Super-Twisting Fuzzy Sliding Mode Observer (ST-FSMO) for Piston Position in Stirling Linear Motors

Free-piston Stirling linear generation systems are widely used for thermal power generation. Due to operation in high-pressure environments and compact internal structures, reliable displacement sensors cannot achieve real-time position detection. During system operation, controlling the mover’s motion position is indispensable for ensuring safety. This paper implements a sensorless observation method for position information based on Stirling generation system principles, which establishes a super-twisting sliding mode observation model using the linear motor, adopts an integral terminal sliding surface to reduce errors, utilizes a sigmoid function as the switching function to decrease chattering, simultaneously, employs fuzzy control based on the sliding variable and its rate of change to output sliding mode gains, achieving gain adaptation, uses SOGI to simultaneously perform filtering and integration functions, mitigating integral saturation caused by DC bias in collected data. Simulation results demonstrate that this method stably tracks the mover’s unique information under varying operating conditions of the Stirling linear generator, with both amplitude and phase tracking errors less than $5 \%$, meeting control system requirements.