Distributed Piecewise Approximation Economic Dispatch for Regional Power Systems Under Non-Ideal Communication

• Published in: IEEE access 2019, Vol.7, p.45533-45543
• Main Authors: Hao, Ran, Lu, Tianguang, Wu, Qiuwei, Chen, Xinyu, Ai, Qian
• Format: Article
• Language: English
• Subjects: Algorithms; Approximation; Communication; Communication uncertainties; distributed economic dispatch; Economics; Electric power distribution; Generators; Mathematical analysis; on/off status; Optimization; piecewise approximation; Power dispatch; Power systems; Privacy; Protocols; Regional development; Topology; Unit commitment
• ISSN: 2169-3536; 2169-3536
• DOI: 10.1109/ACCESS.2019.2908680

Appropriate distributed economic dispatch (DED) strategies are of great importance to manage wide-area controllable generators in wide-area regional power systems. Compared with existing works related to ED, where dispatch algorithms are carried out by a centralized controller, a practical DED scheme is proposed in this paper to achieve the optimal dispatch by appropriately allocating the load to generation units while guaranteeing consensus among incremental costs. The ED problem is decoupled into several parallel sub-problems by the primal-dual principle to address the computational issue of satisfying power balance between the demand and the supply from the distributed regional power system. The feasibility test and an innovative mechanism for unit commitment are then designed to handle extreme operation situations, such as low load level and surplus generation. In the designed mechanism, the on/off status of units is determined in a fully distributed framework, which is solved using the piecewise approximation method and the discrete consensus algorithm. In the algorithm, the push-sum protocol is proposed to increase the system adaptation on the time-varying communication topology. Moreover, consensus gain functions are designed to ensure the performance of the proposed DED under communication noise. Case studies on a standard IEEE 30-bus system demonstrate the effectiveness of our proposed methodology.