40220802 (Power System Forecasting)Course Number: 40220802

Program: Undergraduate program

Type: Elective

Credits: 2

Term Offered: Spring

Prerequisite(s): Principles of Electric Circuits, Electric Machinery Fundamentals

Instructor(s): Chongqing Kang


Kang Chongqing, Xia Qing, and Liu Mei, Power System Load Forecasting, China Electric Power Press, 2007. (in Chinese)

Course Description:

The course is about forecasting methods and techniques, which is becoming attractive topics in power system. It covers the scale of long-term electricity demand forecasting, short-term load forecasting, ultra-short-term load forecasting, bus load forecasting, electricity market price forecasting, wind power forecasting. Forecasting principles and mathematical models are taught with examples in real power systems. The frontiers and new development of power system forecasting techniques are also introduced briefly in the course.

Course Objectives and Outcomes:

Numbers in brackets are linked to department educational outcomes

1.Students could learn the concept and theories of forecasting. [1]

2.Students should skillfully master forecasting methods and technology in practice of power system, Students should grasp the key model and data flow of the forecasting problem in power system. [10,11]

3.Students get trained in forecasting modeling capabilities and physical mechanism analyzing capabilities. [5,9]

4.Students have a better understanding of power system practical situation. [8]

Course Topics:

1.Basic concepts and principles of forecasting

2.Regression theory and forecasting problem

3.Forecasting methods based on timing trend extrapolation

4.Long-term load forecasting models and methods

5.The annual and monthly load forecasting

6.Time series ARIMA model

7.Load characteristics analysis and evaluation

8.Short-term load forecasting models and methods

9.Holiday load forecasting method

10.Bus load forecasting

11.Electricity price forecasting

12.Wind power forecasting

Experiment(s): Numerical experiments, which is the entitled as projects as follows.



l  Competitive load forecasting experiment

Based on the understanding of basic principles and mathematical model of load forecasting, student are supposed to design a forecasting model and implement it using programs. The load data given is from actual power system. The experiment includes long-term load forecasting short-term load forecasting. There is no standard answer (data identification, model choice, model parameters) to the problem, Project assessment is according to forecasting accuracy and the quality of experiment reports.

Course Assessment:

There are three ways of assessment in this course.

  Type A:

Final exam score, 90 points.

Attendance assessments, 10 points.

  Type B:

Competitive load forecasting experiment, 90 points.

Attendance assessments, 10 points.

  Type C:

Innovative research papers on forecasting models and methods, 90 points.

Attendance assessments, 10 points.