Controlling spark timing for consecutive cycles to reduce the cyclic variations of SI engines

Abstract

Minimization of the cyclic variations is one of the most important design goal for spark-ignited engines. Primary motivation of this study is to reduce the cyclic variations in spark ignition engines by controlling the spark timing for consecutive cycles. A stochastic model was performed between spark timing and in–cylinder maximum pressure by using the system identification techniques. The incylinder maximum pressure of the next cycle was predicted with this model. Minimum variance and generalized minimum variance controllers were designed to regulate the in–cylinder maximum pressure by changing the spark timing for consecutive cycles of the test engine. The produced control algorithms were built in LabView environment and installed to the Field Programmable Gate Arrays (FPGA) chassis. According to the test results, the in–cylinder maximum pressure of the next pressure cycle can be predicted fairly well, and the spark timing can be regulated to keep the in–cylinder maximum pressure in a desired band to reduce the cyclic variations. At fixed spark timing experiments, the COVPmax and COVimep were 3.764 and 0.677%, whereas they decreased to 3.208 and 0.533% when GMV controller was applied, respectively.