| Abstract: |
This thesis has presented an assessment to the capability of FACTS devices in controUing and enhancing the electromechanical oscillations in power systems. The study focused on two different types of FACTS devices. First kind is a series device depends on thyristor controlled reactor, namely the thyristor controlled series capacitor (TCSC). The other is one of the most recent developed FACTS devices that depends on voltage source converters, namely the unified power flow controller (lJPFC). The contribution of these two devices, and their damping controllers, in system damping is analyzed from the point of view of damping and synchronizing torque coefficients.Indices have been proposed to detect the oscillatory problem and to determine the suitable location and feedback signal for damping controllers. Artificial intelligence based technique are used to tune FACTS device damping controllers and to coordinate between them. Genetic algorithm is used to maximize damping ratio of system mechanical modes by searching in the controllers’ parameter space for the optimum values. Neuro-Fuzzy network is trained to adapt the parameters of each controller, so that maximum damping is obtained over wide range of operating condition.1- FACTS devices do not provide new transmission capability but they certainly remove constraints and enable improved use of existing transmission network, and are capable of fast control.2- This thesis introduced a detailed analysis of TCSC and UPFC models and damping controllers, and how these models can be linked to the dynamic models of both single machine infinite bus and multi-machine power systems.3- It has been shown that tuning of various controllers gains have to be carried out judiciously. Controller gains that may be acceptable for certain operating condition may have negative effect on system damping for other opending conditions.4- This thesis demonstrated that TCSC and UPFC controllers, appropriately tuned and placed make them a valuable alternative to the traditional PSS controller. However, an over all co~1 benefit analysis would be necessary for a specific application for oscillation control combined with other application aspects, such as voltage control and power flow control, given their high relatively costs of FACTS• controllers when compared to PSS controllers.5- Damping torque analysis has been a powerful technique in studying power system oscillation for decades. Besides damping torque analysis, synchronizing torque coefficient due to FACTS devises and their damping controllers is also investigated. It has been shown that, UPFC affects the synchronizing torque coefficient more than TCSC, which is supported by the frequency of system oscillation for the same disturbance discussed in chapt~r 7.6- The increasing trend in the use of FACTS controllers in power system could lead to some undesirable interaction between FACTS and power system controllers or among FACTS controllers. This problem is thought as an optimization problem in w.!Uch it is required to seek for the appropriate controllers’ parameters that work together to maximize mechanical modes damping ratio. To solve this problem, conventional GA is used, GA seeks for the best parameters combination for each operating point of the power system. The problem faced when dealing with GA is the searching time, to over come this problem, GA is used ofT-line to generate the suitable parameters of damping controllers for each operating point. These data are feed to ANFISs which work on-line, so that searching time is not a matter of concern since GA itself is not working on-line.7- A novel hybrid teclmique based on ANFIS is proposed to adapt FACTS devices damping controllers’ gains to improve the damping characteristics of power systems over a wide range of operating conditions. The proposed ANFISs were trained based on real-time measurements of local signals available at FACTS devices location such as line power and current. Damping controllers’ gains can be determined by the ANFISs, which makes the proposed stabilizcr relatively simple and suitable for practical on-line implementation.8- FACTS device, either TCR-based or VSC-based, should be placed between areas that participate in system critical modes.9- The robustness of the feedback control signal for various operating conditions is examined. Line power, for multi-machine system, is found to be the robust feedback control signal to damping controllers for both lCSe and UPFC compared to other local signals such as bus voltage and line current and reactive power.10- Simulation results show that the proposed genetic based ANFISs damping controllers is very effective means for improving the dynamic performance of the power system compared to fixed-gains controllers.1- Network equations required to link lCSC and UJ>FC dynamic models to power system are presented for single-machine infinite bus and for multi-machine power systems.2- The problem of linel1f controller in power systems in Lbat they don not curry power. system nonlinearity is solved by introducing each operating point of thesystem to ANFISs. This solution makes the damping controllers adaptive without computation burden, which is the maIn disadvantage ofnonJinear controllers.4- Previously, GA was used to maximize modes damping ratio considering all working range of power system at a time. In this case, controllers’ parameters are optimal if we consider the whole operating conditions, but are not for each operating point. In our thesis, for every operating point GA is used to search for controllers’ parameters, parameters found at certain operating point are different than those found at others. Much effort is preferred to be done off line so that’”controllers’ behaviour will not deteriorate when operating on line. ANFISs arerequircd in this case to recognizc the appropriatc parameters for each operating point.5- Evaluation of available local feedback signals to FACTS devices damping controllers is done not only for the normal operating point but also for different operating points.6- Using ANFISs for each danlping controller makes them decentralized controllers and the problem 0 f signal communication is eliminated.From the results achieved throughout the work of this thesis, the following points can be investigated.1- The study can be extended to examine ability of FACTS devices with magnetic energy storage to improve power systems transient stability.2- Economical assessment should be applied to investigate the benefits of FACTS devices versus other options such as network reinforcement.3- Ability of different types of FACTS device on voltage control and stability should be carried out.
|
|
|