Carles
Pedrals
Hi! I am Carles and I am a Engineering Physics graduate and Software Engineering major. Below, you will find a summary of my education, experience and some developed projects!
Skills
- - AI Development tools, Pytorch, Pytorch Geometric, TensorFlow
- - Quantum Computing algorithms development, Qiskit
- - Developed also in Java, NodeJS and others
- - Web development, HTML, CSS, JS, React
- - People skills and leadership working asa Restaurant Manager
Relevant Experience
Development of a custom energy consumption forcasting AI model for microgrids and individual users in a national power-grid.
Design of quantum algorithms for logistic problems applications.
Lead developer in a program utilizing AI for traffic forecasting using realtime information on car flux per road and from the traffic light configuration
Implementing this AI-driven tool to facilitate decision-making in the traffic light control room during a pilot trial.
Portfolio
This project is operational, and can be accessed at the url above. For a full showcase of functionalities, credentials are needed. Contact me via the email provided in my CV!
More accurate description on the tool is provided here
QAOA algorithm is tested on an energy grid problem, aiming at the allocation of renewable energy sources on the Scandinavian power grid and evaluating different strategies for performing parameter optimization of QAOA ansatz circuits. This process combines fine-tuning parameters using both quantum and classical methods. Multiple extensions inspired by literature such as the Multiangle Approach, initial parameter distributions, and diverse methodologies are explored and their efficiency is discussed, highlighting Gaussian distribution as a great naive option. A new approach to improve convergence using neural networks is proposed and many techniques are applied to envision an efficient quantum algorithm capable of handling real-time demand of power grid partitioning.