Design Of Nonlinear Robust Control For Power Systems Based On Minimax Method

Modern power systems are large scale dynamic systems including different kinds of generator sets, multitudinous substations, loads, networks with various voltage grades, and transmission lines in varying mode. With the expansion of the scale of electric net-work, more generators have been employed and the stability problems become promi-nence. The mathematical models of power systems are strongly nonlinear, multidimen-sional and complex equations for the complex dynamic characteristics belonging to the system itself and its various units. Thus, to enhance the operation dynamic performance, the thorough study on applying nonlinear control theory to power system robust control shows high research values and application values.Based on previous works in literature, by taking the nonlinear models of power sys-tems, this dissertation mainly solves the instability caused by sudden changes in the sys-tem structure or parameters.A further study is presented on the transient stability when systems suffer large disturbances. The proposed strategies ensure that the controller keeps satisfactory performance, the negative influences of disturbances are reduced, and the sys-tem is insensitive to effects of large disturbances. The main contributions are as follows:For the nonlinear excitation and steam valve coordinated control problem, a large dis-turbance attenuation control algorithm based on Backstepping method is derived. Focus on the absence of strict feedback conditions, an inverse scheme is provide via adjusting the design steps of virtual feedback control law, accordingly the nonlinear characteristics are completely reserved. The disturbance attenuation is achieved by adopting Minimax method. Ahead of the control law design, a test function is constructed to calculate the critical disturbance which has a dominant influence to the system. The ability of distur-bance rejection of system is improved for taking into account the worst case disturbance.The disturbance attenuation control strategy based on Minimax and Backstepping is extended to n-order nonlinear systems. The failure structure of strict triangular conditions produced in test function is modified by dynamic tuning the variables contained in regular outputs in each inverse step. Meanwhile, a parameter projection mechanism is employed to handle the parameter uncertainties, and the adaptive law forces the parameter estimate within a prior specified interval. The proposed method is applied to turbo generator steam valve whole-range control systems. A comparison simulation study is carried out and the results show that the method has superior performances on enhancing transient stability and improving dynamic quality.An improved disturbance attenuation control method is investigated for nonlinear static var compensator (SVC) system. The class-κ function is introduced during the feedback control law design to improve the transient response of the closed-loop sys-tem without exploiting a large control gain in a long term. A specific nonlinear function is selected to re-parameterize the original uncertain parameter, the estimate of uncertainty is indirectly obtained via the estimators of auxiliary variable. The parameter estimation is forced within a prior specified interval without compromising on control smoothness.Digital control becomes a tendency in industry field due to the rapid development of computer technology. If the continuous-time controller is applied to a sampled-data system directly, the power system may be transferred to unstable state with the increase of sampling time. To avoid the unstable trend, the large disturbance sampling attenuation method based on Minimax and Backstepping is proposed. This method improves the robustness and the insensitivity to large disturbances of the system. The proposed strategy is applied to the nonlinear static synchronous compensator (STATCOM) sampled-data system, and at the same time, the transient performance of system is strengthened.An adaptive disturbance attenuation controller is designed based on Hamiltonian function method for nonlinear thyristor controlled series capacitor(TCSC) with uncer-tainty and external disturbances. And an L2 gain disturbance attenuation algorithm is derived. The restriction of assumption inequality is eliminated by introducing the min-imax method into the dissipative Hamiltonian systems, meanwhile the conservativeness brought by conventional disturbance treatment is significantly reduced. Fully considering the prior information of the uncertain parameter, the parameter projection mechanism is used to design adaptive law to enhance the efficiency of the parameter.Concluding remarks of this dissertation are presented at the end, and some interesting problems deserved to further study are also pointed out.