Cyclic torque imbalance detection in gasoline engines using a uniform second-order sliding mode observer

Engine torque imbalance is a wide-ranging problem which is caused due to variance of the combustion mixture in the engine cylinder. In this research work, cyclic torque imbalance detection is carried out by formulating a uniform second-order sliding mode observer using the First Principle–based Engine Model. Oscillations in the crankshaft angular speed were modeled in the novel First Principle–based Engine Model, which were suppressed in the Mean Value Engine Models. Cyclic torque imbalance is simulated at multiple instances by varying the injected fuel mass. Estimation of the net piston force is carried out for cyclic torque imbalance detection using rotational dynamics of the engine model. This force is treated as unknown input to the torque production subsystem of the model. Cyclic torque imbalance detection is validated using the GT-Power engine model. Variations in the cyclic torque were detected proximate to actual values which demonstrated validity of the proposed technique.