| Demands for high output performance, stringent fuel consumption and emission regulations have become some of the major challenges for engineers in the automotive field, especially control engineers. Precise control of air-fuel ratio during both steady and transient engine operation is the key for better engine performance and reduced exhaust emissions.; In this thesis, a nonlinear, fuel injected spark-ignition engine numerical model is developed. This physically motivated model includes intake manifold air dynamics, fuel wall-wetting dynamics, load speed effects, crank shaft dynamics and process delays inherent in a four-stroke engines. The developed model is unique in that it incorporates all of the above noted phenomenon. A sliding mode controller is then designed based on a linearized version of the mathematical engine model. The linear model is compared favorably with a non linear model in open loop simulated throttling tests. The speed of the system response is limited due to the delay time from the air-fuel ratio (AFR) metering location to the sensor using a conventional feedback control design. Modern control and estimation theory was used to design a high bandwidth closed-loop AFR controller. The sliding mode controller is robust to rapidly changing throttle maneuvers and disturbances. In order to implement the inherent state feedback control scheme, an optimal state estimator is developed.; The sliding mode control method offers a strong potential for future engine control problems, since: it results in a relatively simple control structure; it is robust to model errors; and it can be easily adapted to a wide family of engines. |
