Constrained Consensus-based Iterative Algorithm for Economic Dispatch in Power Systems

This paper considers the distributed economic dispatch problem in power systems. A novel constrained consensus-based iterative algorithm is proposed to cooperatively search the optimal incremental cost. The algorithm is carried out by alternately executing a continuous-time finite-time consensus algorithm and employing local projection operations only when consensus variables reach agreement. At each iteration, the initial values of consensus variables are restored locally based on the deviation between the consensus value in the last iteration and its projection on local constraints. Compared to existing methods, our method only needs a single consensus algorithm and does not require continuous projection operations. Thus each node in our algorithm only transmits a single variable to neighbours and the search time of the optimal solution is significantly reduced. Convergence analysis is given, and numerical examples are presented to show the effectiveness of our method with comparison to some existing results.