A Model-Based Short-term Load Forecast Methodology for Aggregated Power Consumption of Thermostatically Controlled Appliances in DSM

Yazdkhasti, Pegah; Diduch, C. P.

doi:10.1109/ecce47101.2021.9595608

Cited by 2 publications

(3 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This equation mirrors (2), where H a (0) = α and H s (0) = β. Now, by approximating the aggregated demand with a LTI system, the dynamic response of the system can be captured by [24]:…”

Section: Aggregated Loadmentioning

confidence: 99%

“…The details of the model learning process is presented in [24]. As a short summary, the learning procedure initiates when the forecast error surpasses a predefined threshold level.…”

Section: Learning Modelmentioning

confidence: 99%

“…Determining H s when a small change is applied to T s , and the effect of changes in T a can be estimated and removed using H a . The formulations for these steps are elaborated in detail in [24].…”

Section: Learning Modelmentioning

confidence: 99%

See 2 more Smart Citations

Smart Agent-Based Direct Load Control of Air Conditioner Populations in Demand Side Management

Yazdkhasti,

Cárdenas–Barrera,

Diduch

2024

Adv. sci. technol. eng. syst. j.

Self Cite

View full text Add to dashboard Cite

The integration of fluctuating renewable resources such as wind and solar into existing power systems poses challenges to grid reliability and the seamless incorporation of these resources. To address the inherent variability in renewable generation, direct load control emerges as a promising method for demand-side management. Thermostatically controlled appliances, like air conditioners, hold a significant role in this approach. However, effective direct load control necessitates accurate load magnitude estimation and the potential for load shifting. In this paper, we introduce a smart-agent architecture that employs a mathematical model to forecast aggregated power consumption behavior, even when changes are introduced by the controller. To assess system performance, a numerical simulator was developed, demonstrating the system’s adaptability to changes, its self-retraining capability, and its continuous improvement in predicting aggregated power consumption.

show abstract

“…This equation mirrors (2), where H a (0) = α and H s (0) = β. Now, by approximating the aggregated demand with a LTI system, the dynamic response of the system can be captured by [24]:…”

Section: Aggregated Loadmentioning

confidence: 99%

“…The details of the model learning process is presented in [24]. As a short summary, the learning procedure initiates when the forecast error surpasses a predefined threshold level.…”

Section: Learning Modelmentioning

confidence: 99%