Embed
Email

Lectures of Load Forecasting_nnt

Document Sample

Shared by: cuiliqing
Categories
Tags
Stats
views:
6
posted:
11/2/2011
language:
English
pages:
89
SHORT TERM LOAD FORECASTING

USING NEURAL NETWORKS AND

FUZZY LOGIC









George G Karady

Arizona State University





06/2000 Short-term load forecasting 1

Short Term Load Forecasting

Content

 Overview of Short term load forecasting

• Introduction

• Definitions and expected results

• Importance of Short-term load forecasting

• Impute data and system parameters required for load forecasting

• Concept

• Major load forecasting techniques

• Concept of STLF model Development

 Statistical methods

• (Multiply liner regression,

• stochastic time series e.t.c)





06/2000 Short-term load forecasting 2

Short Term Load Forecasting

Content

 Artificial Neural Networks

• Building block: a feed forward network

• Load forecasting engine

• Results

• Numerical example

 Fuzzy logic and evolutionary programming









06/2000 Short-term load forecasting 3

Overview of Short Term

Load Forecasting (STLF)





06/2000 Short-term load forecasting 4

Introduction

 The electrical load increases about 3-7%

per year for many years.

 The long term load increase depends on

the population growth, local area

development, industrial expansion e.t.c.

 The short term load variation depends on

weather, local events, type of day

(Weekday or Holiday or Weekend) e.t.c.

06/2000 Short-term load forecasting 5

Introduction

 The building of a power plant requires:

• 10 years (Nuclear)

• 6 years (Large coal-fired)

• 3 years (combustion turbine)

 The electric system planing needs the

forecast of the load for several years.

 Typically the long term forecast covers a

period of 20 years

06/2000 Short-term load forecasting 6

Introduction

 The planning of maintenance, scheduling

of the fuel supply etc. calls for medium

term load forecast .

 The medium term load forecast covers a

period of a few weeks.

 It provides the peak load and the daily

energy requirement



06/2000 Short-term load forecasting 7

Introduction

 The number of generators in operation,

the start up of a new unit depends on the

load.

 The day to day operation of the system

requires accurate short term load

forecasting.





06/2000 Short-term load forecasting 8

Introduction

 Typically the short term load forecast

covers a period of one week

 The forecast calculates the estimated load

for each hours of the day (MW).

 The daily peak load. (MW)

 The daily or weekly energy generation.

(MWh)

06/2000 Short-term load forecasting 9

Introduction

 The utilities use three types of load

forecasting:

• Long term (e.g. 20 years)

• Medium term. (e.g. 3-8 weeks)

• Short term (e.g. one week)



 This lecture presents the short term load

forecasting techniques

06/2000 Short-term load forecasting 10

Definitions and Expected Results

(Big picture)







 The short term load forecasting provides

load data for each hour and cover a

period of one week.

 The load data are:

• hourly or half-hourly peak load in kW

• hourly or half-hourly values of system

energy in kWh

• Daily and weekly system energy in kWh

06/2000 Short-term load forecasting 11

Definitions and Expected Results

(Big picture)







 The short term load forecasting is

performed daily or weekly.

 The forecasted data are continuously

updated.

 Typical short-term, daily load forecast is

presented in the Table in the next page.

(Salt River Project, SRP)



06/2000 Short-term load forecasting 12

Definitions and Expected Results

(Big picture)





Typical short-term, daily load forecast from Salt

River Project. Hourly Load in MW

1/01/85

2153 2075 2001

2787 2677 2545



01/02/85

1930 1893 1893

2721 2613 2500



01/03/85

1892 1839 1835

2657 2526 2412



01/04/85

1856 1803 1806

2716 2620 2537



01/05/85

1793 1720 1672

2509 2314 2170

06/2000 Short-term load forecasting 13

Definitions and Expected Results

(Big picture)





Typical short-term, daily load forecast from Salt

River Project. Hourly Load in MW

1/01/85

2153 2075 2001

2787 2677 2545



01/02/85

1930 1893 1893

2721 2613 2500



01/03/85

1892 1839 1835

2657 2526 2412



01/04/85

1856 1803 1806

2716 2620 2537



01/05/85

1793 1720 1672

2509 2314 2170



06/2000 Short-term load forecasting 14

Importance of Short-term Load

Forecasting



 Provide load data to the dispatchers for

economic and reliable operation of the

local power system.

 The timeliness and accuracy of the data

affects the cost of operation.

