Home | | Flexible AC Transmission Systems | Co-Ordination of Multiple Controllers using Linear - Control Techniques

Chapter: Flexible Alternating Current Transmission System - Co-Ordination of FACTS Controllers

Study Material, Lecturing Notes, Assignment, Reference, Wiki description explanation, brief detail

Co-Ordination of Multiple Controllers using Linear - Control Techniques

1. Introduction 2. The Basic Procedure for Controller Design

Co-Ordination of Multiple Controllers using Linear – Control Techniques

 

1. Introduction

 

Ø   The term coordination does not imply centralized control; rather, it implies the simultaneous tuning of the controllers to attain an effective, positive improvement of the overall control scheme.

 

Ø   It is understood that each controller relies primarily on measurements of locally available quantities and acts independently on the local FACTS equipment.

 

 

2. The Basic Procedure for Controller Design

 

The controller-design procedure involves the following steps:

1. Derivation of the system model;

2. Enumeration of the system-performance specifications;

3. Selection of the measurement and control signals;

4. Coordination of the controller design; and 

5. Validation of the design and performance evaluation.

 

Step 1: Derivation of System Model

 

Ø   First, a reduced-order nonlinearsystem model must be derived for the original power system and this model should retain the essential steady-state and dynamic characteristics of the power system .

 

 

Ø   Then, the model is linearized around an operating point to make it amenable to the application of linear-control design techniques. If a controllermust be designed for damping electromechanical oscillations, a further reducedlinear model is selected that exhibits the same modal characteristics over the relevant narrow range of frequencies as the original system.

 

Ø   In situations where linearized-system models may not be easily obtainable, identification techniques are employed to derive simple linear models from time-response information.

 

Step 2: Enumeration of the System – performance Specifications

Ø   The damping controller is expected to satisfy the following criteria.

 

1.     It should help the system survive the first few oscillations after a severe system disturbance with an adequate safety margin. This safety factor is usually specified in terms of bus-voltage levels that should not be violated after a disturbance.

 

2. A minimum level of damping must be ensured in the steady state after a disturbance.

 

3. Potentially deleterious interactions with other installed controls should be avoided or minimized.

 

4. Desired objectives over a wide range of system-operating conditions should be met (i.e., it should be robust).

 

Step 3: Selection of the Measurement and Control Signals

 

Ø   The choice of appropriate measurement and control signals is crucial to controller design.

 

Ø   The signals must have high observability and controllability of the relevant modes to be damped, and furthermore, the signals should only minimally affect the other system modes.

 

Ø   The selection of these signals is usually based on system-modal magnitudes, shapes, and sensitivties—all of which can be obtained from small-signal-stability analysis.

  

Step 4: Controller Design and Coordination

 

Ø   The FACTS controller structures are usually chosen from industry practice. Typically, the controller transfer function, Hj(s), of controller j is assumed to be


Ø   This transfer function consists of a gain, a washout stage, and a pth-order leadlag block, as well as low-pass filters. Alternatively, it can be expressed as

 

Ø   Although the basic structure of different controllers is assumed as from the preceding text, the coordination of controllers involves the simultaneous selection of gains and time constants through different techniques.

 

Ø   Doing so permits the system-operating constraints and damping criteria to be satisfied over a wide range of operating conditions.

 

Ø   The coordination techniques may use linearized models of the power system and other embedded equipments, capitalizing on the existing sparsity in system representation.

 

Ø   This model may be further reduced by eliminating certain algebraic variables yet still retaining the essential system behavior in the frequency range of interest.

 

Ø   Eigenvalue analysis–based controller-optimization and -coordination techniques are applicable to power systems typically with a thousand states occurring when full modal analysis must be performed. However, sometimes a limited number of electromechanical modes must be damped; hence the eigenvalue analysis of a selected region can be performed even for relatively larger power systems.

 

Step 5: Validation of the Design and performance Evaluation

 

Ø   Even though the controller design is performed on the simplified system model, the performance of the controller must still be established by using the most detailed system model.

 

Ø   The controller should meet the specifications over a wide range of operating conditions and consider all credible contingencies.This validation is generally performed with nonlinear time-domain


Ø    simulations of the system.


Study Material, Lecturing Notes, Assignment, Reference, Wiki description explanation, brief detail


Copyright © 2018-2020 BrainKart.com; All Rights Reserved. Developed by Therithal info, Chennai.