Chapter 4 - A smart and efficient IoT-AI and ML-based multifunctional system for multilevel power distribution management

Abstract

With the integration of machine learning (ML), artificial intelligence (AI), and Internet of Things (IoT), incredible automation in energy generation, transmission and utilization process becomes more intelligent and smarter. So, it is the ultimate time to move toward renewable sources of energy from non-renewable energy sources and it is necessary to adopt energy efficient policies. We are currently dependent on coal-fired energy generation, which leaves a heavy carbon and ash foot-print on the atmosphere. This chapter provides an in-depth insight about how the consumer pattern of usage behavior. From the consumption study, the hour-by-hour demand forecasting can be carried out to provide outage less supply using various traditional, ML, and deep learning-based forecast algorithms. An IoT-AIML based architecture is proposed in this chapter toward energy management, outage management, power loss, and fault and power theft detection. A new feature is incorporated using AI and ML integration that describes how a consumer can be educated in terms of their non-prioritized electrical appliances during peak consumption hours and suggests to shift the consumption for later. Thus, the consumers can give grant for those appliances to be in rest, so that demand can be reduced up to a remarkable extent during peak demand hours. By performing analysis on the electrical equipment’s discriminately, each of its load is analyzed and helps in usage switching. Health of the distribution transformers and power quality are monitored remotely as well and various fault prediction and the architecture implementation feasibility has been discussed. The implementation of this system will turn power grids into smart grids. This chapter also provides an insight toward pole transformer monitoring, effective implementation, and negotiation of consumption performed by smart sensing meters. It also discusses the feasibility of distributional equipment’s health detection and prediction.