• Example: The increase of accuracy of the

forecast by 1% reduced the operating cost

by L 10M in the British Power system in

1985

06/2000 Short-term load forecasting 15

Importance of Short-term Load

Forecasting



 The forecasted data are used for:

• Unit commitment.

– selection of generators in operation,

– start up/shut down of generation to minimize

operation cost

• Hydro scheduling to optimize water release

from reservoirs

• Hydro-thermal coordination to determine the

least cost operation mode (optimum mix)

06/2000 Short-term load forecasting 16

Importance of Short-term Load

Forecasting



 The forecasted data are used for:

• Interchange scheduling and energy purchase.

• Transmission line loading

• Power system security assessment.

– Load-flow

– transient stability studies

using different contingencies and the predicted

loads.

06/2000 Short-term load forecasting 17

Importance of Short-term Load

Forecasting



 These off-line network studies:

• detect conditions under which the system is

vulnerable

• permit preparation of corrective actions

– load shedding,

– power purchase,

– starting up of peak units,

– switching off interconnections, forming islands,

– increase spinning and stand by reserve

06/2000 Short-term load forecasting 18

Impute Data and System

Parameters for Load Forecasting



 The system load is the sum of individual

load.

 The usage of electricity by individuals is

unpredictable and varies randomly.

 The system load has two components:

• Base component

• Randomly variable component



06/2000 Short-term load forecasting 19

Impute Data and System

Parameters for Load Forecasting



 The factors affecting the load are:

• economical or environmental

• time

• weather

• Unforeseeable random events









06/2000 Short-term load forecasting 20

Impute Data and System

Parameters for Load Forecasting



Economical or environmental factors

• Service area demographics (rural, residential)

• Industrial growth.

• Emergence of new industry, change of farming

• Penetration or saturation of appliance usage

• Economical trends (recession or expansion)

• Change of the price of electricity

• Demand side load management

06/2000 Short-term load forecasting 21

Impute Data and System

Parameters for Load Forecasting



 The time constraints of economical or

environmental factors are slow,

measured in years.

 This factors explains the regional variation

of the load model (New York vs. Kansas)

 The load model depends on these slow

changing factors and has to be updated

periodically

06/2000 Short-term load forecasting 22

Impute Data and System

Parameters for Load Forecasting



Time Factors affecting the load

 Seasonal variation of load (summer,

winter etc.). The load change is due to:

– Change of number of daylight hours

– Gradual change of average temperature

– Start of school year, vacation



 Calls for a different model for each

season

06/2000 Short-term load forecasting 23

Impute Data and System

Parameters for Load Forecasting



Typical Seasonal Variation of Load

Summer peaking utility









06/2000 Short-term load forecasting 24

Impute Data and System

Parameters for Load Forecasting



Time Factors affecting the load

Daily variation of load. ( night, morning,etc)









06/2000 Short-term load forecasting 25

Impute Data and System

Parameters for Load Forecasting



Weekly Cyclic Variation

• Saturday and Sunday significant load reduction

• Monday and Friday slight load reduction

• Typical weekly load pattern:









06/2000 Short-term load forecasting 26

Impute Data and System

Parameters for Load Forecasting



Time Factors affecting the load

 Holidays (Christmas, New Years)

• Significant reduction of load

• Days proceeding or following the holidays

also have a reduced load.

– Pattern change due to the tendency of prolonging

the vacation





06/2000 Short-term load forecasting 27

Impute Data and System

Parameters for Load Forecasting



 Weather factors affecting the load

 The weather affects the load because of

weather sensitive loads:

• air-conditioning

• house heating

• irrigation





06/2000 Short-term load forecasting 28

Impute Data and System

Parameters for Load Forecasting



Weather factors affecting the load

The most important parameters are:

• Forecasted temperature

• Forecasted maximum daily temperature

• Past temperature

Regional temperature in regions with

diverse climate



06/2000 Short-term load forecasting 29

Impute Data and System

Parameters for Load Forecasting



Weather Factors Affecting the Load

The most important parameters are:

• Humidity

• Thunderstorms

• Wind speed

• Rain, fog, snow

• Cloud cover or sunshine



06/2000 Short-term load forecasting 30

Impute Data and System

Parameters for Load Forecasting



Random Disturbances Effects on Load

 Start or stop of large loads (steel mill,

factory, furnace)

 Widespread strikes

 Sporting events (football games)

 Popular television shows

 Shut-down of industrial facility

06/2000 Short-term load forecasting 31

Impute Data and System

Parameters for Load Forecasting



 The different load forecasting techniques

use different sets of data listed before.

