Innovative network planning for active electric distribution networks with advanced machine learning algorithms
In order to reach the energy strategy goals for 2050, a massive penetration of small-scale renewable energy sources (ie. solar panels) and a new type of consumers (ie. electric cars) should be integrated into the electric distribution network. Thus, increasing the uncertainty and complexity of generation and load forecasting. Due to these changes, traditional passive "worst-case" network planning methods will lead to extensive capital investments with a high probability that the resulted network would be largely overdesigned and underutilized. Actual restrictions for network operators are the computational time that electric network calculation requires and the lack of research in the computer science field. In this Thesis, new probabilistic network planning methods using Machine Learning approaches will be evaluated and a solution will be presented for more efficient, reliable, and cost-effective network planning.
Email: yamshid.farhat@bkw.ch
Related Literature:
[1] Artificial intelligence techniques for enabling Big Data services in distribution networks: A review. (A complete review of the existing research on the field of AI for electric distribution network)
[2] Power Grid Management via Semi-Markov Afterstate Actor-Critic. (A novel approach modeling the network operation problem as a reinforcement learning task. I would like to explore the potential of this idea for Network planning instead of operation)
[3] Statistical Machine Learning Model for Stochastic Optimal Planning of Distribution Networks considering a Dynamic Correlation and Dimension Reduction. (Implementation of Probabilistic Network calculations reducing the problem dimension with statistical machine learning theories)
[4] Neural Networks for power flow: Graph neural solver. (Design of a Graph NN Solver for electric power flow calculations able to massive improve time-performance in comparison with conventional Newton-Raphson solvers.)