Convergence Filters for Efficient Economic MPC of Non-dissipative Systems
By: Defeng He , Weiliang Xiong , Shiqiang He and more
Potential Business Impact:
Makes machines run better and use less energy.
This note presents a novel, efficient economic model predictive control (EMPC) scheme for non-dissipative systems subject to state and input constraints. A new conception of convergence filters is defined to address the stability issue of EMPC for constrained non-dissipative systems. Three convergence filters are designed accordingly to be imposed into the receding horizon optimization problem of EMPC. To improve online computational efficiency, the variable horizon idea without terminal constraints is adopted to compromise the convergence speed, economic performance, and computational burden of EMPC. Moreover, sufficient conditions are derived to guarantee the recursive feasibility and stability of the EMPC. The advantages of the proposed EMPC are validated by a classical non-dissipative continuous stirred-tank reactor.
Similar Papers
Varying Horizon Learning Economic MPC With Unknown Costs of Disturbed Nonlinear Systems
Systems and Control
Makes machines run better and cheaper.
Multi-Timescale Model Predictive Control for Slow-Fast Systems
Systems and Control
Makes robots move faster by simplifying their thinking.
Adaptive Event-Triggered MPC for Linear Parameter-Varying Systems with State Delays, Actuator Saturation and Disturbances
Systems and Control
Makes machines smarter, using less power.