Predictive model control design for a small-scale steam engine based multi-source CHP system

Sammanfattning

To improve the efficiency in decentralized energy systems, this thesis studiesmodel predictive control (MPC) strategies for a supercritical rankine cyclesystem, with water as the working fluid. This system allows various heatsources to achieve waste heat recovery, ensuring power supply in off-gridscenarios and emergencies. A physical model of this SCR system is built fromscratch in MATLAB comprising the boiler, steam engine, mass flow controlmechanism, etc., simulating the key thermodynamic processes of industrialconcerns.PID control and MPC are implemented and comparatively evaluated withperformance metrics of reference tracking ability and control efforts underdifferent test-rigs through the system simulation. Results indicate that MPCoutperforms PID in terms of overshoot reduction, convergence speed, andcontrol effort minimization, due to its advantages in managing multipleconstraints and control output limiting.