Professional Documents
Culture Documents
• Linear Models
• Linear Regression
• Time Series Approach
• Nonlinear Models
• Artificial Neural Networks
• Nonlinear Regression
• Fuzzy Approach
• Bayesian Network Approach
Weekend Week Days
Source-RTE France
Daily Consumption
Second Peak
First Peak
Source-RTE France
Determining factors
• Calendar
Seasonal variation
Daily variation
Weekly Cycle
Holidays
• Economical or environmental
• Weather
Temperature
Cloud cover or sunshine
Humidity
• Unforeseeable random event
Hidden layer
f (∑ wk xk )
Forecasted Load
Output layer
1
f (x) =
1 − e−x
f ′ ( x ) = f ( x ) (1 − f ( x ) )
Input layer
STLF Using ANN (1st Approach)
LOAD IN MEGAWATT
4
x 10 Actual Load Hour 23/12/2005 20/01/2006
8 00:00 70500 64200
01:00 67300 60800
02:00 68900 61700
03:00 66500 59000
7.5
04:00 64400 57100
05:00 64100 57000
06:00 66900 60300
08:00 1800 0
09:00 1100 -300
2000
10:00 -700 -600
11:00 -100 0
12:00 700 -200
0
13:00 -2500 -1500
14:00 -2600 -2200
The input I11, I12 has five membership functions each. I11
represents the load increment at the kth hour and I12
represents the forecasted load increment at the same
hour. The forecasted load increment was obtained using
the traditional ANN.
Results
“The best way to predict
the future is to invent it”
THANKS