Prospect of Adapting Artificial Intelligence in Smart Grids for Developing Countries

By Vishalya Sooriarachchi

With the increased application of renewable energy in the energy market and the continuous variations undertaken to implement a highly controllable and interactive power system, the present energy systems are becoming extremely complex. The development of deep learning and core technologies such as multicore processors and math accelerators for neural networks, has made Artificial Intelligence (AI) a massively industrialized component in the modern technological arena. To support the existing systems, and to extend the flexibility and applicability of Smart Grids, AI has been therefore naturally adapted. However, though the adaptation of such technologies is highly recommended by the developed countries, the prospect of the developing countries in terms of redefining the existing energy system and adapting AI and other such technologies to provide improved flexibility in the system is yet at a considerably low level. Although the implementation of AI related technologies in the existing energy grids of developing countries is constituted with a number of complications due to the lack of data samples, imperfections in the available infrastructure, and reliability issues, an AI enabled smart grids can enable the optimization of the power supply, analysis of behavior in the power usage by the users, and predetermined diagnosis of faults.

Artificial Intelligence and Smart Grids

The electric power grid has seen numerous changes and developments according to the requirements of mankind and technological advancements in the world. The developments procedure has been highly integrated with the various research findings, as well as to facilitate the power requirements, balancing the energy demand, increasing the efficiency of the system, reducing energy wastage, minimizing the carbon footprint, and increasing cost effectivity.

With the increasing developments seen in the field of Information Technology and its related field, the classic power grids are currently turning into much smarter adaptations. Thus, in the pathway of progress adaptions such as smart grids and microgrids have begun to play an important role, yet however, at present, the concepts are further developing. The smart functionality within the energy grid undoubtedly plays the major role in the present context, but further increasing its efficiency in terms of cost and energy management is a major goal.

Hence, Artificial Intelligence (AI) has been deemed suitable in balancing the control and supply of energy within the network. It has been able to improve the performance of smart grids in terms of dynamic clustering, increasing price evolution, improving fault isolation mechanisms, creating self-organization capabilities, and adapting self-diagnosis techniques in the system.

Challenges, Benefits, and the Role of AI in Smart Grids

The past rush in developing the Artificial Intelligence related activities and adapting them into the respective revenue and cost saving operations, has begun the rise of a novel energy network scheme which is heavily powered by concepts and related methodologies of AI [1]. Novel questions and challenges arose through the adaptation such as uncertainty in forecasted loads, errors in calculations, and the requirement of high computations resources. It was also capable of providing more advanced methodologies and energy supply systems which includes autonomous decision-making capabilities, insensitiveness to the market structure, and the process adaptation of providing better results in high constrained optimization problems.

Most of the developed countries in the world are already adapting AI based technologies for control and monitoring of the systems while also using them in aspects such as [2], i. addressing the changes in the power grid required to improvise to the supplemental changes by cutting energy waste, facilitating and accelerating the use of clean and renewable energy sources, and improving the planning, operation, and control of the power systems, ii. information sharing and two-way communication, iii. Configuration of ownership and systems through energy suppliers and iv. Net metering and related activities.

The use of the automated and AI-based system has tremendously affected the effectivity, accuracy (dependent upon the provided data), and the efficiency of the power system while enabling the embracing of a cyber-physical combination onto the energy systems.

Some of the adapted AI techniques in the smart grid and its related developmental procedures includes [1];

• Managing the grid users and controllers
• System based operation strategies for the grid
• Power supply optimization
• Consensus-based intelligent distribution techniques
• Machine learning and deep learning enabled costing mechanisms
• Intelligent energy storage systems
• Intelligent voltage profile regulation techniques using smart algorithms

Prospect for Developing Countries

The grid network in most of the developing countries are decades-old and comprise of techniques that are extremely inefficient. For example, most of the countries are unable to locate the exact positions of power breakdowns and outages unless the consumers inform the relevant authorities. In such circumstances, the implementation of the smart grids has been identified as a much effective and more appropriate methodology to support the requirements of the countries.

Many developed countries, especially those in the Asian region, have implemented smart grids in order to suffice to the requirements of the energy users while implementing mechanisms to lower the man power spent on the provision of proper energy supply to the community.

However, considering the requirements of the society, the increasing population, and the massively increasing and developing industries and related technological perspectives in the countries, it is more viable to ensure the smart grid is enhanced with more sophisticated and advanced techniques to further improve the system [3]. Unlike developed countries, the developing countries have the opportunity to embrace the techniques that have already been implemented in the major countries in the world, thereby enhancing the pathway of progression of the energy network by employing multi-directional and multi-channel networks to enhance the one-way or two-way flow of energy (along with the related communication and such mechanisms) among energy suppliers and customers or prosumers [4].

References

1. S. Ali and B. Choi, "State-of-the-Art Artificial Intelligence Techniques for Distributed Smart Grids: A Review", Electronics, vol. 9, no. 6, p. 1030, 2020. Available: 10.3390/electronics9061030 [Accessed 23 October 2020].
2. J. Li, Y. Zhao, C. Sun, X. Bao, Q. Zhao and H. Zhou, "A Survey of Development and Application of Artificial Intelligence in Smart Grid", IOP Conference Series: Earth and Environmental Science, vol. 186, p. 012066, 2018. Available: 10.1088/1755-1315/186/4/012066 [Accessed 23 October 2020].
3. N. Bassiliades and G. Chalkiadakis, "Artificial Intelligence Techniques for the Smart Grid", Advances in Building Energy Research, vol. 12, no. 1, pp. 1-2, 2017. Available: 10.1080/17512549.2017.1410362 [Accessed 23 October 2020].
4. A. Ferreira, P. Leitao and P. Vrba, "Challenges of ICT and artificial intelligence in smart grids", 2014 IEEE International Workshop on Intelligent Energy Systems (IWIES), 2014. Available: 10.1109/iwies.2014.6957039 [Accessed 23 October 2020].

This article edited by Mehrdad Boloorchi

For a downloadable copy of the November 2020 eNewsletter which includes this article, please visit the IEEE Smart Grid Resource Center.

Vishalya Sooriarachchi is an undergraduate student of Sri Lanka Technological Campus (Sri Lanka). She will be receiving her B.Sc. degree in Electronics and Power Systems Engineering in 2020. Her undergraduate research is on the use of “Demand Response Management in Sri Lanka for power crisis”. She has also presented a research on “Adapting Micro-grid based distributed generation in Sri Lanka” for the International Research Conference, held at SLTC, 2020. Her current research interests are application of Control systems and AI in energy industry, development of Smart grids and distributed generation, renewable energy integration, sustainable development and energy management. She is hoping to pursue a career in research specializing Smart Grids and energy management.