Browsing by Author "Paudyal, Sumit"

Now showing 1 - 4 of 4

Data-Driven Local Control Design for Dead Band Control of Load Tap Changers
(IEEE, 2024) Savasci, Alper; Ceylan, Oguzhan; Paudyal, Sumit
This study presents an off-line optimization-guided machine learning approach for coordinating the local control rules of on-load tap changers (OLTCs) and step-voltage regulators (SVRs). Based on a bang-bang control rule, these legacy devices autonomously regulate the feeder voltage around the nominal level by varying the tap position in the lower or raise direction. The characterizing parameter of the local control rule is the dead band, which affects the number of tap switching in operation and is directly related to the economical use life of the equipment. The bandwidth is typically set within a standard voltage range and is generally kept constant in daily operation. However, adjusting the bandwidth dynamically can prevent excessive tap switching while maintaining satisfactory voltage regulation for varying loading and distributed generation conditions. Our approach aims to set the bandwidth parameter systematically and efficiently through a machine learning-based scheme, which is trained with a dataset formed by solving the distribution network optimal power flow (DOPF) problem. The performance of learning the bandwidth parameter is demonstrated on the modified 33-node feeder, which is promising for integrated voltage control schemes.
Citation - Scopus: 2
Data-Driven Methods for Optimal Setting of Legacy Control Devices in Distribution Grids
(IEEE Computer Society, 2024) Savasci, Alper; Ceylan, Oǧuzhan; Paudyal, Sumit
This study presents machine learning-based dispatch strategies for legacy voltage regulation devices, i.e., onload tap changers (OLTCs), step-voltage regulators (SVRs), and switched-capacitors (SCs) in modern distribution networks. The proposed approach utilizes k-nearest neighbor (KNN), random forest (RF), and neural networks (NN) to map nodal net active and reactive injections to the optimal legacy controls and resulting voltage magnitudes. To implement these strategies, first, an efficient optimal power flow (OPF) is formulated as a mixed-integer linear program that obtains optimal decisions of tap positions for OLTCs, SVRs, and on/off status of SCs. Then, training and testing datasets are generated by solving the OPF model for daily horizons with 1-hr resolution for varying loading and photovoltaic (PV) generation profile. Case studies on the 33-node feeder demonstrate high-accuracy mapping between the input feature and the output vector, which is promising for integrated Volt/VAr control schemes. © 2024 Elsevier B.V., All rights reserved.
Citation - WoS: 1
Citation - Scopus: 2
Optimal Dead Band Control of Occupant Thermostats for Grid-Interactive Homes
(IEEE, 2024) Savasci, Alper; Ceylan, Oguzhan; Paudyal, Sumit
Efficient and grid-aware management of home-scale heating, ventilation, and air conditioning (HVAC) systems is one of the key enablers of demand-side management (DSM) and associated grid services in the residential sector. HVACs regulate the indoor temperature around a set point through a thermostat operating within a closed-loop control scheme. Conventional thermostats typically have a built-in temperature dead band or differential where the thermostat is idle, and HVAC stays at the most recent state (On/Off). The temperature dead band is an important control parameter that can help save energy as well as preventing frequent On/Off switching cycles leading to excessive wear and tear on the equipment. However, strategic and dynamic adjustment of the dead band can be a challenging task for an occupant. This paper proposes a mixed-integer linear program (MILP)-based tuning scheme to optimally determine the dead band. The novelty in this formulation is the inclusion of thermostat hysteresis curve modeled by piecewise techniques for tuning the dead band accurately. The proposed formulation is solved as a receding horizon manner for normal as well as under a demand response (DR) event and has been found it can achieve up to 10% reduction in energy consumption without degrading the regulation performance significantly.
Supervised Learning-Driven Dead Band Control of Occupant Thermostats for Energy-Efficient Residential HVAC
(Elsevier, 2026) Savasci, Alper; Ceylan, Oguzhan; Paudyal, Sumit
Heating, ventilation, and air conditioning (HVAC) systems play a crucial role in demand-side management (DSM) by shaping residential electricity consumption and enabling flexible, grid-responsive operation. Thermostats in HVAC systems regulate indoor temperature as part of a closed-loop control framework, typically incorporating a fixed temperature dead band-a range around the setpoint where no action is taken-to reduce energy use and prevent frequent cycling of the HVAC system. Although essential for efficiency and equipment longevity, fixed dead bands limit adaptability, as dynamically adjusting them under varying environmental conditions remains challenging for occupants. To address this limitation, we propose a machine learning (ML)-based dead band tuning framework that optimally adjusts thermostat settings in real time. The method integrates conventional optimization with data-driven modeling: a mixed-integer linear programming (MILP) model is first used to gen erate optimal dead band values under measured outdoor temperature records (diverse seasonal weather scenarios) which are then employed to train the ML-based predictor to learn a real-time discrete dead band decision policy that approximates the MILP-optimal hysteresis-aware decisions. Among the evaluated models, Random Forest demonstrates superior predictive performance, achieving a mean squared error (MSE) of 0.0399 and a coefficient of determination (R2) of 95.75 %.