Learning-assisted Diagnosis and Control of Electric Distribution Network

The digital transformation of modern power systems enables more robust and transparent system planning and operation. This process is highly reliant on big data, which is widely available through energy generation to utilization. The distribution networks are exposed to high challenges due to the massive integration of renewables and electric vehicles at low voltages, but opportunities do exist due to the widely available infrastructure for data collection and smart control at the distribution level. For example, the phase measurement unit (PMU) for better system observation in the transmission system has been extended to the distribution network, where standard protocols are available for high precision data transmission in a sub-second timescale. Artificial intelligence (AI) technologies become the main enabler to utilize these data. This provides revolutionary solutions covering fault diagnosis, control and almost all aspects of modern electric power systems. The values of AI in distribution networks are widely recognized especially for the increased complexity of distribution systems, where active management and autonomous operation are urgently needed. The typical mechanism-based modeling and model-based control lack adaptability under an evolving environment, e.g., network topology changes that are more frequent at the distribution level. However, the existing AI technologies as well as the associated machine learning algorithms cannot be directly transplanted into the area of power systems. To fulfill the strict technical requirements in power systems for security, generalization capability, interpretability, out-of-sample performance, robustness and reliability are the main issues to be addressed before their massive deployment.

Therefore, this Research Topic aims to provide a contribution in line with this area of research, i.e., innovative approaches and schemes for learning-assisted diagnosis and control for the medium to low voltage distribution networks. Original work focusing on the applications of state-of-art AI technologies in the power and energy related fields are invited. Both high-quality Original Research and Review Articles are welcome about the latest progress and potential research applications of the relevant areas with interests in modeling, control, monitoring and diagnosis of the distribution network using AI-related techniques.

Potential topics include but are not limited to the following:
• Data-driven operation and control of the distribution network
• Optimal Planning of renewable and energy storage systems in distribution network
• Fault diagnosis and location of distribution network using the learning-assisted approach
• Data-driven state estimation of electric distribution network
• Information and communication technologies (ICT) enabling AI in distribution network