Piston-trajectory tracking for combustion stabilisation and generation consistency in a free-piston engine generator

The free-piston engine generator (FPEG) converts fuel chemical energy directly into electrical power via combustion-driven linear piston motion. However, without a crankshaft, the piston trajectory is not kinematically constrained, leaving the compression ratio—and consequently cold-start capability and combustion stability—unregulated. This study addresses the problem in two stages. First, closed-loop trajectory-tracking experiments on a dual-cylinder prototype demonstrate that stabilising the piston trajectory improves downstream generation: proportional–integral–derivative (PID) tracking increases indicated work by 7.2% and output electrical power by 5.0%, while reducing their cycle-to-cycle variability from 10.08% to 6.45% and from 3.16% to 2.10%, respectively. The underlying mechanism is thermodynamic: controlled motion extends the top dead centre (TDC) residence time from 1.4 to 1.7 ms, allowing combustion pressure to build near minimum volume and enlarging the high-pressure portion of the pressure–volume cycle. Because PID's fixed gains limit adaptability across a wide operating envelope, a simulation platform—validated against the experimental data—is developed to compare adaptive sliding-mode control (ASMC) with fixed-parameter sliding-mode control (SMC). Across 20 simulated operating conditions spanning stroke coefficients δ = 0.9–1.05 and shape factors Ω = 0.7–0.9, ASMC consistently outperforms SMC without per-condition re-tuning, improving TDC accuracy by 0.01–0.22 mm and indicated work by 0.24–1.13 J/cycle. ASMC also recovers fastest from combustion misfire and friction mutation, reducing operating-frequency variability by 21.21% and 27.83%, respectively, relative to SMC. These quantified margins provide criteria for controller selection and a basis for FPEG control-system design, pending hardware validation of the adaptive strategies.