Font Size: a A A

Researches On Excitation Control Based On Iterative Learning Control Theory

Posted on:2006-04-14Degree:DoctorType:Dissertation
Country:ChinaCandidate:M XuFull Text:PDF
GTID:1118360212967695Subject:Control Science and Engineering
Abstract/Summary:PDF Full Text Request
In the power system, excellent excitation control system can not only guarantee reliability and stability of generator's, but also improve effectively technique and economic characteristic of the generator and its joined power systems. It is good to stability improvement in power system. Abroad uses of the fast excitation control have made the new excitation controller has important meaning. Because of its high nonlinearity, dynamics is changed evidently when running points changed (such as loads are fluctuated in a large scale or a serious failure is happened) in the power system. Here a linear controller cannot satisfy the stabilities, except halting machines, resistance arrester, tossing loads and such urgent control managements as to assure the stabilities. It is necessary to design nonlinear controllers considering nonlinearities of power systems based on nonlinear models of power systems. Based on the mechanism of iterative learning control, it is applied to excitation control with considering strong nonlinearities in the paper. Excitation control is realized with iterative leaning control, so dynamics performances are improved such as immoderate modulation restraint, fast convergence and shorter transition times.An iterative learning control configuration for synchronous machine's excitation in a single machine to infinite system has been built in the paper. The convergence of D and PID type iterative learning laws for excitation control analyzed. Convergent conditions are deduced by rigorous mathematical proving. Characteristics of excitation controllers have been improved with iterative learning, and a stronger ability for maintaining synchronous machine's terminal voltage is obtained, with better adjusting effect, simple realization. They have better convergence and stability comparing to normal PID, without effects to system's stabilities.Iterative learning control laws with variable gains have been presented. PID-PI and PID-PID type learning laws are used for excitation control in a single machine to infinite and two machines system. The terminal voltage of generators can be maintained in a scale of ±5%,and convergence is greatly speeded up, the power angle affluence is increased, which is in favor for stability improvement without effects to synchronous of power systems.Iterative learning control theory is applied to the design of excitation controllers. Memory and modify to u_d(t) by iterative learning control are changed to the ones to u(t).A new fitting method for excitation control is represented. The restriction of perfect tracking in limited time span is removed. The controller designed in this way has simple...
Keywords/Search Tags:iterative learning control, synchronous machine, excitation control, stability, learning law, convergence
PDF Full Text Request
Related items