Formulation and analysis of a rule-based short-term load forecasting algorithm
Article Abstract:
Load forecasting, an important factor in electric utility operation, is expected to become more complex. The forecasting technique must be intelligent, sensitive to changes in the environment, and adaptive. Meeting these requirements requires a knowledge-based load-forecasting approach. The algorithm predicts the load for a one-hour to 168-hour time frame by using the expert systems knowledge and historical relationship between the load, temperature, wind speed, day type, and hours of the day. The algorithm is based on load and weather variables, day types, and seasons. The seven-day algorithm's data is made up of four weeks of hourly historical load and weather data plus one week's predicted weather information. The one-day algorithm's data is made up of load and weather information for the forecast day and the previous seven days. The seven-day load forecast's main use is for unit commitment and hydrothermal scheduling.
Publication Name: Proceedings of the IEEE
Subject: Electronics
ISSN: 0018-9219
Year: 1990
User Contributions:
Comment about this article or add new information about this topic:
On a constrained LMS dither algorithm (technical)
Article Abstract:
An LMS algorithm is introduced that reduces the problem of signal cancellation in adaptive array systems when the jammer power is relatively large and the jammer is correlated with the desired signal. The LMS algorithm exhibits steady-state performance similar to that of the Widrow dither algorithm, but mechanical motion of the receiver array is not required. The new algorithm might well be useful in beam-forming applications because of its excellent performance.
Publication Name: Proceedings of the IEEE
Subject: Electronics
ISSN: 0018-9219
Year: 1987
User Contributions:
Comment about this article or add new information about this topic:
Analysis of an algorithm that resolves oscillatory data into smooth components
Article Abstract:
The Ksienski algorithm does not accurately extract smooth functions from oscillatory data, but reuse of the algorithm for interpolation of the resulting smooth functions can produce good results. Achieving accuracy, though, requires that the original functions to be linear. Knowledge of the frequency of the oscillations is also required to generate smooth mean and amplitude functions. Equations, utilization, and validation of the algorithm are described.
Publication Name: Proceedings of the IEEE
Subject: Electronics
ISSN: 0018-9219
Year: 1988
User Contributions:
Comment about this article or add new information about this topic:
- Abstracts: Status of device-qualified GaAs substrate technology for GaAs integrated circuits. The development of the first LSI GaAs integrated circuits and the path to the commercial market
- Abstracts: Design and performance analysis of adaptive optical telescopes using laser guide stars. Sodium resonance flourescence lidar applications in atmospheric science and astronomy
- Abstracts: Solution and linear estimation of 2-D nearest-neighbor models. The Solution of Municipal Rapid Transit
- Abstracts: On the generation of random numbers from heavy-tailed distributions
- Abstracts: Networks for approximation and learning. Fast frequency synthesis by PLL using a continuous phase divider. Chebyshev Phase Approximation