Subbhodeep Basu

– Hands-on experience in end-to-end data flow in an Industry 4.0 environment with the Niagara emalytics platform for IoT – Ensured historical energy consumption data quality by developing custom components for IoT based building management system – Reduced manual effort and time requirements to synchronize sensor readings by developing a NumericalOffset extension – Automated energy flow visualization by building custom data pipelines for Sankey Diagrams – Documentation in Sphinx for all developed components

Topic: Optimal power scheduling for hybrid grid-connected networks with machine learning. - Designed a optimised power scheduling systems based on Machine Learning and Reinforcement learning and spot prices of electricity in Germany -Simulated Battery operation (Charging and discharging ) -Simulated output of a planned PV plant -Load and PV generation forecasting with RandomForest and XGBoost -Results indicate possible 40% cost reduction and 17% reduction in grid dependency

Client: Northwestern Mutual – Ensured data quality and consistency as a Database administrator for a Business-critical SQL database. – Resolved customer requirements by creating targeted SQL queries for complex data requirements. – Automated manual workload as a part of value-added projects using technologies like React, NodeJS, and Selenium – Supported business-critical applications for the client, working with diverse tech-stacks and Monitoring/Dashboard technologies such as - Kibana, Dynatrace and Vigil.

– Analyzed short- and long-term storage technologies (batteries, hydrogen, CAES, etc.) for the integration of renewable energies. – Evaluated the contribution of storage systems to grid stability, security of supply and CO2 reduction in Germany (with a focus on NRW and Paderborn). – Investigated political framework conditions, economic factors and technological trends for a sustainable energy storage energy storage strategy until 2040.

– Developed a SSVEP-based BCI system to control a GoPiGo robot using EEG signals. – Implemented signal preprocessing (Butterworth filter, ICA), feature extraction (DWT, PSD) and dimension reduction (PCA). – Trained classification models (SVM, KNN) and deep learning models (LSTM-RNN) for the recognition of control commands – Designed a user-friendly GUI for data acquisition, visual stimulation, model training and robot control.

Specialisation - Signal and information processing Courses: Advanced System Theory, Modelling and Simulation, Physical Systems for AI, Management of Technical Projects, Advanced Control, Data Driven Innovation and Engineering, Statistical Machine Learning, Statistical Signal Processing, Digital Image processing, Optimisation Based Control, Solar Electric Energy Systems, Energy Transition

-Project Goal: Design and simulate a DC drive system powered by a buck-boost converter, with speed control of a DC motor via Arduino-generated PWM signals. -Core Concept: A buck-boost converter modifies DC input voltage to a higher or lower output voltage, controlled by switching operations (MOSFET) governed by Arduino PWM. -Implementation: The system was modeled using MATLAB (for power circuit simulation) and Proteus (for PWM signal generation), achieving stable motor speed and torque.

Courses: Electrical Circuit Design, Control Systems, Power Systems, Electrical Machines, Digital Electronics, Digital Signal Processing, AI and Soft Computing, Data Structures and Algorithms, Database Management Systems, C Programming, Power Generation Economics, Engineering Graphics, Power Electronics, Electrical Drives