A parallel dual-type algorithm for a class of quadratic programming problems and applications

In this paper, we present a parallel dual-type (PDT) algorithm for solving a strictly convex quadratic programming problem with equality and box constraints. The PDT algorithm is suitable for distributed implementation and can be used as a basic optimization module for handling optimization problems of large distributed systems. Besides, combining the proposed algorithm with a successive quadratic programming (SQP) method, we can solve constrained nonlinear programming problems such as power-system state estimation with power-flow balance constraints on no generation and no-load buses. We have demonstrated the computational efficiency of our method, by comparing with the benchmark commercial NCONF and QPROG routines and the state-of-the-art parallel algorithm through the implementation in the sequential version of Sparc workstation and the parallel version of PC network in solving constrained state estimation problems within IEEE 30-bus and IEEE 118-bus systems.