Artificial intelligence-powered controllers that mimic the human brain could strengthen the grid

As traditional power plants are replaced by intermittent sources such as solar and wind, maintaining grid stability has become a complex engineering challenge. hussein khanA doctoral thesis at the University of Vaasa, Finland, presents an advanced AI-based control strategy to ensure the reliability and resilience of local power grids.

The power system is undergoing a major transformation as fossil fuel-based power generation is gradually replaced by inverter-based renewable energy. This change introduces inherent uncertainties and low inertia that significantly complicate grid operation and voltage stability in AC and DC microgrids.

Hussein Khan tackles these challenges in his doctoral thesis in electrical engineering. By using artificial neural networks (ANN), Khan has developed a controller that can predict and compensate for changes in the grid in real time, outperforming traditional control methods.

– ANN is inspired by the human brain, which processes information through interconnected neurons. This biomimetic approach allows the system to learn from different scenarios and adapt to the unpredictability of solar and wind power generation, Kahn said.

Cost-effective solution with sensor optimization

Traditional systems rely on multiple physical sensors to monitor voltage, current, and other parameters, increasing cost and increasing the number of potential points of failure. Khan’s AI-driven approach demonstrates that sophisticated software can compensate for fewer hardware components.

– By effectively training the neural network, the system can provide the same reliable results with just one sensor instead of two. This results in fewer physical parts that can fail, leading to cost optimization and improved overall reliability, Kahn said.

While AI-based control improves efficiency and reduces hardware requirements, introducing intelligent controllers into critical infrastructure also brings new considerations.

– The main concern is that AI behaves like a black box. You can see the inputs and outputs, but it doesn’t always fully explain what’s happening under the hood. Still, the controller performed very well in our tests and was rigorously validated in real time, Khan said.

Khan’s research supports the broader goal of creating a carbon-neutral energy system in the coming decades. AI-based control could enable power grids to integrate more renewable energy in the future by improving stability and reducing hardware requirements.

dissertation

Khan, Hussein (2026) Advanced predictive and AI-based converter control strategies for AC and DC microgrids. Acta Wasaensia 580. Doctoral dissertation. University of Vaasa.

Publication PDF