International Journal of Electronics, Communication & Instrumentation Engineering Research and Development (IJECIERD) ISSN 2249-684X Vol.

3, Issue 2, Jun 2013, 45-50 TJPRC Pvt. Ltd.

APPLICATION OF NEURAL NETWORK IN LOAD SHEDDING AND SOME PREDICTABLE FUNCTIONING OF LOAD SHEDDING METHODS
B. SUNEETHA1, S. HEMACHANDRA2, B. GOWRI PRASAD3 & A. JAYANTH4
1,3,4 2

Department of Econe, Sree Vidyanikethan Engineering College, Tirupati, Andhra Pradesh, India

Associate Professor& Head, Sree Vidyanikethan Engineering College, Tirupati, Andhra Pradesh India

ABSTRACT
There are many faults which alarm the voltage and frequency stability to the power systems. This voltage and frequency instability causes dispersal of a power system into sub-systems. The major problem of an electric power system is due to disproportion between loads and generated power. This leads to instability in the power system and hence resulting to a black out. A load shedding method may be used to avoid such problem by reducing the total load. . This paper presents a fast and optimal adaptive load shedding method, for power system using Artificial Neural Networks (ANN). The proposed method is able to determine the necessary load shedding in all steps simultaneously and is much faster than conventional method.

KEYWORDS: Optimal Adaptive Load Shedding Method, Artificial Neural Networks, Breaker Interlock Scheme INTRODUCTION
In general, load shedding can be defined as the quantity of load that must almost instantly be isolated from a power system to keep the remaining portion of the system operational. An electric power system is a bulky interconnected system that produces, transmits and distributes an electric energy to different consumers. Stability of the power system is of a great apprehension, since it is subjected to different disturbances that may cause a limited or complete system collapse if no tolerable action is taken to prevent it. Therefore, many techniques have been developed to make the power system survives during disturbances and continue to operate. One common disturbance is the disparity between generation and load due to an overload situation caused by generator outage or loss of transmission lines. Large changes in generation capacity through the loss of a generator or main impacts the dynamic response of the prime mover and can produce severe generation and load imbalance, resulting in rapid frequency decline. For some switching disturbances (that results in a loss of generation or system islanding condition), the cascading effects may be of the primary concern if the load shedding action is not set and timed correct. Load shedding is a process in which the electrical authority handles the deficiency of the electrical power being consumed by the people. Shedding is done to decrease the load being consumed by the people through several sub stations which are associated to the main power station. The main premise behind the proposed method is to enlarge a computerized procedure for controlling the load- shedding time period in a organized way so that in the shedding management process, manual work may be minimized.

PREDICTABLE FUNCTIONING OF LOAD SHEDDING

This section is a review of load shedding techniques that have been devised over a number of years each having its own set of applications and drawbacks.

46

B. Suneetha, S. Hemachandra, B. Gowri prasad & A. Jayanth

Breaker Interlock Scheme One of the simplest techniques of carrying out load shedding is Breaker interlock scheme. For example a source breaker would be interlock via hard wired or remote signals to a set of load breakers that have been pre-selected to trip. The signals are mechanically sent to load breakers to open, when generator breaker is misplaced for any motive. This scheme is very fast as there is no processing time required.

Figure 1 In Figure 1, the load is supplied by a combination of a generator and a power grid. A disturbance outside the facility causes the main breaker to operate and open. This would isolate the system from the power grid causing the system load to be supplied solely by the local generator (STG1). Although, all the decisions about the amount of load to be shed were made long before the fault occurred, breaker interlock load shedding possesses number of intrinsic drawbacks. It is very hard to change load priority, since the actions for load shedding are hardwired and amount of load shedding is calculated for the worst case scenario. Load shedding can be done at only one stage More loads are shed than necessary The operation of this kind of load shedding system will most likely shut the entire industrial facility downwards in a non-orderly way. This unintended outage may result in processing equipment spoil, reduced equipment lifetime, or worse Plant restarting may be delayed because of the requirement to shut down and then restart other remote facilities that have been affected by the loss of the main facility, before the main facility can be started Modifications to the system are very costly

Under Frequency Relay Method Frequency relays react to the disturbances but do not detect disturbances. They react to the disturbances commonly to both large and small systems. When the first stage is reached, the relay waits for a fixed period of time, to

Application of Neural Network in Load Shedding and Some Predictable Functioning of Load Shedding Methods

47

avoid nuisance tripping, and then trips one or more load breakers .If the frequency continues to the next stage to be reached and after an additional time delay, opens other load breakers.

Figure 2 The frequency relay based load shedding has a number of inherent drawbacks as listed below Slow response time of frequency relays Incorrect load may be dropped causing undesirable blackouts The unavailability of information for future changes and enhancement of the system will significantly reduce the protection system performance Programmable Logic Controller Based Load Shedding With a Programmable Logic Controller (PLC) scheme, load shedding is limited based on the total load versus the number of under frequency conditions. With common type of PLC based on the system frequency deviations and/or other triggers. PLC based load shedding scheme offers many advantages over the frequency based scheme since they have access to information about the actual operating status of the power systems. The drawback of this PLC is that the monitoring of the power system is limited to the sections of the system that are connected to the data acquisition system. This drawback is further compounded by the implementation of predefined load Priority tables in the PLC.

