الفهرس 
Power System Stability
Practical issues of wide-area damping control designOnly 14 pages are availabe for public view
 |  | from | 316 |  |  |
| | | from | 316 | | |
Abstract
Electromechanical oscillations have been noticed in power systems once the generators are connected to provide greater power capacity and supply reliability. These oscillations are reflected in the generator mechanical axes relative motions and coupled by voltage and power oscillations. Some of large-scale electric power systems characteristics, such as highly variable generation patterns and heavy loading, weak transmission, tend to increase the probability of wide-area electromechanical oscillations appearance. Therefore, the wide-area controller is used here to emphasize the possible co-existence of inter-area oscillations and local modes of different frequencies that might appear simultaneously in different parts of the power systems. Such oscillations endanger the secure operation of power systems and if not efficiently controlled can lead to line tripping, generator outages, and even large-scale blackouts. In general, the electromechanical oscillations are caused by insufficient system damping. Due to the limited availability of the current power-system network-structure, most of improvement for electromechanical oscillations damping focus on setting different controllers such as Power System Stabilizers (PSSs), Thyristor Controlled Series Compensators (TCSC), and so on. These damping controllers usually use local measurements at their inputs, and their control parameters are determined in off-line. However, increasing uncertainties brought by the growing complexity resulting from new power flow control devices and renewable generation, make the damping effects of these controllers become questionable. Moreover, the controllers installed into different areas are to be further coordinated to attain better global damping performances. Therefore, it still needs for more efficient and more widely coordinated damping control. For this purpose, this thesis proposes a supplementary damping control. It treats damping control as a multi-step optimization control problem with discrete dynamics and costs, which calculates supplementary input signals for existing controllers using learning based methods. At each control time, the controller collects current system states, solves the optimal control problem and superimposes the calculated supplementary signals on the outputs of the controller. With the help of these supplementary signals, it forces the controlled system response to adapt the desired trajectory. These supplementary signals allow adaptively adjusting and coordinating a subset or all of the existing damping controllers. In this thesis, In order to carry out on-line assessment of interarea oscillation, data from Phasor Measurement Units (PMUs) is utilized. Intelligent algorithms for optimal location of PMUs are developed. Based on PMU information, model-free learning methods Reinforcement Learning (RL) with fuzzy logic algorithm is applied to coordinate Flexible Alternating Current Transmission Lines (FACTS) devices and PSSs. The location of PMU and FACTS allocation are chosen to achieve observability and controllability of the power system. The scheme of the RL with fuzzy logic is investigated.