Robust Multivariable Feedback Control of Natural Gas-Diesel RCCI Combustion

Advanced combustion concepts such as Reactivity Controlled Compression Ignition (RCCI) demonstrate very high thermal efficiencies combined with ultra low NOx emissions. As RCCI is sensitive for operating conditions, closed-loop control is a crucial enabler for stable and robust combustion. The feedback design is complex due to the coupling between inputs, such as intake manifold temperature, blend ratio (BR), and air-to-fuel ratio (λ), and, on the other hand, combustion parameters. Contrary to earlier, parallel SISO approaches, a multivariable feedback control strategy is applied, which uses a new combination of control parameters. Following the linearisation of a validated multi-zone RCCI combustion model, a MIMO feedback control strategy is designed. This cycle-to-cycle control strategy uses diesel injection timing, diesel fuel quantity, and natural gas (NG) fuel quantity information to track the desired engine load, ignition delay, and a new control parameter: BR. The designed controller is validated with simulation cases for disturbance rejection and simultaneous multivariable reference tracking. Simulation results demonstrate robust performance, effective decoupling action, and fast settling time (within 3 engine cycles) of the designed controller.