Wednesday 19 March 2025
The quest for efficient electric motor control has been a long-standing challenge in the field of engineering. Electric motors are ubiquitous, powering everything from household appliances to industrial machinery, and their efficient operation is crucial for reducing energy consumption and mitigating environmental impact. In recent years, researchers have turned to artificial intelligence (AI) to improve the performance of electric motor controllers.
One such approach involves using neural networks, a type of AI that can learn and adapt to new situations. Neural networks have been successfully applied to various fields, including image recognition and natural language processing. Now, scientists have developed a tiny neural network (TinyNN) specifically designed for microcontrollers, the small computers found in many modern devices.
The TinyNN is optimized for use in electric motor control systems, allowing it to learn from data and make adjustments in real-time. This enables the controller to adapt to changing conditions, such as fluctuations in power supply or variations in motor load. The result is more efficient operation, reduced energy consumption, and extended lifespan of the motor.
To test the TinyNN, researchers implemented it on a microcontroller-based development board and connected it to a permanent magnet synchronous motor (PMSM). PSMCs are commonly used in applications such as electric vehicles, industrial machinery, and renewable energy systems. The team then simulated various scenarios, including changes in speed, load, and power supply.
The results were impressive. The TinyNN was able to reduce overshoots and deviations in the motor’s response by up to 87.5% compared to traditional proportional-integral (PI) controllers. PI controllers are widely used in electric motor control systems due to their simplicity and ease of implementation. However, they have limitations, including a tendency to oscillate or undershoot.
The TinyNN’s performance was evaluated using various metrics, including mean squared error (MSE), which measures the difference between the desired output and the actual output. The TinyNN achieved an MSE of 0.02 in one test case, compared to an MSE of 1.21 for the PI controller.
While the results are promising, there is still much work to be done before TinyNNs can be widely adopted in electric motor control systems. For example, researchers need to develop more efficient algorithms and optimize the network’s architecture for specific applications.
Despite these challenges, the potential benefits of using AI in electric motor control systems are significant.
Cite this article: “Artificial Intelligence Boosts Electric Motor Efficiency”, The Science Archive, 2025.
Electric Motor Control, Artificial Intelligence, Neural Networks, Microcontrollers, Permanent Magnet Synchronous Motors, Proportional-Integral Controllers, Mean Squared Error, Electric Vehicles, Industrial Machinery, Renewable Energy Systems
