The application of multi-agent technology on transient stability assessment of Iraqi super grid network

INTRODUCTION

For reliable service, a bulk electricity system must remain intact and be capable of withstanding a wide variety of disturbances. Therefore, it is essential that the system be designed and operated so that the more probable contingencies can be sustained with no loss of load (except that connected to the faulted element). So, the most adverse possible contingencies must not result in uncontrolled, widespread and cascading power interruptions (1). In the past, when systems were smaller and less complicated informal methods of security, analysis and control, were performed by system operators based upon experience and knowledge of system. Modern power systems are quite large and more extensively interconnected making the task of security analysis and control difficult for the system operator (2). In case where transient stability is an issue, the conventional methods of stability analysis by a time domain iterative process are too far slow for online operation. This led researchers to explore fast direct methods to analyze transient stability of electric power systems (3). A direct method for transient stability analysis is defined as a method that is able to determine stability without explicitly integrating differential equations describing the post-fault system (4). Among different methods proposed, the standard Energy Approach has received a great deal of attention.

In this research a prediction agent is designed to predict the stability of the network under consideration during the abnormal conditions by applying the PEBS method for transient stability. The predicted values are linked to a control agent that will apply the control action to stabilize the system. This control action is carried out by increasing the generating power of the generating unit in the network and analyze the network by a load flow program then fed the new situation of power flow to the predicted agent so as to predict the stability of the network. The process will be repeated until reaching the stable situation

THE MULTI-AGENT STRUCTURE

The idea of designing the new structure was to increase the generation of the generating units in steps in the case of instability of the system after a disturbance. This increment will be added to the input data file of the load flow program (the control agent) to obtain anew power flow which is fed to the transient stability program (the predictive agent) so as to check the stability of the system. The process will continue until stable situation is reached. Figure 1 shows the proposed multi-agent's structure.

[FIGURE 1 OMITTED]

Scenario of the multi-agent technique can be illustrated as:

* Three-phase fault occurred on each generation buses

* Clearing the fault according to the calculated Critical Clearing Time (CCT)

* Observe the stability of the system

* Apply control agent to the disturbed generators

Prediction agent and control agent are linked to a measurement agent, which updates the values of the prediction and provides active power measurements for the control agent as shown in Fig. 2. This agent is represent the HQ agent in the proposed agent structure of Iraq control system (5).

[FIGURE 2 OMITTED]

The predicted agent: The proposed prediction agent is adapted from applying a transient stability program using DML/EAF (The Direct Method of Lyapunov/The Energy Approach Function. The interest in direct methods for transient stability analysis stems from the problem of real-time prediction of instability and control.

In PEBS Approach Function, the transient energy method can easily explained from the expression of the total energy. Most of the stability concepts can be interpreted by considering a ball sliding without friction in a bowl having a shape similar to that of the potential energy surface Vpe([delta]) as depicted in Fig. 3. During the fault-on period, an additional energy is injected into the system in the same way as the ball in the bowl is given energy when it is initially pushed. During the post-fault period, the total energy remains constant.

[FIGURE 3 OMITTED]

In Fig. 3, A1 represents the excess of kinetic energy injected into the system during the time period where the fault is on and A2 represents the total energy present in the system at the clearing time tc. The stability of the system is determined by the ability of the post-fault system to convert the excess kinetic energy Vke([omega]c). If the kinetic energy at clearing time exceeds the difference between the potential energy at clearing time and that at the u.e.p, the system will be unstable. Formally, we have:

[V.sub.Ke]([[omega].sub.e1] + [V.sub.Pe] ([[delta].sub.e1])<[V.sub.Pe]([[delta].sub.u]) (1)

This inequality is the mathematical way of stating the Energy Approach. The above inequality must hold not only at clearing time, but also during the post-fault time, that is:

[V.sub.Ke]([omega]) + [V.sub.Pe] ([delta])<[V.sub.Pe]([[delta].sub.u])(2)

According to the inequality (2), a transient stability criterion can be defined and extended to a multi-machine system as follows: