EFFECT OF LOAD FREQUENCY CONTROL ON MULTI-AREA POWER SYSTEM NETWORK

CHAPTER 1 INTRODUCTION
1.1 INTRODUCTION

As demand fluctuates from its normal operating value the state of the system changes. To maintain the system at normal operating state different types of controllers based on classical linear control theory has been developed in the past. Most load frequency controllers are primarily composed of an integral controller. The integrator gain is set to a level that compromises between fast transient recovery and low overshoot in the dynamic response of the overall system. This type of controller is slow and does not allow the designer to take into account possible nonlinearity in the generator unit. Automatic Generation Control. Automatic generation control, sometimes called load-frequency control (LFC), regulates power system in terms of maintaining scheduled system frequency and scheduled net interchange. Automatic generation control is implemented as a closed-loop feedback controller. The error signal is determined either as a computed area control error (ACE) for a control area or a given area requirement (AR) in some power pool control structures. Positive ACE indicates over generation; positive AR indicates under generation. The ACE calculation is based on frequency deviation

from schedule, net interchange deviation, or a composite tie-line bias. In tie-line bias control mode, interconnected control areas jointly participate in maintaining frequency, which is uniform among areas, but are individually responsible for maintaining each areas scheduled net interchange.

Area control error is a noisy signal and so requires processing. Processing also includes provision for proportional, integral, and anticipatory (or derivative) control characteristics for AGC as a feedback controller. Integral control is necessary to prevent long-term offset in frequency and to ensure that ACE crosses zero (the normal set point) frequently. System control requirements thus determined from processed ACE are allocated to generating units based on several criteria.

CHAPTER 2

CONVENTIONAL LOAD FREQUENCY CONTROL OF A SINGLE AREA SYSTEM

2.1 INTRODUCTION
In this chapter, we will discuss about the single area network and the load frequency of the single area network. Single area system is the system where all the generators are running coherently i.e., they are in unison. All generators are connected to form a state grid, regional grid and the national grid. Load dispatch centers are required for the control of power flow in these grids. It is feasible to divide a very large power system, say a national grid into subareas in which all the generators are assumed to be tightly coupled, i.e., they swing in unison with change in load or due to speed changer setting. All the machines in an area exporting power can be reduced to an equalent generator, therefore, for developing a suitable control strategy, we will consider that a given area forms a single generator, speed governor and load system.

When ever there is an increase in load on a generating unit, more amount of real power is to be supplied, which is immediately received from the kinetic energy power in rotating part, thereby reducing the KE of angular velocity or the speed of the machine. There will be a change in speed which is a measure of real power in balance. The change in speed is sensed by a speed governing mechanism and controls the position of inlet valve to the prime mover, thereby controlling the steam water supplied to turbine.

Consequently the machine comes back to the nominal speed and hence frequency. This action is a slow process, since mechanical elements are involved and usually the time involves is one to two seconds. The maximum permissible change in frequency i.e., f is about .

2.2 Speed governing system

The Speed governer is the main primary tool for the load frequency control whether the machine is used alone to feed a smaller system or whether it is a part of a most elaborated arrangement. A schematic arrengment of the main features of the main features of a speed governer system of the kind used on steam turbine is shown in the figure [2.2].

Fig 2.2. Speed governing system

Its main parts of the speed governing system are as follows:

Fly ball speed governor:

This is the heart of the system, which senses the change in the speed i.e, frequency. As the speed increases the fly balls move outwards and the point B on the linkage mechanism moves upwards the reverse happens when the speed decreases.

Hydraulic Amplifier:
It consists of a pivot valve V and an oil servo motor. With this arrangement hydraulic amplification is achieved, i.e., a low power pilot valve movement is converted into high power level movement of the oil servo motor piston. Hydraulic amplification is necessary, so that the steam valve could be operated against high pressure steam.

Speed changer:
It provides a steam state power output setting for the turbines. Its upward movement opens the upper pilot valve so that more steam is admitted to the turbine under steady state conditions. Hence more steady power output. The reverse happens for downward movement of speed changer.

Linkage mechanism:
A, B, C is the rigid link provided at the point B and C, D, E is another rigid link at point D. the function of this link mechanism is to control the steam valve V, only one valve V, but actually steam turbines have a number of them, which is opened in succession. With as smaller as possible in order to maintain high efficiency. Further via link D E we get a feedback from the movement of the steam valve.

2.3 Speed governing system model:

Now we shall develop the mathematical model based on small deviations around the nominal steady state. With the help of the speed changer, we command a power increase Pc. Due to this command, the change in the position of the linkage point A is given by

XA=K1. Pc

[2.3 a]

Now with the increase in the increased speed the fly balls of the governor move outwards, thus causes the change in the link point B. The change in the position of the point B with the change in the frequency f be XB. The change in the position of the linkage point B is given by

XB= K2. f

[2.3 b]

The change in the position of link C i.e., XC due to the links A and B is given by the following equation

XC= -K1. Pc + K2. f

[2.3 c]

The change in the position of point d, is contributed by the movement XC and XD. Thus we may write the net movement of link point D as

XD= K3. XC + K4. XE

[2.3 d]

Due to the change in the positions of all the links in the mechanism the change in the position of link E is given by

XE= - K5.1/s. XD(s)

[2.3 e]

By eliminating XC(s) and XD(s), we obtain the following equation

The above equation may be expressed as

In standard block diagram symbols we can represent the equation as shown in the figure, which represents the linearised model of speed governor mechanism

10

Fig 2.3(a). Block diagram representation of speed governor mechanism

Turbine model:
There is an increase in turbine power, Pt due to the change in the valve position, X5, which will result in an increased generator power Pg. If the incremental loss is neglected then Pt = Pg. A non- reheat steam turbine in characterized by a single gain factor Kt and a single time constant Tt. We can thus write

11

In the standard block diagram symbols we can represent the above equation, as shown in the following figure, which diagram represents the linearised model of a non-reheat turbine.

Fig.2.3(b). Block diagram representating the linearized model of a non-reheat turbine

Generator load model:

12

We shall proceed now to develop the mathematical model of an isolated generator, which is only supplying local load and is not supplying power to another area via a tie line. Suppose there is a real power change of Pd. Due to the action of the turbine controllers, the generator increases its output by the amount of Pg.

The total mega watt rating of the generator is given by

Or

Where the variables Pg and B are noe measured in per unit of S 3ph. Taking the laplace transform of above equation, we get

13

Where,

14

We can represent the above equation in the block diagram form as shown below

Fig.2.3(c). Block diagram representation of generator load model.

15

16

CHAPTER 3 CONVENTIONAL LOAD FREQUENCY CONTROL OF TWO AREA SYSTEM

3.1 INTRODUCTION

In this chapter, after discussing the block diagram of a two area system and the test system details, simulation results are presented. In a single area system mechanical power is produced by a turbine and delivered to a synchronous generator serving different users. The frequency at the output of the generator is mainly determined by the turbine steam flow .It also affected by changes in user power demands which appear, therefore, as electric perturbations. The electric load on the bus suddenly increases, the generator shaft slows down, and the frequency of the generator decreases. The control system immediately detect the load variation and command the steam admission valve to open more so that the turbine increases its mechanical power production ,counteracts the load increase and brings down the shaft speed and hence the generator frequency back to its nominal value. Steam enters the turbine through a pipe that is partially obstructed by a steam admission valve. In a steady state the opening of the valve is determined by the position of the device called the speed changer. The reference valve, or set point, of the turbine power in steady state is called the reference power and is denoted by Pref. When the load on the bus suddenly changes, the shaft speed is modified, and a device called the speed regulator acts through the rigid rods to move the steam valve. Note a similar effect is produced by temporarily modifying the reference power (which justifies the name speed changer).In practice, both control schemes are used simultaneously. Amplifying stages (generally

17

hydraulic) are introduced to magnify the outputs of the controllers and produce the forces necessary to actually move the steam valve. The speed regulator is a proportional controller of gain 1/R [10].In conventional systems, an integral controller of gain KI sums the frequency fluctuations f and uses the result as a controller signal to the speed changer to raise or lower the reference power. By combining two control loops [Fig 3.1], we get a parallel PI (proportional - integral)
Electric Bus 1

Area 1 Load Tie-Line Power Flow Signal Tie-Line

Area 2 Electric Bus 2 Load

Fig 3.1. Conventional Two Area System Basic Block Diagram

Controller capable of driving frequency fluctuations to zero whenever a step load perturbation is applied to the system [10]. Since most devices in power systems are nonlinear, one usually likes to linearize the plant and to think of different variables in terms of their fluctuations about given operating points. Nonlinearities are then modeled by making the parameters of the linearized system functions of

18

the operating point. The resulting small signal models consist of linear operators having variable parameters whose values depend upon the state of the system. The last step in modeling consists of replacing all small signals by their Laplace transform and to represent the linearzed devices by transfer functions.

3.2TEST SYSTEM AND SIMULATION RESULTS The conventional load frequency control of two-area system presented in the previous section is simulated using SIMULINK. This section gives the data used and the simulation results. These are used for comparing the performance of artificial neural network and dynamic neural network based load frequency control. The simulink model of the conventional load frequency control of two area system is given in Fig. 3.2.The data used for the simulation of these systems is given in Table 3.1 and the percentage of the loads applied on two areas is given in this section. Simulations are carried out to analyze the frequency response in both the areas and the tie-line power deviations from either area by applying a small disturbance of 1% in area 1. This model is used as reference model for the remaining load frequency controllers for adjusting the parameters of the neural networks. Table 3.1 Parameter values for two area system KI,1 =.425 KH,1 =0.1 TH,1=0.08 KT,1 =1 TT,1 =0.3 KP,1 =1 KI,2 =.425 KH,2 =0.1 TH,2=0.08 KT,2 =1 TT,2 =0.3 KP,2 =1

19

TP,1 =19 B1=.425 R1 =2.4

TP,2 =19 B2 =.425 R2 =2.4

Percentage of load applied on area 1 Percentage of load applied on area 1 Synchronization coefficient TO

1% 0% 0.0707

20

21

3.3INFERENCE From the plots, it is observed that the step load change of 1% disturbance in Area-1 resulted in 2.5% change of frequency from its nominal value in Area-1 and 1.75% change in Area-2. The steady state frequency value of zero is reached in approximately within 11 seconds. The tie-Line power variations (from Area-1 to Area-2) is approximately 0.6% for a given load disturbance in the two area test system.

3.4SUMMARY In this chapter PI (proportional-Integral) based load frequency control model is defined and the various blocks representing the model to control the frequency of a two area power system has been detailed. Standard PI based load frequency model [10] parameter values and the required time constant values of the linearzed model of the turbine; generator and Hydraulic amplifier is given. The simulink model of a two area power system is simulated to analyze the frequency deviations in both the areas and the tie-line power variations from both the areas for a step load change of 1% in Area-1 and zero step load change in Area-2.

22

CHAPTER - 4

INTRODUCTION TO SIMULINK
4.1 INTRODUCTION Simulink is a software package for modeling, simulating, and analyzing dynamical systems. It supports linear and nonlinear systems, modeled in continuous time, sampled time, or a hybrid of the two. For modeling, Simulink provides a graphical user interface (GUI) for building models as block diagrams, using click-and-drag mouse operations. Models are hierarchical, so we can build models using both top-down and bottom-up approaches. We can view the system at a high level, then double-click on blocks to go down through the levels to see increasing levels of model detail. This approach provides insight into how a model is organized and how its parts interact. After we define a model, we can simulate it, using a choice of integration methods, either from the Simulink menus or by entering commands in MATLAB's command window. Using scopes and other display blocks, we can see the simulation results while the simulation is running. In addition, we can change parameters and immediately see what happens, for "what if" exploration. The simulation results can be put in the MATLAB workspace for post processing and visualization. Simulink can be used to explore the behavior of a wide range of real-world dynamic systems, including electrical circuits, shock absorbers, braking systems, and many other electrical, mechanical, and thermodynamic systems. Simulating a dynamic system is a two-step process with Simulink. First, we create a graphical model of the system to be simulated, using Simulink's model editor. The model depicts the time-dependent mathematical relationships among the systems inputs, states, and outputs. Then, we use Simulink to simulate the behavior of the system over a specified time span. Simulink uses information that you entered into the model to perform the simulation.

23

4.2 Block Diagram A Simulink block diagram is a pictorial model of a dynamic system. It consists of a set of symbols, called blocks, interconnected by lines. Each block represents an elementary dynamic system that produces an output either continuously (a continuous block) or at specific points in time (a discrete block). The lines represent connections of block inputs to block outputs. Every block in a block diagram is an instance of a specific type of block. The type of the block determines the relationship between a block's outputs and its inputs, states, and time. A block diagram can contain any number of instances of any type of block needed to model a system. Blocks represent elementary dynamic systems that Simulink knows how to simulate. A block comprises one or more of the following: 1) A set of inputs, 2) A set of states, and 3) A set of outputs. A block's output is a function of time and the block's inputs and states (if any). The specific function that relates a block's output to its inputs, states, and time depends on the type of block of which the block is an instance. Continuous Versus discrete Blocks Simulink's standard block set includes continuous blocks and discrete blocks. Continuous blocks respond continuously to continuously changing input. Discrete blocks, by contrast, respond to changes in input only at integral multiples of a fixed interval called the block's sample time. Discrete blocks hold their output constant between successive sample time hits. Each discrete block includes a sample time parameter that allows you to specify its sample rate. The Simulink blocks can be either continuous or discrete, depending on whether they are driven by continuous or discrete blocks. A block that can be either discrete or continuous is said to have an implicit sample rate. The implicit sample time is continuous if any of the block's inputs are continuous. The implicit sample time is equal to the shortest input sample time if all the input sample times are integral multiples of the shortest time. Otherwise, the input sample time is equal to the

24

fundamental sample time of the inputs, where the fundamental sample time of a set of sample times is defined as the greatest integer divisor of the set of sample times. Simulink can optionally color code a block diagram to indicate the sample times of the blocks it contains, e.g., black (continuous), magenta (constant), yellow (hybrid), red (fastest discrete), and so on. The block contains block name, icon, and block library that contain the block, the purpose of the block. 4.3 Simulink Block Libraries Simulink organizes its blocks into block libraries according to their behavior. 1) The Sources library contains blocks that generate signals. 2) The Sinks library contains blocks that display or write block output. 3) The Discrete library contains blocks that describe discrete-time components. 4) The Continuous library contains blocks that describe linear functions. 5) The Math library contains blocks that describe general mathematics functions. 6) The Functions & Tables library contains blocks that describe general functions and table look-up operations. 7) The Nonlinear library contains blocks that describe nonlinear functions. 8) The Signal & Systems library contains blocks that allow multiplexing and demultiplexing, implement external input/output, pass data to other parts of the model, and perform other functions. 9) The Subsystems library contains blocks for creating various types of subsystems. 10) The Block sets and Toolboxes library contains the Extras block library of specialized blocks.

25

4.4 Sub Systems Simulink allows to model a complex system as a set of interconnected subsystems each of which is represented by a block diagram. We create a subsystem using Simulink's Subsystem block and the Simulink model editor. We can embed subsystems with subsystems to any depth to create hierarchical models. We can create conditionally executed subsystems that are executed only when a transition occurs on a triggering or enabling input. 4.5 Solvers Simulink simulates a dynamic system by computing its states at successive time step solver a specified time span, using information provided by the model. The process of computing the successive states of a system from its model is known as solving the model. No single method of solving a model suffices for all systems. Accordingly, Simulink provides a set of programs, known as solvers, that each embody a particular approach to solving a model. The Simulation Parameters dialog box allows us to choose the solver most suitable for our model. 4.5.1 Fixed-Step and Variable-Step Solvers Fixed-step solvers solve the model at regular time intervals from the beginning to the end of the simulation. The size of the interval is known as the step-size. We can specify the step size or let the solver choose the step size. Generally decreasing the step size increases the accuracy of the results while increasing the time required to simulate the system. Variable-step solvers vary the step size during the simulation, reducing the step size to increase accuracy when a model's states are changing rapidly and increasing the step size to avoid taking unnecessary steps when the model's states are changing slowly. Computing the step size adds to the computational overhead at each step but can reduce the total number of steps, and hence simulation time, required to maintain a specified level of accuracy for models with rapidly changing or piecewise continuous states.

26

4.5.2 Continuous and Discrete Solvers Continuous solvers use numerical integration to compute a model's continuous states at the current time step from the states at previous time steps and the state derivatives. Continuous solvers rely on the model's blocks to compute the values of the model's discrete states at each time step. Mathematicians have developed a wide variety of numerical integration techniques for solving the ordinary differential equations (ODEs) that represent the continuous states of dynamic systems. Simulink provides an extensive set of fixed-step and variable-step continuous solvers, each implementing a specific ODE solution method. Some continuous solvers subdivide the simulation time span into major and minor steps, where a minor time step represents a subdivision of the major time step. The solver produces a result at each major time step. It use results at the minor time steps to improve the accuracy of the result at the major time step. Discrete solvers exist primarily to solve purely discrete models. They compute the next simulation time-step for a model and nothing else. They do not compute continuous states and they rely on the model's blocks to update the model's discrete states. We can use a continuous solver, but not a discrete solver, to solve a model that contains both continuous and discrete states. This is because a discrete solver does not handle continuous states. If you select a discrete solver for a continuous model, Simulink disregards your selection and uses a continuous solver instead when solving the model. Simulink provides two discrete solvers, a fixed-step discrete solver and a variable-step discrete solver. The fixed-step solver by default chooses a step size and hence simulation rate fast enough to track state changes in the fastest block in our model. The variable-step solver adjusts the simulation step size to keep pace with the actual rate of discrete state changes in our model. This can avoid unnecessary steps and hence shorten simulation time for multirate models. 4.6 Model Execution Phase

27

In the simulation model execution phase, Simulink successively computes the states and outputs of the system at intervals from the simulation start time to the finish time, using information provided by the model. The successive time points at which the states and outputs are computed are called time steps. The length of time between steps is called the step size. The step size depends on the type of solver used to compute the system's continuous states, the system's fundamental sample time, and whether the system's continuous states have discontinuities (Zero Crossing Detection). At the start of the simulation, the model specifies the initial states and outputs of the system to be simulated. At each step, Simulink computes new values for the system's inputs, states, and outputs and updates the model to reflect the computed values. At the end of the simulation, the model reflects the final values of the system's inputs, states, and outputs. At each time step: 1) Simulink Updates the outputs of the models' blocks in sorted order. Simulink computes a block's outputs by invoking the block's output function. Simulink passes the current time and the block's inputs and states to the output function as it may require these arguments to compute the block's output. Simulink updates the output of a discrete block only if the current step is an integral multiple of the block's sample time. 2) Updates the states of the model's blocks in sorted order. Simulink computes a block's discrete states by invoking its discrete state update function. Simulink computes a block's continuous states by numerically integrating the time derivatives of the continuous states. It computes the time derivatives of the states by invoking the block's continuous derivatives function. 3) Optionally checks for discontinuities in the continuous states of blocks. Simulink uses a technique called zero crossing detection to detect discontinuities in continuous states. 4) Computes the time for the next time step. Simulink repeats steps 1 through 4 until the simulation stop time is reached.

28

4.6.1 Block Sorting Rules Simulink uses the following basic update rules to sort the blocks: 1) Each block must be updated before any of the direct-feed through blocks that it drives. This rule ensures that the inputs to directfeed through blocks will be valid when they are updated. 2) Non direct-feed through blocks can be updated in any order as long as they are updated before any direct-feed through blocks that they drive. This rule can be met by putting all non direct-feed through blocks at the head of the update list in any order. It thus allows Simulink to ignore non direct-feed through blocks during the sorting process. The result of applying these rules is an update list in which non direct-feed through blocks appear at the head of the list in no particular order followed by direct-feed through blocks in the order required to supply valid inputs to the blocks they drive. During the sorting process, Simulink checks for and flags the occurrence of algebraic loops, that is, signal loops in which an output of a direct-feed through block is connected directly or indirectly to one of the block's inputs. Such loops seemingly create a deadlock condition since Simulink needs the input of a direct-feed through block in order to compute its output. However, an algebraic loop can represent a set of simultaneous algebraic equations (hence the name) where the block's input and output are the unknowns. Further, these equations can have valid solutions at each time step. Accordingly, Simulink assumes that loops involving direct-feed through blocks do, in fact, represent a solvable set of algebraic equations and attempts to solve them each time the block is updated during a simulation. 4.7 Determining Block Update Order During a simulation, Simulink updates the states and outputs of a model's blocks once per time step. The order in which the blocks are updated is therefore critical to the validity of the results. In particular, if a block's outputs are a function of its inputs at the current time step, the

29

block must be updated after the blocks that drive its inputs. Otherwise, the block's outputs will be invalid. The order in which blocks are stored in a model file is not necessarily the order in which they need to be updated during a simulation. Consequently, Simulink sorts the blocks into the correct order during the model initialization phase. In order to create a valid update ordering, Simulink categorizes blocks according to the relationship of outputs to inputs. Blocks whose current outputs depend on their current inputs are called direct feed through blocks. All other blocks are called non direct-feed through blocks. Examples of direct-feed through blocks include the Gain, Product, and Sum blocks. Examples of non direct-feed through blocks include the Integrator block (its output is a function purely of its state), the Constant block (it does not have an input), and the Memory block (its output is dependent on its input in the previous time step). Simulink allows you to assign update priorities to blocks. Simulink updates higher priority blocks before lower priority blocks. Simulink honors the priorities only if they are consistent with its block sorting rules. Some of SIMULINK blocks, which are used in this thesis, are given below. 1) Abs:

Purpose: Output the absolute value of the input. Description: The Abs block implements the equation y = , where U is U the input and y is the output. 2) Clock:

Purpose: Provide and display the system time.

30

Description: The clock block serves two purposes. When opened, it provides a window that continuously displays the system time as a simulation progress. This time also available as the output of the clock for any blocks whose behavior depends on time. 3) Constant:

Purpose: Inject a constant value. Description: The output of the constant block is the value of the specified parameter independent of time. 4) Demux:

Purpose: Separate a vector signal into scalar signals. Description: The Demux block separates a vector line into several scalar lines. The number of the outputs of the SIMULINK block must match the width of the input line. If the number of outputs does not match the input width, SIMULINK signals an error for the block. 5) Derivative:

Purpose: Output the time derivative of the input. Description: The Derivative block approximates the derivative of its input with du/dt, where du is the change in input value since the previous simulator time step, and dt is the change in time since the previous simulator time step. The accuracy of the results depends on the size of the time steps taken in the simulation. 6) Discrete Transfer Fcn:

Purpose: Implement a discrete transfer function.

31

Description: Discrete Transfer Fcn block requires the specification of vectors for the numerator and denominator. The vectors elements are treated as the coefficients of the numerator and denominator polynomials in descending order of z. this block represents the method used in the control system toolbox, while the filter block represents the method used in signal processing toolbox. The two methods are identical when the numerator is the same length as the denominator. 7) Discrete Zero-Pole:

Purpose: Provide a discrete transfer function in zero-pole form. Description: This block implements a discrete single input-single output (SISO) zero-pole-gain in terms of the delay operator z. 8) Fcn:

Purpose: Complete a general-purpose function of the input. Description: The Fcn block applies the specified function to its input. This function must take the form of a legal expression in C. It can be made up of the numbers, binary operators, unary operators, parentheses, relational operator, functions of 1 variable, functions of 2 variable and MATLAB variables. 9) From Workspace:

Purpose: Read data from a matrix. Description: The specified matrix must have two or more columns. The first column must contain monotonically increasing points. Each additional column contains data points corresponding to the time point in its row. 10) Gain:

32

Purpose: Multiply its input by a constant. Description: The Gain block implements Y=KU, where Y is the output, U is the input, and K is the specified gain. The Gain block displays scalar gain data entered as variable or a constant. The block displays the text as it appears in the dialog box. 11) In port: Purpose: Provide a link to an external input and for linearization. Description: In ports are the links from the outside world into a system. Inside a subsystem block, there is an in port corresponding to each input port on the block. A signal that arrives at an input port on a subsystem block flows out of the corresponding in port within that block. The imports within a subsystem block must be numbered consecutively, starting with 1. 12) Integrator:

Purpose: Integrate a signal. Description: The Integrator block integrates its input. Its single parameter is its initial condition. The output of the integrator is simply its state, the integral. The initial value can be entered as a constant or a variable. 13) Look-Up Table:

Purpose: Perform piecewise linear mapping of the input. Description: The Look Up Table block implements the piecewise linear mapping described by its two vector parameters. Input values are located in the input vector and corresponding output values are taken from the same position in the output vector. Linear interpolation is used for input values falling between specified input values.

33

14) Mux:

Purpose: Group several scalar lines into a vector line. Description: The Mux block groups several scalar lines into a single vector line. The width of the vector line is the same as the number of inputs specified. If the number of inputs does not match the output width, SIMULINK signals an error for the block. 15) Out port: Purpose: provide a link to an external output and for linearization. Description: The Out port block provides a mechanism for labeling a systems outputs. In a subsystem, output ports correspond to outputs on the subsystem block. 16) Product:

Purpose: Multiply inputs together Description: The Product block multiplies the value of each input to produce the output. If the inputs are vectors, an element-by-element multiplication is performed. The number of inputs parameter can be represented with a constant or a variable. 17) Relay:

Purpose: Switch output between two values. Description: The Relay block allows the output to switch between two specified values. When the relay is on, it remains on until the input drops below the specified value of Input for off. When the relay is off, it remains off until the input exceeds the specified value of Input for on.

