A novel sensorless position sensing for free-piston engine generation using observation coil

Accurate piston position information is crucial for the control and efficient operation of Free piston engine generator (FPEG). Conventional position sensors, however, are costly, which has restricted the development of FPEG. This paper proposes a novel sensorless position observation method based on the structural characteristics of FPEG. By leveraging observation coils to establish a relationship between magnetic flux variation and displacement, this proposed approach enables real-time monitoring of piston position. An equivalent magnetic network model is developed to analyze the flux variation within the observation coil, and a series of displacement observation experiments were conducted on a test bench. The results show that the calculated magnetic flux exhibits a strong linear relationship with the actual mover position. The proposed method achieves real-time position estimation across various operating strokes with a signal delay of less than 1 ms and a detection error below 4%. This technique demonstrates excellent potential for improving the performance, robustness, and commercial feasibility of FPEG systems.