Causality Theory & Advanced Machine Learning in Power System Applications

With the recent advances in the field of Machine Learning, there have been several developments in modern power systems computations. Meanwhile, causality has also received a lot of recognition in various fields such as economics, social sciences, biology, etc. In this thesis, we focus on integrating causal theory into power system applications also utilizing machine learning algorithms for validation and predictive modeling in the areas of resilience and power system planning. This thesis proposes a novel causality analysis approach called the Causal Markov Elman Network (CMEN) integrated with deep neural networks to characterize the interdependencies and interrelationships between various heterogeneous time-series from multi-network infrastructure networks. The CMEN performance, which comprises of inputs filtered by Markov property, successfully characterizes various multivariate dependencies in an urban environment. The thesis also proposes a novel hypothesis of characterizing joint information between interconnected systems such as electricity and transportation networks. The proposed methodology and the hypotheses are then validated by Information Theory distance-based metrics. To optimize the performance of CMEN, a deep learning algorithm is also adopted and named as Deep Neural Network Causal model (DNNC). For cross-validation, the CMEN & DNNC causal models are applied to a case study application of the electricity load forecasting problem using actual data from the City of Tallahassee, Florida. Another case study application of the proposed causal methodology is characterizing the co-dependency between different infrastructure networks based on real-world data from Hurricane Hermine and Hurricane Michael.