摘要 |
PURPOSE:To enable the control of learning, regardless of whtehter a feedback compensation coefficient is skipped or not, by increasing a learnt value when the actual ratio of fuel to air is lower than a set value or when the feedback compensation coefficient is the upper limit value, and by decreasing the learnt value when the actual ratio is higher than the set value or when the compensation coefficient is the lower limit value. CONSTITUTION:When the feedback control of the air-fuel ratio of an engine is under way, a set time of T seconds has elapsed since the last calculation of a learnt value, a throttle valve 16 is fully open and the actual air-fuel ratio is judged to be higher than a set value, a learnt value correspoinding to the case that an air-fuel ratio feedback compensation coefficient FAF is the lower limit value of 0.8 is divided by K to use the quotient as an actual learnt value. When the throttle valve 16 is entirely closed, the actual air-fuel ratio is lower than the set value and the compensation coefficient FAF is the upper limit value of 1.2, the learnt value corresponding to the present time is multiplied by K to use the product as an actual learnt value. The control of learning is thus always enabled even if the compensation coefficient is not skipped, to improve the fuel efficiency, operation performance and exhaust gas purification performance of the engine. |