Clustering Techniques in Load Profile Analysis for Distribution Stations PDF

0 votes
 
Abstract—The demand characteristic is the most important 
one in analyzing customer information. In a distribution 
network, there is in any moment certain degree of uncertainty 
about busses loads, and consequently, about load level of 
network, busses voltage level, and power losses. Therefore, it is 
very important to estimate first of all the load profiles of buses, 
using available data (measurements effectuated in distribution 
stations). The results obtained for various distribution stations 
demonstrate the effectiveness of the present method in 
overcoming the difficulties encountered in optimal planning 
and operation of distribution networks. 
 
Index Terms—load profile, clustering techniques, data flow 
analysis, power consumption, distribution station 
I.  INTRODUCTION 
Electric distribution networks have a large number of load 
busses, even if we were to take into consideration only the 
busses with substations. The consumers connected in the 
network busses are also very numerous, heterogeneous as 
the absorbed powers, using their technologists, social 
behaviors, enforcing the particular loop of thing. 
In addition, the multiple participants in the electricity 
market need new business strategies for providing value 
added services to customer. They need, therefore, accurate 
customer information about  the electricity demand. The 
Demand characteristic is the most important one for 
analyzing customer information.  
These difficulties are eliminated if for the distribution 
networks analysis a daily load curve is used for each bus, 
within characteristic regimes  (winter and summer, working 
day and weekend day).  
The models for electric loads determinations are different, 
depending on the networking tips: urban, rural or industrial. 
Variation in time of the electric load reflects the graphic of 
daily, seasonal and annual load, which indicate the real 
electric energy consumption. 
In this paper, load profile data, which can be collected by 
means of the automatic meter  reading system, are analyzed 
in order to get demand patterns of customers. The load 
profile data include electricity demand at a 15 minutes 
interval. An algorithm for clustering similar patterns is 
developed using the load profile data. As a result of the 
classification, representative curves for the same groups are 
generated. The demand characteristics of the groups are 
further discussed. 
II.  GENERAL CONSIDERATIONS 
Since the storage of electric energy on a large scale is not 
possible, the main role of the power network is to transport 
the demanded energy to consumers. Therefore, it is very 
important to study and analyze the evolution of the load in 
order to operate and design the power network. All the other 
decisions are based on the consumed energy such as the load 
forecast, the voltage control, determination of the peak load 
for various types of consumer, calculations of the power 
losses or power losses estimations, proper tariff design, etc. 
 
 
Figure 1. Load profile for different seasons. 
 
The main causes generating load modifications are: 
•  weather conditions: the season, the daily 
temperatures, the speed of the wind, etc; (Fig. 1) 
•  demographic factors: the growth rate of the 
population, the number of the inhabitants in a 
certain area, the birth rate, etc; 
•  economic factors: the gross national product, the 
labor productivity, the economy development 
rate, the level of life quality and a very important 
element: the price of energy.  
The evolution in time of these parameters has a strong 
random character. At a certain  moment, the more or less 
accidental realization of these parameters directly influences 
the load and its variation change tendency influences in a 
decisive way the load curves. 
 
 

 

asked in Education by

Please log in or register to answer this question.

Vietnam-Cruise.com
Great deals for all cruises on our website