Research On Methodology Of Preventive Control With Transient Stability Constraints For Power System

TSCOPF(Transient Stability Constrained Optimal Power Flow) is the core problem of prevention and control of power system, which is a large-scale nonlinear and complicated optimization problem and can connect the economic and safety of power plant. This paper establishes a mathematical model considering transient stability constraints in prevention and control, and study the intelligent optimization algorithm and optimal control theory in the application of the prevention of the power system.The dissertation firstly studies an cooperative micro-particle swarm optimization in high-dimensional complex problems. The partition optimization and co-evolution is introduced into the standard particle swarm optimization algorithm, which can effectively solving high dimension optimization problems. And it has better global optimal search capability, the optimization effect is better than that of the same size and mass of particle swarm optimization algorithm. Besides, the paper establish stability criterion according to the center of inertia angle and perfect the mathematical model of the prevention and control problem. The application of particle swarm algorithm solve the preventive control problem, the characteristic of the algorithm is required to evolve the particle number less, good optimization effect, the simulation examples show that the particle swarm algorithm proposed has better optimization effect.In addition, this paper made a research on the application of optimal control theory in the prevention problem. First, according to the optimal control theory, an effective approach, to calculate the gradients of transient stability constraints with respect to all control variables, is proposed by simultaneously computing generator dynamic equations and co-state equations. Besides, we set the power angle of generator initial values as the control variables, the gradient on the transient stability constraints are re-derived,the derivation process is effectively simplified computation of the gradient, after the gradient obtained by solving a nonlinear optimization problem,the generator output is revise, we validate the effectiveness of algorithm through an example.Some of the problems that should be researched on are proposed at the end of the paper, and these are the research direction for the future work.