34

18) Scope:

Purpose: Display signals during simulation. Description: While the simulation is running, the Scope block displays the output of the block driving it. Opening a scope block produces a scope window. The title of this window matches the name of the block. 19) Step Fcn:

Purpose: Generate step function. Description: The Step Fcn block provides a step between two arbitrary levels at a specified time. If system time is less than the step time parameter, the blocks output is the initial value. For time greater than or equal to the step time, the output is the final value. 20) Subsystem:

Purpose: Group blocks into a subsystem Description: Subsystem blocks represent one system within another system. Any set of blocks and lines can be converted to a Subsystem block with the Group command on the options menu. The Group command removes all selected objects from the active window and replaces them within a Subsystem block. This new block, when opened, redisplays all of the grouped objectives. 21) Sum: Purpose: Sum outputs together Description: The Sum block accepts scalar inputs only and adds the value of each input to produce a scalar output. The list of signs parameter can be represented with a constant or a combination of the symbols (+-). 22) Switch:

35

Purpose: Switch between two inputs. Description: The Switch block uses its second input to control which of its other two inputs is propagated to its output. If the signal on the second input is greater than or equal to zero, the first input is propagated; otherwise, the third input drives the output. 23) To Workspace:

Purpose: Write data to a matrix. Description: The To Workspace block creates a matrix with the specified name containing the data from the vector input line. The matrix will have the same number of columns as the vector width of the input with each channel of data being written to a different column. One row of data is written for each simulation step containing the state of each element of the vector input. Every time a simulation is finished, any existing matrix of the same name is over written; this block does not append data. The matrix is not available in the workspace until the simulation is terminated. 24) Zero-pole:

Purpose: Implement a transfer function specified in terms of poles and zeros. Description: the Zero-Pole block implements a system with the specified zeros, poles, and gains in the Laplace operator s.
0.1

36

37

CHAPTER 5 CONCLUSION AND FUTURE SCOPE WORK

5.1 CONCLUSION In this work investigations related to the applications of multilayer feed forward network models that are applied to load frequency control of a two area system have been carried out. These investigations reveal that this control approach causes lesser frequency drop in the two areas than the conventional controllers. It has also been observed that the oscillations in frequency and tie line power following load disturbances damp out more rapidly .

5.2 SCOPE FOR FURTHER WORK Application of ANN based LFC to multi-area power systems. Investigations related to the hardware implementations

38

REFERANCES

Beau fays F, Abdel-Magid Y, Widrow B. Application of neural networks to load frequency control in power systems. Neural Networks 1994;7(1):pp.183198. Chaturvedi DK, Satsangi PS, Kalra PK. Load frequency control: A generalized neural network approach. Int J Electr Power Syst 1999; 21(6): pp 40515. Becerikli Y, Konar AF, Samad T. Intelligent optimal control with dynamic neural networks. Neural Networks 2003; 16: pp 2519. Djukanovic M, Novicevic M, Sobajic DJ, Pao Y-P. Conceptual development of optimal load frequency control using artificial neural networks and fuzzy set theory. Int J Eng Intell Syst Elec Eng Commun 1995;3(2):pp 95108. Olle L.Elgerd ELECTRICAL ENERGY SYSTEM THEORY; AN INTRODUCTION second edition. Tata McGraw-Hill publishing company limited. New Delhi Konar AF, Becerikli Y, Samad T. Trajectory tracking with dynamic neural networks, In: Proc 1997 IEEE Int Symp on Intelligent Control (ISIC_97), Istanbul/Turkey, 1997. Hopfield JJ. Neural networks and physical systems with emergent collective computational abilities. Proc Nat Acad Sci 1982; 79:pp 25548. Hopfield JJ. Neurons with graded response have collective computational properties like those of two-state neurons. Proc Nat Acad Sci 1984; 81: pp 308892.

39

Edgar TF, Himmelblau DM. Optimization of chemical processes. McGraw-Hill; 1988. Anderson JA, Rosenfeld. Neurocomputing: foundation of research. Cambridge, MA: MIT Press; 1988.www.neurosolutions.com

10

40