Valve Flowrate Model Correction Using Richness Sensor Deviation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the flowrate model of a valve, particularly in the context of a canister purge valve for internal combustion engines, face challenges due to manufacturing variations, which affect the accuracy of canister load estimation and fuel supply management, requiring a precise modeling of the flowrate as a function of the applied command.
Innovation Solution
A method that iteratively corrects a flowrate model by using deviation information from a pre-existing reduced area controller, eliminating the need for additional sensors, and allowing for in-situ determination and periodic correction of the model to account for variations and environmental effects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manufacturing variations are present in valve production, then valve flowrate characteristics deviate from nominal models, but using additional sensors to compensate increases device complexity and cost
Solution Approach 1:
The system uses existing sensors (reduced area controller, richness sensor) to automatically detect and correct valve model deviations without requiring additional measurement devices. The engine's normal operation provides the test conditions needed for self-diagnosis and model correction.
Solution Approach 2:
The system implements a feedback loop where the richness sensor deviation is used to calculate a correction factor that adjusts the valve flowrate model. This closed-loop approach continuously refines the model accuracy using operational data from the engine system.
2Measurement precision
If a precise flowrate model is used for canister load estimation, then fuel supply management accuracy improves, but model determination requires additional testing and calibration time
Solution Approach 1:
The system performs preliminary model correction during engine startup or idle periods when the engine is already running. The correction process uses normal engine operation conditions to determine valve characteristics, avoiding the need for separate testing procedures.
Solution Approach 2:
The model determination and correction process occurs continuously during normal engine operation rather than requiring a separate calibration phase. The system uses ongoing engine data to continuously refine the valve model, turning operational time into calibration time.
3Reliability
If the valve model is corrected iteratively using engine operation data, then model accuracy improves over time, but the control algorithm complexity increases
Solution Approach 1:
The system applies a simplified correction factor based on richness sensor deviation rather than implementing a full iterative optimization algorithm. This partial correction approach provides sufficient accuracy improvement without the computational burden of complex control algorithms.
Data Source
AI summary
Disclosed are a method and device for determining a model of flowrate for a valve, including: instantiation of the model initially equal to an approximate model; choosing flowrate setting; determination of an initial command corresponding to the flowrate setting by the model; application of the initial command to the valve; observation of a discrepancy indicative of a difference between the flowrate setting and a flowrate achieved; if the discrepancy is not zero, application of a correction to the initial command, and return to the application step; if the discrepancy is zero, correction of the model by replacing, for the flowrate setting, the initial command by the corrected command, where the flowrate passing through the valve ends up in a combustion engine admission collector including a small-area controller regulating the flowrate of air admitted, and where the discrepancy of the observation step is the deviation of the small-area controller.