The introduction of inexpensive cylinder pressure sensors provides new oppo
rtunities for precise engine control. This paper presents a spark advance c
ontrol Strategy based upon cylinder pressure in spark ignition engines. It
is well known that the location of peak pressure (LPP) reflects combustion
phasing and can be used for controlling the spark advance. The well-known p
roblems of the LPP-based spark advance control method are that many samples
of data are required and there is loss of combustion phasing detection cap
ability due to hook-back at late burn conditions. To solve these problems,
a multi-layer feedforward neural network is employed. The LPP and hook-back
are estimated, using the neural network, which needs only five output volt
age samples from the pressure sensor. The neural network plays an important
role in mitigating the A/D conversion load of an electronic engine control
ler by increasing the sampling interval from 1 degrees crank angle(CA) to 2
0 degrees CA. A proposed control algorithm does not need a sensor calibrati
on and pegging (bias calculation) procedure because the neural network esti
mates the LPP from the raw sensor output voltage, The estimated LPP can be
regarded as a good index for combustion phasing, and can also be used as an
MBT control parameter. The feasibility of this methodology is closely exam
ined through steady and transient engine operations to control individual c
ylinder spark advances. The experimental results have revealed a favorable
agreement of optimal combustion phasing in each cylinder.