Engine Mass Air Flow Estimation Using Regression and Flow Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Internal combustion engines face challenges in accurately controlling the air-to-fuel ratio (AFR) during transient operations due to time lags in sensor readings and the complexity of multiple actuated components, which can lead to inefficiencies and emission issues, and existing solutions are susceptible to sensor failures.
Innovation Solution
A method that uses a combination of regression and flow models to estimate engine mass air flow (MAF) by monitoring fluid pressures and valve positions, allowing for redundant sensor inputs and adaptive control strategies to maintain accurate AFR control even in the presence of sensor failures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If MAF is determined using dedicated MAF sensors or indirect calculation based on intake-manifold pressure sensors, then MAF measurement is available, but time lags occur during transient events affecting AFR control accuracy
Solution Approach 1:
The system performs preliminary estimation of MAF using a flow model that incorporates valve positions and fluid pressures before actual fuel injection occurs. This preliminary action allows the control system to have an advance estimate of MAF during transient events, eliminating the time lag problem associated with waiting for sensor readings to update.
Solution Approach 2:
The invention introduces an intermediary flow model that mediates between the physical sensor readings and the control decisions. This flow model uses valve positions and fluid pressures as intermediate parameters to calculate MAF, providing a bridge that delivers timely estimates without waiting for direct sensor updates during transient conditions.
2Adaptability or versatility
If multiple actuated components (throttle valve, EGR valve, compressor bypass valve, turbine waste-gate valve) are used to control engine parameters, then engine performance and emissions are optimized, but the complexity of estimating MAF increases due to multiple changing valve positions
Solution Approach 1:
The system segments the MAF estimation problem by treating each valve (throttle, EGR, compressor bypass, turbine waste-gate) as an independent component with its own position parameter. The flow model calculates the contribution of each valve separately and combines them, making the complex multi-valve system manageable through modular estimation of individual valve effects on air flow.
Solution Approach 2:
The invention changes the parameters used for MAF estimation from direct sensor readings to a combination of valve positions and fluid pressures. By using valve positions (which are directly controllable and measurable) along with pressure measurements, the system transforms a complex direct measurement problem into a calculable parameter-based estimation that accounts for multiple actuated components.
3Productivity
If traditional AFR control systems are used, then normal operation is maintained, but sensor failures cause the engine to enter reduced power mode resulting in power loss
Solution Approach 1:
The system prepares for potential sensor failures by having a pre-developed flow model that can estimate MAF using alternative parameters (valve positions and fluid pressures). This beforehand cushioning ensures that when sensor failures occur, the system already has a backup estimation method ready, preventing the need to enter reduced power mode and maintaining engine productivity.
Solution Approach 2:
The invention creates a virtual copy of the MAF measurement function through the flow model. Instead of relying solely on physical MAF sensors, the system implements a software-based copy that replicates the MAF measurement capability using valve positions and pressure data. This virtual copy provides redundancy, ensuring continuous operation even when physical sensors fail.
Data Source
AI summary
A system and method for determining an engine mass air flow (MAF) for use in an engine air to fuel ratio (AFR) calculation to operate an engine includes monitoring engine operation, determining in the electronic controller a first estimation of engine MAF based on a regression model, determining in the electronic controller a second estimation of engine MAF based on a flow model, and selecting the first or second estimation of engine MAF based on an operating state of the engine. Each estimation can use various engine parameters interchangeably to provide a robust system against sensor failures.