| 摘要 |
PURPOSE:To enable control of an air-fuel ratio with further high precision at a stage in that learning is progressed, by a method wherein, when a difference between a learning correction factor before correction and a noble learning correction factor is below a given value, a feedback correction factor is clamped to a reference value. CONSTITUTION:A learning correction factor correcting means H learns a deviation from the reference value of a feedback correction factor, a learning correction factor responding to a region under the running state of an engine is corrected in a direction in which the deviation is decreased to rewrite a data in a memory means C. In which case, a feedback correction factor clamp means I compares a learning correction factor before correction with a novel learning correction factor, and when a difference therebetween is below a given value, a feedback correction factor, set by a feedback correction factor setting means E, is clamped to a reference value to send it to a fuel injection amount computing means F. Namely, feedback control is stopped, and is brought into an open loop, and control precision of an air-fuel ratio is improved as a learning result is effectively utilized.
|