 Two -three years of data is required for

the validation and development of a new

forecasting program.

 The practical use of a forecasting

program requires a moving time window

of data

06/2000 Short-term load forecasting 32

Impute Data and System

Parameters for Load Forecasting



 The moving time window of data

requires:

• Data covering the last 3-6 weeks

• Data forecasted for the forecasting period,

generally one week









06/2000 Short-term load forecasting 33

Impute Data and System

Parameters for Load Forecasting



 The selection of long periods of historical

data eliminates the seasonal variation



 The selection of short periods of historical

data eliminates the processes that are no

longer operative.





06/2000 Short-term load forecasting 34

Impute Data and System

Parameters for Load Forecasting



 The forecasting is a continuous process.

 The utility forecasts the load of its service

area.

 The forecaster

• prepares a new forecast for everyday and

• updates the existing forecast daily

 The data base is a moving window of data

06/2000 Short-term load forecasting 35

Major Load Forecasting

Techniques



 Statisticalmethods

 Artificial Neural Networks

 Fuzzy logic

 Evolutionary programming

 Simulated Annealing and expert system

 Combination of the above methods





06/2000 Short-term load forecasting 36

Major Load Forecasting

Techniques



 The statistical methods will be discussed

briefly to explain the basic concept of the

load forecasting



 This lecture concentrates on load

forecasting methods using neural

networks and fuzzy logic



06/2000 Short-term load forecasting 37

Concept of STLF Model

Development



 Model selection

 Calculation and update of model

parameters

 Testing the model performance

 Update/modification of the model if the

performance is not satisfactory





06/2000 Short-term load forecasting 38

Concept of STLF Model

Development



 Model selection

• Selection of mathematical techniques that

match with the local requirements

 Calculation and update of model

parameters

• This includes the determination of the constants and

• selection of the method to update the constants

values as the circumstance varies. (seasonal

changes)

06/2000 Short-term load forecasting 39

Concept of STLF Model

Development



 Testing the model performance

• First the model performance has to be

validated using 2-3 years of historical data

• The final validation is the use of the model in

real life conditions. The evaluation terms

are:

– accuracy

– ease of use

– bad/anomalous data detection

06/2000 Short-term load forecasting 40

Concept of STLF Model

Development



 Update/modification of the model if the

performance is not satisfactory

• Due to the changing circumstances (regional

gross, decline of local industry etc.) the

model becomes obsolete and inaccurate,

• Model performance, accuracy has to be

evaluated continuously

• Periodic update of parameters or the change

of model structure is needed

06/2000 Short-term load forecasting 41

Artificial Neural Networks

for STLF





06/2000 Short-term load forecasting 42

Artificial Neural Networks

for STLF



 Several Artificial Neural Network (ANN)

based load forecasting programs have

been developed.

 The following neural networks were

tested for load forecasting:

• Feed-forward type ANN

• Radial based ANN

• Recurrent type ANN

06/2000 Short-term load forecasting 43

Building Blocks of

a Feed Forward Network



A Feed forward Three-Layered

Perceptron Type ANN was selected to

demonstrate the short term load

forecasting technique.

 The selected network forecasts:

• Hourly loads

• Peak load of the day

• Total load of the day.



06/2000 Short-term load forecasting 44

Building Blocks of

a Feed Forward Network



 The forecasting with neural network will

be demonstrated using a feed forward

three-layered network to forecast the

peak load of the day.

 The network has:

• one output: (load in kW)

• three input: (previous day max. load and

temperature, forecasted max. temperature)_

06/2000 Short-term load forecasting 45

Building Blocks of

a Feed Forward Network



 The structure of the Feed Forward

Three-Layered Perceptron Type ANN is

presented on the next page.

 The network contains:

• i = 1.. 3 input layer nodes

• j = 1…5 hidden layer nodes

• k=1 output layer nodes



06/2000 Short-term load forecasting 46

Artificial Neural Networks

for STLF

Y









w1 w2 w3 w4 w5



1 2 3 4 5



W11 W12 W25 W35



1 2 3

X1 X2 X3



06/2000 Short-term load forecasting 47

Building Blocks of

a Feed Forward Network



 The inputs are:

• X1 previous day max. load

• X2 previous day max. temperature

• X3 forecasted max. temperature

 Wij weight factor between input

and hidden layer

 wj weight factor between hidden

layer and output

06/2000 Short-term load forecasting 48

Building Blocks of

a Feed Forward Network



A sigmoid function is placed in the nodes

(neurons) of the hidden layer and output

node.

