Cold Start Fuel Richness Correction Using Learned Injection Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for fuel richness correction in heat engines during cold starts are inadequate as they rely on non-operational richness probes, leading to incomplete and insufficient adaptive physical models, which cannot accurately reset fuel richness measurements, affecting the reliability and robustness of cold start corrections.
Innovation Solution
A method that corrects fuel injection quantities based on learned values from previous cold starts, using a classification system of physical parameters to determine the most suitable injection quantity, while inhibiting corrections if the learned data is obsolete, by comparing the engine's aging state over time and ensuring the chosen quantity meets predetermined conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If adaptive physical models are used to compensate for component dispersions and aging, then measurement accuracy is improved, but the models provide false values during cold start when the richness sensor is not operational
Solution Approach 1:
The patent applies preliminary action by pre-calculating and storing corrected fuel injection quantities for various cold start conditions before the actual cold start occurs. The system uses learned data from previous cold starts and stores corrected injected fuel quantities classified by physical parameters (temperature, humidity, altitude) so that during a new cold start, the appropriate pre-computed correction can be immediately applied without waiting for the richness sensor to become operational.
Solution Approach 2:
The patent introduces an intermediary approach by using learned correction factors and classified fuel quantity data as intermediaries between the non-operational richness sensor and the fuel injection system. During cold start, instead of directly relying on the non-operational sensor, the system uses pre-learned correction data that mediates the fuel richness control until the sensor becomes operational.
2Reliability
If corrected fuel injection quantities are learned from previous cold starts, then fuel richness correction is improved, but obsolete learned data may be applied leading to increased pollutant emissions
Solution Approach 1:
The patent implements feedback by continuously monitoring engine operating conditions and comparing them against stored learned data. The system uses feedback mechanisms to determine whether learned correction data is still valid or has become obsolete due to changes in engine conditions, aging, or environmental factors. This feedback loop prevents the application of obsolete corrections that would increase pollutant emissions.
Solution Approach 2:
The patent applies parameter changes by monitoring variations in physical parameters (temperature, humidity, altitude, engine age) and using these changes to determine the validity of learned correction data. When parameters change beyond certain thresholds, the system recognizes that learned data may be obsolete and adjusts its behavior accordingly, preventing the application of outdated corrections.
3Ease of operation
If the richness sensor is used for fuel richness regulation, then fuel control is simplified, but the sensor is not operational during cold start when it has not reached its operating temperature
Solution Approach 1:
The patent applies preliminary action by pre-computing and storing corrected fuel injection quantities for cold start conditions before the richness sensor becomes operational. The system uses learned data from previous cold starts to establish baseline corrections that can be applied immediately during cold start, eliminating the need to wait for the sensor to reach operating temperature.
Solution Approach 2:
The patent uses copying by creating virtual copies of operational data from previous cold starts. Instead of waiting for the actual richness sensor to provide measurements during cold start, the system copies and reuses corrected fuel quantity data from previous cold start events, effectively replicating the functionality of the non-operational sensor.
Data Source
Figure 1
Figure 2
AI summary
Method for correcting the richness of fuel in a heat engine (1) during a cold start, with a richness sensor (6) in the exhaust line that is not yet operational, characterised in that it comprises correcting a quantity of fuel injected into the engine (1) during a cold start in progress according to corrected quantities of injected fuel which have been learned from previous cold starts, stored and classified according to first physical parameters, the selected corrected quantity of injected fuel being the quantity with a classification according to the first physical parameters which most closely approximates the cold start in progress, the method verifying whether predetermined conditions (120) are satisfied so as to authorise the correction, and the predetermined conditions comprising inhibition conditions (122) prohibiting the correction when the learning of the selected adjusted quantity of injected fuel is judged to be obsolete with respect to the cold start in progress.