Search

This paper introduces a novel bi-objective routing model based on Artificial Intelligence of Things (AIoT) principles. Our model not only aims to minimize vehicle transportation costs and prevent time window violations but also endeavors to mitigate environmental pollutants. This study addresses the complex challenge of optimizing routes for heterogeneous vehicle fleets usingAIoT technology. Analyzing the bi-objective model using AI tools (MOSCA and NSGA II), we unveil a fascinating trade-off: as energy consumption decreases, system costs increase. Employing robust optimization techniques, we validate the model's performance under pessimistic conditions characterized by rising uncertainty rates. Notably, heightened uncertainty correlates with increased objective function values. Through a series of diverse test cases, we observe that MOSCA demonstrates superior efficiency, notably outperforming in NP, MD, and T indices. Our findings offer valuable insights for practitioners, policymakers, and researchers in the domains of transportation optimization, AIoT, and environmental sustainability. [ABSTRACT FROM AUTHOR]

Proposing a robust designed facility location is one of the most effective ways to hedge against unexpected disruptions and failures in a transportation network system. This paper considers the combined facility location/network design problem with regard to transportation link disruptions and develops a mixed integer linear programming formulation to model it. With respect to the probability of link disruptions, the objective function of the model minimizes the total costs, including location costs, link construction costs and also the expected transportation costs. An efficient hybrid algorithm based on LP relaxation and variable neighborhood search metaheuristic is developed in order to solve the mathematical model. Numerical results demonstrate that the proposed hybrid algorithm has suitable efficiency in terms of duration of solution time and determining excellent solution quality. [ABSTRACT FROM AUTHOR]

Uncertainties of renewable power generation, load, and electricity price imposes serious challenges in optimal scheduling of microgrids (MGs). To cope with this issue, advanced models should be developed to optimize scheduling problem of MGs under uncertainties. However, the optimality and robustness of solutions as well as the computational burden are still a matter of concern when applied to scheduling of a practical MG. To provide a model which is adjustably robust and computationally tractable and also financial effective at the same time, this paper proposes a novel uncertainty handling framework based on stochastic p -robust optimization method for day-ahead scheduling of MGs including renewable and non-renewable distributed generations, combined heat and power systems, energy-storage systems, as well as demand response providers. The objective is to find optimal scheduling of MG in a way that the expected profit is maximized while the regret in each scenario is bounded. The effectiveness of the proposed model is tested on a MG test system thorough different case studies. Simulation results demonstrate that the robustness of the scheduling results is enhanced significantly without substantial decrease in the expected profit of MG. • A comprehensive day-ahead energy management model is developed for CHP-based microgrids. • The model is transformed to an MILP one to ensure near-optimal solutions. • All uncertainties including market price, load and renewable intermittencies are modeled. • A new stochastic p-robust optimization framework is proposed to properly cover uncertainty risks. [ABSTRACT FROM AUTHOR]