 The sigmoid equation for an arbitrary Z

function is:

1

1  e Z



Y output :maximum load

06/2000 Short-term load forecasting 49

Building Blocks of

a Feed Forward Network



 Inputs Xi are multiplied by the

connection weights (Wij) and passed on to

the neurons in the hidden layer.

 The weighted inputs (Xi*Wij) to each

neurons are added together and passed

through a sigmoid function.

 Input of hidden layer neuron 1 is:

X1 W11  X2 W21  X3 W31

06/2000 Short-term load forecasting 50

Building Blocks of

a Feed Forward Network



 The output Hj of the jth hidden layer

node is:





Hj  1

n

W X

i 1 ij j

1 e

06/2000 Short-term load forecasting 51

Building Blocks of

a Feed Forward Network



 Inputs Hj are multiplied by the

connection weights (wk) and passed on to

the neurons in the output layer.

 The weighted inputs (Hj*wk) to each

output neurons are added together and

passed through a sigmoid function.

 Input of output neuron is:

H1 W1  H2 W2  H3 W3  H4 W4  H5 W5



06/2000 Short-term load forecasting 52

Building Blocks of

a Feed Forward Network



 The output Y is:





Y 1

h

w H

j j

1 e j 1

06/2000 Short-term load forecasting 53

Training of the

Feed Forward Neural Network



 The described neural network is trained

using historical data.

 Typical data set contains 2-3 years of

load and weather data.

 Error back propagation (BP) method is

used for the training.

 During the learning the weights are

adjusted repeatedly.

06/2000 Short-term load forecasting 54

Training of the

Feed Forward Neural Network



 The output produced by the ANN in

response to inputs are repeatedly

compared with the correct answers

 Each time the weights are adjusted

slightly by beck-propagating the error at

the output layer through the ANN

 Equations for the training are presented

in the next page

06/2000 Short-term load forecasting 55

Training of the

Feed Forward Neural Network



 The equations used for the training are:

• Weight is between input and hidden layer:

n 1

 Wij   Y ( Yactual  Y ) (1  Y ) H j 

n

Wij

 w j (1  H j ) X j



• Weight is between hidden layer and output:

w n1  w n   Y ( Yactual  Y ) (1  Y ) H j

j j





06/2000 Short-term load forecasting 56

Training of the

Feed Forward Neural Network



 In the equations:

• Yactual is the true value of the output load

•  is learning factor (0.3-0.8)

• n is the number of learning cycles

• Xj is the input value belongs to Yactual

A numerical example demonstrates the

use of neural forecasting method.

06/2000 Short-term load forecasting 57

Training of the

Feed Forward Neural Network



 The over training has to be avoided using

the cross validation method:

• The training set is divided into two parts.

(Part 1 : two years data, Part 2: one year

data)

• Part 1 is used to train the network, by

passing the data through the network.

• Few hundred times pass represents a

training period.



06/2000 Short-term load forecasting 58

Training of the

Feed Forward Neural Network



 The over training of the network has to

be avoided:

• Part 2 is used to check the effectiveness of

the training.

• After each training period the error is

calculated when the network is supplied by

the input data of Part 2.

• The increase of error indicates over training,

when the training has to be stopped

06/2000 Short-term load forecasting 59

Load Forecasting Engine



 The EPRI developed a Load Forecasting

Engine using 24 Neural networks.

 One network forecasts the load for each

hour of the day.

 The networks are grouped into four (4)

categories depending on time of the day.

 The categories have different inputs.



06/2000 Short-term load forecasting 60

Load Forecasting Engine



The construction of the engine shows the four

groups of neural networks.









06/2000 Short-term load forecasting 61

Load Forecasting Engine



 The four categories are:

• Category 1. Nine neural networks. Forecasts

the load between 1-9AM .

• Category 2. Nine neural networks. Forecasts

the load between 10AM -2PM and 7 -10PM .

• Category 3. Four neural networks. Forecasts

the load between 3-6 PM .

• Category 4. Two neural networks. Forecasts

the load between 11-12 PM .

