Research Projects

The integration of behind-the-meter (BTM) load and generations poses significant challenges for operating electric distribution grids, particularly under limited visibility due to restricted access and private ownership, increasing uncertainty in managing generation and load during normal and outage scenarios. Existing methods assume complete network observability, synchronized updates, and accurate data on BTM loads and distributed energy resources (DERs), which are often unavailable in practice, especially in low-voltage grids. The diverse operating modes of DERs, variability in appliances, and limited deployment of smart meters and sensors further deteriorate this issue, making networks partially visible. This needs stringent research for managing such networks, enabling efficient operation and restoration during pre- and post-outage conditions. This is crucial for enhancing grid resilience and reliability globally, including in India. Objectives: 1. Enhanced situational awareness using data-driven estimation of network topology and BTM resources with minimal sensor data. 2. Multi-period robust frameworks for minimizing network losses while managing real-time uncertainties in load and DER power generation. 3. Rapid detection of outages to reduce downtime and enhance system reliability. 4. Swift power restoration in de-energized or islanded sections via network reconfiguration and DER coordination, improving resilience without overburdening the grid. Hypothesis: 1. Limited deployment of smart meters and measuring devices is sufficient to develop effective management strategies for partially observable networks. 2. Integrating data-driven estimation with robust optimization can enable efficient operations and full observability of DNs under practical constraints. Research Works: 1. WP1: Data collection and design model based reference framework 2. WP2: Data-driven estimation of BTM resources and topology 3. WP3: Robust loss minimization algorithm for pre-outage conditions 4. WP4: Outage detection and fast post-outage power restoration 5. WP5: ADNMS testbed setup and HIL validation. Deliverables and their significance: 1. An efficient data-driven estimation algorithm by merging both supervised and unsupervised learnings to enhance situational awareness 2. Robust optimization frameworks that minimize network losses and reduce load restoration times in pre- and post-outage scenarios, respectively. 3. An efficient ADNMS laboratory testbed for successful demonstration with standard and real-life DNs If successful, the project will advance the fundamental understanding of managing partially observable DNs and significantly contribute to practical applications. It will enable DNs to operate efficiently under uncertainty, enhance resilience against outages, and provide scalable solutions for grid modernization. The findings could shape the future of DN management globally, particularly in regions with limited resources for widespread sensor deployment