Power System Operation And Control

The Problem of power system operation and control is studied in this thesis. The main work consists of two parts: First, on the problem of power system operation, including: 1. Critical Clearing Time (CCT) assessment based on Functional Link Net (FLN) ;2. Artificial Neural Net (ANN) applied to estimation for short-circuit characters in three-phase power system, such as impact current, equivalent coefficient and thermal stress etc; 3. A hybrid ANN-Lagrangian relaxation approach for generating unit commitment , and for the disadvantages of BP algorithm, a new conjugate gradient algorithm without line searches is used to speed up convergemce. Secondly, including. 1. High level FLN and general inversion (GI) algorithm are used to adjust the parameters of power system stabilizer (PSS) in realtime; 2. The inverse system method for the control of fast valving and nonlinear excitation , and ANN is compensated for the uncertainity of system parameters ; 3. Block Pulse Function (BPF) is applied in analysis and synthesis of optimal control for PWM systems, and a piecewise constant solution is obtained; 4. Study of the method for designing intelligent robust regulator for high-precision and tremely-law-speed system, and to overcome low-speed jerking motion effectively, trace ramp input signal rapidly, eliminate steady-state error fastly. In addation, an A/D converter using feedforward nearal network model is proposed, and a coupled ANN-ES scheme for power system security assessment and preventive control is also proposed.