06/2000 Short-term load forecasting 62

Load Forecasting Engine



 The inputs in four categories are:

• Category 1. Forecast for early morning

– general input

– load in the last three-four hours

– temperature in the last three-four hours

• Category 2. Forecast for off peak hours

– general input

– forecast temperature of previous hours

– yesterday’s load and temperature of hours close

to this hours

06/2000 Short-term load forecasting 63

Load Forecasting Engine



 The inputs in four categories are:

• Category 3. Forecast for afternoon peak hours

– general input

– forecast temperatures of previous and feature

hours close to this hours.

– yesterday’s load and temperature of hours close

to this hours

• Category 4. Forecast for late night hours

– general input

– forecast temperatures for the four proceeding

hours

06/2000 Short-term load forecasting 64

Load Forecasting Engine



 The general input variables are

• same hour load, temperature and humidity

of one day ago. (3)

• same hour load, temperature and humidity

of two days ago. (3)

• same hour load and temperature seven (7)

days ago (2)





06/2000 Short-term load forecasting 65

Load Forecasting Engine



 The general input variables are

(continuation):

• same hour forecast temperature and relative

humidity of the next day (2)

• day of the week index (Sunday 01, Monday

02 etc.)







06/2000 Short-term load forecasting 66

Load Forecasting Engine



 The load forecasting engine has one

output for the hourly load

 The extended forecast uses the forecasted

values. E.g. The two-day ahead forecast

uses values obtained by the one-day

ahead forecast.

 The forecast can be updated each hour

using the recent load and weather data

06/2000 Short-term load forecasting 67

Load Forecasting Engine



 The weights in the neural network are

adjusted daily.

 The retraining uses the actual load and

weather data of the past few days.

 The retraining helps to follow the trends,

changes of weather patter e.t.c





06/2000 Short-term load forecasting 68

Load Forecasting For Holidays

 The load during the holidays has

different patterns and is significantly

reduced.

 The forecast is inaccurate because of the

small number of historical data.

 The holiday is treated as

• Saturday if the shopping centers are open

• Sunday if the shopping centers are not open

06/2000 Short-term load forecasting 69

Hourly Weather Forecast



 The weather service provides forecasts

for:

• daily maximum and minimum temperature

• daily maximum and minimum relative

humidity

• rain and fog

• maximum wind speed

 No hourly data are provided.

06/2000 Short-term load forecasting 70

Hourly Weather Forecast



 EPRI developed an hourly temperature

and humidity forecasting engine.

 Single neural network with

• 28 inputs :

– hourly temperature of the previous day)

– high and low temperature of the two previous

day

• 24 outputs: expected hourly temperatures.

06/2000 Short-term load forecasting 71

Load Forecasting Results









06/2000 Short-term load forecasting 72

Load Forecasting Results



Comparison of forecasted and actual loads









06/2000 Short-term load forecasting 73

Load Forecasting Results



•Accuracy less than 3% for the next days forecast is

considered good

•The longer term forecast accuracy is less (7-8%)









06/2000 Short-term load forecasting 74

Appendix 1







Derivation of Learning

Algorithm



06/2000 Short-term load forecasting 75

Derivation of Learning

Algorithm

 The output of the hidden and output layer and the error

function are:

1 1

H  Y

j n h

  W X   w H

ij j j j

i1 j 1

1 e 1 e



1

E  ( X 2 ,n  Y ) 2

2





06/2000 Short-term load forecasting 76

Derivation of Learning

Algorithm

 The update of the weight factors require iteration

dE dE

dw j   dW i j  

dw j dW i j



 For the calculation of the derivative the following

substitutions are used

h

Y1    w H Y2  1  e Y1 u  ( X 2 ,n  Y )

j j

j 1

n

h1    W X h2  1  eh1

ij i

i1



06/2000 Short-term load forecasting 77

Derivation of Learning

Algorithm

 After substitutions the equations are:



1 1 1

H   

j I h1 h2

  W X 1 e

ij i

i1

1 e



1 1 1

Y  

J 1  e Y1 Y 2

  w H

j j

j 1

1 e

1 1

E  ( X 2 ,n  Y ) 2  u 2

2 2

06/2000 Short-term load forecasting 78

Derivation of Learning

Algorithm

 The derivation of the error function results in :



dE d  1 2  du du du

  2 u  dw  u dw  ( X 2 ,i  Y ) dw

dw j du   j j j









 The derivative of the u function is:



du d dY

 ( X 2 ,i  Y )  

dw j dw j dw j





06/2000 Short-term load forecasting 79

Derivation of Learning

Algorithm

 The derivative of the output function is:



dY



d

Y 2    2

1 dY 2 1 dY 2



1 dY 2



dw j dY 2 dw j Y 2 dw j (1  e ) dw j

Y1 2







1 dY 2 2 dY 2

 2

Y

 

dw j dw j

 J 



   w X 

 j i 



1 e



0 

 

