| 摘要 |
PROBLEM TO BE SOLVED: To implement preferable learning corresponding to individual difference of a pump for delivery amount characteristics of a fuel pump.SOLUTION: An ECU 30, when calculating a first time learning value as a first time learning of delivery amount characteristics in a new pump state after manufacturing of a high pressure pump 13, guards the first time learning value by a first guard range for the first time learning set in advance. Furthermore, the ECU 30 sets a second guard range for re-learning based on the first time learning value after the calculation of the first time learning value. Moreover, the ECU 30, when calculating a re-learning value as the re-learning after performing the first learning, guards the re-learning value by the second guard range. |
