Fuel Injector Transfer Function Learning for Consistency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fuel injectors in internal combustion engines exhibit variability due to manufacturing tolerances and aging, leading to inconsistent fuel injection amounts, resulting in fueling discrepancies across similar injectors.
Innovation Solution
A method is developed to determine and adjust the injector transfer function shape by injecting reference injections followed by pulses at varying widths, allowing for local learning or retrieval of fuel mass transfer function shapes from a cloud, enabling consistent fuel injection across similar injectors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional fuel injection methods are used, then the system is simple to operate, but fuel injection consistency deteriorates due to manufacturing tolerances and aging
Solution Approach 1:
The system uses feedback from fuel rail pressure sensors to monitor actual fuel injection amounts and compares them against target values. The controller adjusts injection parameters based on this feedback to compensate for variations caused by manufacturing tolerances and aging, thereby improving fuel injection consistency without requiring complex hardware modifications.
Solution Approach 2:
The system dynamically changes injection parameters such as pulse width, injection timing, and pressure profiles based on learned transfer function shapes specific to each injector. By adapting these parameters through continuous learning and adjustment, the system compensates for injector variations and maintains consistent fuel delivery despite manufacturing tolerances and aging effects.
2Measurement precision
If individual injector calibration is performed, then fuel injection precision improves, but learning time increases
Solution Approach 1:
The system performs preliminary learning of transfer function shapes during vehicle operation by injecting reference amounts of fuel at multiple pulse widths and measuring the resulting fuel rail pressure drops. This preliminary action captures injector characteristics in advance, allowing the controller to pre-calculate compensation factors that will be applied during normal operation, thereby improving measurement precision without significantly increasing overall learning time.
Solution Approach 2:
The system employs periodic reference injections at different pulse widths to continuously update and refine the transfer function shapes. By conducting these learning cycles periodically during vehicle operation rather than requiring a single lengthy calibration process, the system achieves high measurement precision while distributing the learning time over multiple short intervals, reducing the perceived learning time impact.
3Productivity
If cloud-based data retrieval is used, then learning efficiency improves, but system complexity increases
Solution Approach 1:
The system uses a cloud-based intermediary service that stores and provides transfer function shape data from multiple vehicles. Instead of each vehicle performing complete independent learning, the system queries the cloud intermediary for pre-learned data that can be directly applied or lightly adapted, thereby significantly improving learning speed and productivity while adding only moderate system complexity through cloud communication interfaces.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach reduces the time required to learn the injector transfer function, improves fuel injection consistency, and can be applied to both port and direct fuel injectors while the vehicle is operating, enhancing engine performance and reducing emissions.
Implementation Method 1
sensing a fuel rail pressure drop for at least one injector
Data Source
AI summary
Methods and systems are provided for balancing injector fueling. In one example, a method includes learning portions of a transfer function shape by firing a plurality of injectors at PWs of a set of PWs following a reference injection.