APPARATUS FOR LEARNING CONTROL OF AIR-FUEL RATIO OF AIRFUEL MIXTURE IN ELECTRONICALLY CONTROLLED FUEL INJECTION TYPE INTERNAL COMBUSTION ENGINE

The present invention relates to a learning control apparatus for controlling the air-fuel ratio in an electronic control fuel injection type internal combustion engine wherein the injection quantity (Ti=Tp.times.K.times..alpha.) is basically operated by multiplying the basic injection quantity (Tp), determined from the flow amount of sucked air and the rotation speed, by the air-fuel ratio feedback correction coefficient (.alpha.), and the feedback control of the air-fuel ratio to the aimed target air-flow ratio is performed on the thus operated injection quantity Ti. However, as it takes long time to control the air-fuel ratio to the target air-flow ratio by the feedback control of the integration if the basic air-fuel ratio at the time of .alpha.=1 deviates from the aimed at air-fuel ratio, the value .alpha. at the time in the stationary driving condition and the driving condition per se, are detected, the learning correction coefficient .alpha.o prestored in RAM corresponding to the engine condition is renewed to the new learning correction coefficient .alpha.onew by an operation based on the stored .alpha.o and the detected feedback correction coefficient .alpha., and the next injection quantity is operated by the equation Ti=Tp.times.K.times..alpha.o.times..alpha. thereby the feedback correction coefficient .alpha. could be small. In addition, at the time of acceleration or deceleration of the engine, the acceleration or deceleration correction coefficient Kacc, Kdcl included in constant K is corrected by learning corresponding to the engine driving state, Kacc and Kdcl are pre-stored in RAM and renewed the detected new Kacc and Kdcl, the renewed data is used for the subsequent operation of Ti.