Fuel Injection Control for Engine Learning Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fuel injection amount control systems for internal combustion engines face challenges in maintaining both drivability and injection accuracy due to variations in engine rotation speed and accessory loads, leading to difficulties in accurately learning and correcting injection performance, especially in scenarios where the lockup mechanism is constantly engaged, which can result in incomplete learning processes and decreased reliability.
Innovation Solution
A fuel injection amount control apparatus that includes rotation speed detection, determination of learning conditions, learning-purpose injection, performance value calculation, and correction of injection commands, along with a compulsive signal to force specific load connection states, ensuring the learning process is completed before injection accuracy declines, thereby maintaining both drivability and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the learning process is executed under normal operating conditions with varying accessory loads, then the system can maintain good drivability, but the learning accuracy deteriorates due to fluctuations in engine rotation speed and load states
Solution Approach 1:
The patent segments the learning process into multiple incremental steps rather than attempting to complete the full learning range in one continuous process. The learning is divided into stages where the accessory load is adjusted in discrete increments, allowing the system to capture injection characteristics at different load points while maintaining stable operating conditions during each measurement phase.
Solution Approach 2:
The system performs preliminary stabilization of engine rotation speed before executing each learning measurement. The control apparatus adjusts the injection amount command value to maintain constant rotation speed during the learning process, ensuring that measurements are taken under stable conditions that improve accuracy while still reflecting actual operating characteristics.
2Measurement precision
If the accessory load is stopped to execute the learning process, then the learning accuracy improves, but the drivability deteriorates due to interruption of accessory operation
Solution Approach 1:
The patent implements periodic learning execution during naturally occurring stable operating intervals. The learning process is scheduled to execute during brief periods when operating conditions are favorable, such as during idle-up control phases or transient states, rather than requiring continuous interruption of accessory loads. This periodic approach captures necessary data while minimizing impact on overall system operation.
Solution Approach 2:
The system utilizes naturally occurring stable operating conditions during normal vehicle operation to perform learning measurements. By identifying and exploiting brief intervals where engine parameters are inherently stable (such as during idle transitions or load changes), the system performs self-learning without requiring external intervention or artificial creation of test conditions, thereby maintaining drivability while improving accuracy.
3Ease of operation
If the learning process is delayed until injection accuracy significantly degrades, then drivability is maintained longer, but the time required to complete learning increases beyond permissible limits
Solution Approach 1:
The system performs preliminary learning measurements during brief stable intervals throughout operation, accumulating data incrementally over time. By capturing injection characteristics during each stable period rather than waiting for a dedicated learning window, the system progressively builds the learning information base, ensuring completion within permissible time limits while maintaining normal drivability during intervening periods.
Data Source
AI summary
A fuel injection amount control apparatus for an internal combustion engine includes an ECU that commands a learning-purpose injection when a first learning condition regarding operation state and a second learning condition regarding load connection state are satisfied, and calculates an injection performance value that corresponds to the actual injection amount based on the amount of change in rotation speed, and further determines whether a delay of the learning process is permitted based on whether the learning process, despite occurrence of the delay, can be completed before the injector performance reaches a permissible limit value, and forces, when the delay is not permitted, the load connection state to be a specific connection state so as to satisfy the second learning condition. When it is determined that the delay of the learning process is permitted, the delay is permitted until the two learning conditions are satisfied.