An architectural framework is proposed for the design and construction of hybrid load forecasting systems for electric utilities. This framework consists of the intelligent techniques of artificial neural networks, fuzzy logic, knowledge-based and case-based reasoning. The knowledge-based system is the core of the integration since it is used to supervise the operations of the other intelligent techniques. Experts can also represent their knowledge in rules to refine and validate the results obtained from the other modules of neural networks and case-based reasoning. The framework was implemented on an object oriented real-time expert system shell G2 with General Diagnostic Assistant (GDA) and NeurOn-Line. In this environment, the intelligent techniques are encapsulated in blocks, which communicate with each other via data paths. The blocks can interact with rules in the knowledge base via rule-terminals. Procedures can be invoked by rules.
[1]
Michael T. Manry,et al.
Comparison of very short-term load forecasting techniques
,
1996
.
[2]
Robert J. Marks,et al.
Electric load forecasting using an artificial neural network
,
1991
.
[3]
Kwang-Ho Kim,et al.
Implementation of hybrid short-term load forecasting system using artificial neural networks and fuzzy expert systems
,
1995
.
[4]
Osama A. Mohammed,et al.
Practical experiences with an adaptive neural network short-term load forecasting system
,
1995
.
[5]
Rey-Chue Hwang,et al.
An adaptive modular artificial neural network hourly load forecaster and its implementation at electric utilities
,
1995
.