To enhance industrial park's economic gains and effectively allocate its electricity bill among industrial users with combined heat and power (CHP) units and photovoltaic (PV) panels, this paper proposes a distrib...
详细信息
To enhance industrial park's economic gains and effectively allocate its electricity bill among industrial users with combined heat and power (CHP) units and photovoltaic (PV) panels, this paper proposes a distribution locational marginal price (DLMP)-based bi-level demand management approach. The upper level optimizes dispatching decisions of industrial users with the objective of minimizing their energy bills, and the lower level is a DLMP-based market clearing problem to minimize the two-part tariff cost of the industrial park operator. In order to solve the proposed bi-level model efficiently, it is first equivalently converted into a single-level mathematical programming with equilibrium constraints (MPEC), and then reformulated as a mixed-integer second-order conic programming (MISOCP) model by linearizing bilinear terms. Numerical results demonstrate the effectiveness of our proposed bi-level method in lowering industrial park's electricity bill and achieving effective allocation among users.
The problem of reconfiguration for active distribution systems is formulated as a stochastic mixed-integer second-order conic programming (MISOCP) model that simultaneously considers the minimization of energy power l...
详细信息
ISBN:
(纸本)9781665436137
The problem of reconfiguration for active distribution systems is formulated as a stochastic mixed-integer second-order conic programming (MISOCP) model that simultaneously considers the minimization of energy power losses and CO2 emissions. The solution of the model determines the optimal radial topology, the operation of switchable capacitor banks, and the operation of dispatchable and non - dispatchable distributed generators. A stochastic scenario-based model is considered to handle uncertainties in load behavior, solar irradiation, and energy prices. The optimal solution of this model can be reached with a commercial solver;however, this is not computationally efficient. To tackle this issue a novel methodology which explores the efficiency of classical optimization techniques and heuristic based on neighborhood structures, referred as matheuristic algorithm is proposed. In this algorithm. the neighborhood search is carried out using the solution of reduced MISOCP models that are obtained from the original formulation of the problem. Numerical experiments are performed using several systems to compare the performance of the proposed matheuristic against the direct solution by the commercial solver CPLEX. Results demonstrate the superiority of the proposed methodology solving the problem for large-scale systems.
This paper presents a mixed-integer second-order conic programming (MISOCP) model to solve the reconfiguration problem of electrical distribution systems, considering the simultaneous minimization of total active powe...
详细信息
This paper presents a mixed-integer second-order conic programming (MISOCP) model to solve the reconfiguration problem of electrical distribution systems, considering the simultaneous minimization of total active power losses and improvement of customer-oriented reliability indices. The reliability indices considered in this paper are the system average interruption frequency index (SAIFI), the system average interruption duration index (SAIDI), and the energy not supplied (ENS). Under radiality, the proposed model satisfies the operational constraints of the reconfiguration problem, i.e., the voltage magnitude limits of the nodes and the current capacities of the conductors are not violated. The use of an MISOCP model guarantees convergence to optimality via convex optimization software tools. A multi-objective optimization approach is used to generate a full Pareto front surface that shows the conflict between the active power loss minimization and the improvement of the reliability indices in the reconfiguration problem. Finally, in order to test and verify the proposed methodology, a 43-node test system and a real 136-node system were employed. (C) 2015 Elsevier Ltd. All rights reserved.
暂无评论