Smart Grid

Only available on StudyMode
  • Download(s) : 112
  • Published : May 14, 2013
Open Document
Text Preview
Electrical and Po wer Eng ineering Frontier

Dec.2012, Vo l. 1 Iss. 1, PP. 8 -12

Fuzzy Logic Approach for Short Term Electrical Load Forecasting M. Rizwan1 , Dinesh Kumar2 , Rajesh Kumar3
1, 2, 3

Department of Electrical Engineering, Delh i Technological Un iversity, Delhi 110042, India 1 rizwaniit@yahoo.co.in; 2 dinesh030989@g mail.co m; 3 1991rajesh@gmail.co m

Abstract- The demand of electricity in India is increasing exponentially at the rate of 8-9% per annum. However, the installed power generation capacity of India as on 31 st October 2012 was 209276 MW with a peak power shortage of more than 12%. In addition, the demand of electricity is increasing due to increased population, urbanization and comfort level of the peoples. These indicate that India’s future energy requirements are going to be very high. Keeping in view of aforesaid, proper energy management system is required. In this paper an attempt has been made for short term load forecasting which helps in load management with on line dynamic voltage control, load flow studies and exchange of power as requirement for load frequency control. In this paper, the daily hourly demand of Shahpura, Jaipur, India has been collected from Rajasthan Electricity Board (S hahpura S ubstation), India. To avoid the convergence problems, the input and output load data are scaled down such that they remain within the range of (0.1-0.9). The inputs of the fuzzy logic based models are the hourly electrical demand during the day for the four consecutive days of November 2012 an d the output or forecasted value is the hourly demand of the fifth day. The results obtained from fuzzy logic model has been validated with the actual value and found accurate. The mean absolute percentage error (MAPE) in the forecasted demand is 1.39% in comparison with the desired demand. Keywords- Fuzzy Logic; Load Forecasting; Energy Management

quarterly, half yearly and yearly load forecasting needs. Short Term Load Forecasting (STLF) means days ahead and weeks ahead load fo recasting needs. There are many techniques that could be emp loyed for loads forecasting like linear regression, statistical method, exponential smoothening, neural network based artific ial intelligence technique, fuzzy logic, genetic algorith m, autoregressive model, similar day approach, time series, e xpert system, support vector machine, and data mining model [3-12]. Among these, ANN is widely used when there is no fluctuation in condition like temperature, weather and load. In case of sudden fluctuation in load and temperature, fu zzy logic based approach is used. It has advantage over ANN that it leads with non-linear part of the forecasted load curve as well as it has ability to deal with sudden variation in load i.e. load and temperature variat ion. In addition, fuzzy logic approach is easy and robust. Keeping in view of the aforesaid variation in the inputs, an attempt has been made to develop the fuzzy log ic based model for short term load forecasting. The proposed model is simp le, accurate and incorporates the uncertainties in the input variables. This paper is organized as fo llows: Brief idea about the fuzzy logic is given in Sect ion II. Sect ion III presents the data collection and normalization o f input and output data. Develop ment of fu zzy logic model for short term load forecasting is presented in Sect ion IV. Results are discussed in Section V. Conclusion followed by references is presented in Section VI. II. BASICS OF FUZZY LOGIC The concept of Fuzzy Logic was introduced by Professor Lotfi A. Zadeh at the University of Califo rnia, Berkeley in the 1960's. His goal was to develop a model that could more closely describe the natural language process. This model was intended to be used in situations when deterministic and/or probabilistic models do not provide a realistic description of the phenomenon under study. The fuzzy sets and fuzzy operators are the subjects and verbs of fuzzy logic. But...
tracking img