to purchasers in the wholesale electricity market. The aggregator assigns fixed DR contracts with customers based on three different load reduction strategies. In the presented problem the uncertainty of market price is considered and the risk of aggregator participation is managed in stochastic optimization problem with CVaR. The feasibility of this problem is studied on a case of Alberta electricity market.