Performance Improvement Of Probabilistic PSSs And Order Reduction For Special Oscillation Modes

In power system operation, there are many fluctuations and random factors, such as the variation of load powers and generations, changes in network configuration and system parameters, as well as the measure and forecast errors. In order to improve the stability of power systems in more wide operating condition, the probabilistic technique has been adopted to eigenvalue analysis and power system stabilizer (PSS) design. Based on the present probabilistic analysis methods, an auxiliary control unit is formed with control parameter reasonably determined to improve the PSS robust performance; the order reduction of large power systems for special oscillation modes is presented, and is applied to probabilistic condition.For the purpose of studying large power systems, the part eigenvalue analysis technique and order reduction method are applied to the small signal stability studies. When the eigenvalue analysis is applied to large power systems, the main problem is the calculative precision and running time. In practical application, the number of critical eigenvalues is limited, order reduction can be applied to each critical mode, which will relieve the contradiction between the calculative precision and running time. The order reduction of large power systems for special oscillation modes is presented in this thesis. The coefficient matrix is rearranged by eigenvalue sensitivity values, and reduced in dimension with only the strong correlation generators reserved. Since the reduced system matrix is usually in complex form, the imaginary part is regarded as the correction for real part. With the value of the real part used as the initial value, the accurate computation is carried out by the improved Rayleigh quotient inverse iteration. In the previous Rayleigh quotient inverse iteration method, the left and right eigenvectors are simultaneously iterated, the convergence of this method is not satisfactory for the cases in this thesis. Therefore, the improved form is used, which makes simpler in iteration structure, and better in convergence performance. The effectiveness of this approach is tested on two examples.Continuously, the order reduction of special oscillation mode is introduced to probabilistic eigenvalue analysis, that is, the order reduction model is obtained based on the probabilistic theory, and used to more operating conditions, which provides...