 

 

 

 



 







06/2000 Short-term load forecasting 80

Derivation of Learning

Algorithm

 The derivation of the auxiliary function Y2 gives:



 w j H j

dY 2



d

dw j dY1



1  e Y1

dY1

dw j

 e Y1

dY1

dw j

e

dY1

dw j





 The derivative of Y1 function is:





dY1 d  J 

  wj Hj   Hj

dw j dw j  0

 





06/2000 Short-term load forecasting 81

Derivation of Learning

Algorithm

 Substituting the results in the equations :

J

  w H

j j

j0

e H

dE j

dw      (X  Y)

j dw 2, n 2

j  J 

   w H 

 j j

j0

1  e 

 

 

 



06/2000 Short-term load forecasting 82

Derivation of Learning

Algorithm

 The rearrangement of the output equation results in :

J

  w H

j j 1 Y

j 1 1

e  1

Y Y

 The final equation for the update of dwj is:

(1 Y ) H

dE j

dw    (X  Y)Y 2



j dw 2, n Y

j

(X  Y ) Y (1 Y ) H

2, n j

06/2000 Short-term load forecasting 83

Derivation of Learning

Algorithm

 The derivative of the hidden layer function is:



dH j



d

 h2 1  dh 2   1 2 dh 2   1 h1 2 dh 2 

dw j dh 2 dWi j h 2 dWi j (1  e ) dWi j

1 dh 2 2 dh 2

 2

  Hj

 

dWi j dWi j

 I 



   Wi j X 

 i 

1 e 0



 

 

 

 

 

 



 







06/2000 Short-term load forecasting 84

Derivation of Learning

Algorithm

 The derivation of the auxiliary function h2 gives:



 W



1  e 

Xj

dh 2 d dh1 h1 dh1

ij

dh1

 h1

e e

dWi j dh1 dWi j dWi j dWi j



 The derivative of h1 function is:





dh1 d  I W X    X

  ij j



 0 

j

dWi j dWi j





06/2000 Short-term load forecasting 85

Derivation of Learning

Algorithm

 Substituting the results in the equations :

I

  W X

ij i

e i0 X

dE i

dW      (X  Y ) Y (1  Y )

ij dW 2n 2

ij  I 

   W X 

1  e i  0 i j i 

 

 



 





06/2000 Short-term load forecasting 86

Derivation of Learning

Algorithm

 The rearrangement of the output equation results in :

I

  W X 1 H

ij i

i1 1 j

e  1

H H

j j

 The final equation for the update of dWij is:





dE

dW     (X  Y ) Y (1  Y ) H (1  H ) X

ij dW 2n j j i

ij

06/2000 Short-term load forecasting 87

Derivation of Learning

Algorithm

 Substituting the results in the equations which is used to

iterate the wj value :



dE k

Wikj1  Wikj  dW  Wikj   

ij k

dWi j

Wikj   ( X k ,n  Y k ) Y k ( 1  Y k ) H k ( H k  1) X k

2 j j i









06/2000 Short-term load forecasting 88

Derivation of Learning Algorithm

 The two training algorithms are:





w k 1

j  w   ( X  Y ) Y (1 Y ) H

k

j

k

2n

k k k k

j









Wikj 1  Wikj   ( X k ,n  Y k ) Y k ( 1  Y k ) H k ( H k  1) X k 

2 j j i





 Wikj  dw k  ( H k  1) X k

j j i









06/2000 Short-term load forecasting 89



Related docs
Other docs by cuiliqing
11.1 Exploring Area and Perimeter
Views: 0  |  Downloads: 0
Volusia County
Views: 2  |  Downloads: 0
choosing_topics_and_y10
Views: 0  |  Downloads: 0
CLE Credit - rscrpubs.com
Views: 2  |  Downloads: 0
Meeting Minutes September 8 Final
Views: 0  |  Downloads: 0
nov2411
Views: 3  |  Downloads: 0
EKG Spreadsheet - Geocities.ws
Views: 0  |  Downloads: 0
Gift from Christ to the Church
Views: 0  |  Downloads: 0
By registering with docstoc.com you agree to our
privacy policy

You are almost ready to download!

You are almost ready to download!