PROJECTED TECHNIQUE
The course of action to be followed in this case involves four steps: Recognition of input / output appropriate variables Data set generation Design of the NN Performance evaluation of the neural nets

48

B. Suneetha, S. Hemachandra, B. Gowri prasad & A. Jayanth

Recognition of Input /Output Appropriate Variables

The identification of the variables like inputs and outputs is an important step for a successful application of this technique. Sometimes a pre-processing stage is needed to choose the most significant variables to be used as inputs of a NN. Some of the meaningful variables that have been used as inputs of the NN Active real power generation Active load generation Amount of active load being shed Percentage of exponential type loads being shed Damping factor Power factor These variables provide the NN with valuable information, such that it can make the required assessment with respect to how much the generation load disproportion has been corrected and the influence each load type has on the resulting frequency response.

Generation of Data Sheet

The duplication of a given power systems response through any machine learning technique, like a NN, can only be accurate if the data used to train these structures describes with enough coverage and quality the different operating conditions. Optimally, this data set would include all possible system scenarios; however this would require unrealistic hours of computational time. Therefore the objective of the data generation stage is to capture the breadth of the system operating conditions and performance, while limiting computational and industrial efforts. This data set includes the data used for training a NN and the test data for evaluation purposes.

Design of Neural Network

The initial action in the design of a NN is to establish an architecture that will give in good results. The idea is to use the simplest architecture while maximizing performance. Within this work it was concluded after a few trials that architecture of 2 hidden layers, the first with 14 nodes and the second with 10 nodes, was best suited for this application.

Assessment of NN Performance
In order to get an idea for what kind of performance is to be expected from NN architecture, a preliminary evaluation is needed of its capabilities. The training set data, typically com- posed of % of the OPs from the overall data set, is used to teach the NN and give a relative inclination as to its suitability for that particular application. The test set, which is comprised of the remaining data, is used to evaluate the prediction capabilities and generalizatio n p e r f o r m a n c e o f the structures. The load shedding system can be designed to meet the following objectives: Systems knowledgebase trainable by user defined cases Pattern recognition capability to predict system Response to disturbances Make prompt decisions regarding which loads to shed based on the online status of sheddable Loads. Shed the minimum amount of load to maintain system Stability.

Application of Neural Network in Load Shedding and Some Predictable Functioning of Load Shedding Methods

49

CONCLUSIONS
This paper describes the purpose of a neural network to make a quick and exact prediction of the system performance for a load disproportion disorder followed by a load shedding control action. Tremendous results were obtained in an isolated system shows the applicability of this method for the purpose of evaluating the quality of different load shedding schemes for a pre-specified disturbance. Due to the fast prediction of system behaviour, the dispatch Centre can select the feeders to be disconnected in emergency conditions (meaning the amount of load to disconnect). This can be done by adapting the settings of the feeder by proper programming of PLCS which then disconnect the correct amount of load.

REFERENCES
1. 2. Edward Kim bark, Power System Stability, Wiley- IEEE, February 1, 1995. Warren C .New, Load Shedding, Load Restoration and Generator Protection using Solid Stateand Electromechanical Under frequency Relays, General Electric. 3. M. Sanaye Pasand, H. Seyedi, New Centralized Adaptive Under Frequency Load Shedding

Algorithms, Large Power Engineering System conference, pp. 44-48, 10-12 Oct. 2007. 4. C. T. Hsu, M. S. Kang and C. S. Chen, Design of Adaptive Load Shedding by Artificial Neural Networks, IEE Generation, Transmission, Distribution, Vol. 152, No. 3, pp. 415-421, 2005. 5. M. H. Purnomo, C. A. Patria, E. Purwanto, Adaptive Load Shedding of the Power System Based on Neural Network, IEEE Region 10 Con on Computers, Communications, Control and Power Engineering 3, 1778-1781, Oct. 2002. 6. E. J. Thalassinakis, E. N. Dialynas, D. Agoris, Method Combining Ann's and Monte Carlo Simulation for the Selection of the Load Shedding Protection Strategies in Autonomous Power System, IEEE Trans. on Power System, Vol. 21, No.4, pp. 1574-82, 2006. 7. T. Amraee, B. Mozafari, A. M. Ranjbar, An Improved Model for Optimal Under Voltage Load Shedding, Particle Swarm Approach, IEEE Power Conf. India, April 2006. 8. M. T. Hagan, M. B. Menhaj, Training Feed Forward Networks with the Marquardt Algorithm, IEEE Trans. On Neural Networks, Vol. 5, No. 6, pp. 989-93, 1994.