A closed-loop current measurement system and method based on magnetic balance principle

Abstract

This invention belongs to the field of current measurement, specifically relating to a closed-loop current measurement system and method based on the principle of magnetic balance. The system includes a magnetic core assembly, an induction coil assembly, a magnetic field detection unit, a signal conditioning unit, a digital twin modeling unit, a neural network processing unit, a feedback compensation unit, an active demagnetization control unit, and a power drive unit. The digital twin modeling unit maps the virtual state of the physical magnetic core in real time; the neural network processing unit utilizes transfer learning to deeply integrate prior physical knowledge with real-time data, accurately predicting and correcting nonlinear deviations in the microscopic magnetization state. This invention, through the deep integration of physical logic and artificial intelligence algorithms, achieves dynamic and accurate compensation for hysteresis effects and saturation characteristics, improving measurement linearity and zero-point stability over a wide temperature range. While reducing hardware costs, it ensures the system's dynamic response performance and environmental adaptability in high-frequency harmonic current scenarios.

Images