A Novel Adaptive Supervisory Controller for Optimized Voltage Controlled Demand Response

E. Abraham, H. Marzooghi, J. Yu, V. Terzija

    Research output: Contribution to journalArticlepeer-review


    Recently, it has been shown that voltage-controlled demand response (VCDR), which is based on the principle of the dependence of the active power of demand on the system voltage, can provide ancillary services to future power systems. Voltage control devices used for VCDR can improve system frequency stability to various extents, depending on their locations, load size and load-voltage dependency in the grid. Thus, designing a supervisory controller to guarantee that such devices are optimally utilized for VCDR is necessary, as this can contribute in enhancing system frequency stability. In this paper, a novel adaptive supervisory controller (ASC) is proposed to optimally use VCDR resources in large power systems for such purposes. To ensure the effective operation of the ASC, an assessment of the impacts of on-load tap changer (OLTC) transformers on the system frequency is essential. In this regard, clustering techniques and principal component regression are used as offline tools to evaluate the influences of OLTCs transformers on VCDR in large-scale power systems using the IEEE 39-bus test system. Also, to estimate the effects of the OLTC transformer clusters on VCDR, a comprehensive sensitivity analysis with respect to the gain of the modified OLTC controllers' frequency input is conducted.

    Original languageEnglish
    Pages (from-to)4201-4210
    Number of pages10
    JournalIEEE Transactions on Smart Grid
    Issue number4
    Early online date2 Jul 2018
    Publication statusPublished - Jul 2019


    • Frequency control
    • MIGRATE project
    • k-means clustering
    • model predictive control
    • principal component analysis
    • voltage-controlled demand response


    Dive into the research topics of 'A Novel Adaptive Supervisory Controller for Optimized Voltage Controlled Demand Response'. Together they form a unique fingerprint.

    Cite this