Cylinder Richness Estimation Using Kalman Filter Dynamics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating cylinder richness in engines, particularly diesel engines, face challenges in dynamic operation due to delays and filtering effects from probe placement and operating principles, leading to non-representative measurements and insufficient robustness against component dispersions and drifts.
Innovation Solution
A method that calculates a simple richness model using fuel flow and air flow measurements, applies a Kalman filter for realignment and correction at transient operating points, and monitors probe saturation to adjust the model, incorporating a mapping of wealth drifts to account for engine dynamics and component variations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a probe is used to measure richness, then precise values are obtained in stable conditions, but delays and filtering occur that make measurements non-representative during dynamic operation
Solution Approach 1:
The patent applies dynamics by making the measurement system adaptive to changing operating conditions. The Kalman filter dynamically adjusts its estimation based on current sensor readings and model predictions, allowing the system to track rapid transients in richness while compensating for probe delays and filtering effects through predictive algorithms.
Solution Approach 2:
The patent implements feedback through the Kalman filter architecture, which continuously compares predicted richness values with actual sensor measurements and uses the difference (innovation) to correct future predictions. This closed-loop feedback mechanism compensates for measurement delays and maintains accuracy during dynamic operation by constantly updating the estimation based on new information.
2Speed
If richness is estimated using intake air flow and injected fuel flow parameters, then dynamic response is improved, but dispersions and drifts of engine components reduce accuracy
Solution Approach 1:
The patent uses feedback through the Kalman filter to continuously correct the dynamic richness estimation. The filter compares the calculated richness (from air and fuel flow) with actual probe measurements and uses the discrepancy to identify and compensate for component dispersions and drifts, maintaining accuracy while preserving dynamic response.
Solution Approach 2:
The patent applies parameter changes by using the Kalman filter to dynamically adjust the richness estimation parameters based on identified component variations. The system adapts its model parameters to account for injector drifts, pump variations, and flow meter dispersions, maintaining precision across different operating conditions while preserving fast dynamic response.
3Reliability
If richness is mapped according to speed and torque parameters, then robustness to component variations is improved, but correct dynamics representative of actual cylinder richness are lost
Solution Approach 1:
The patent applies dynamics by replacing static mapping with a dynamic Kalman filter estimation that continuously adapts to changing operating conditions. The filter processes real-time sensor data and model predictions to track actual richness dynamics in each cylinder, preserving the correct temporal behavior while the feedback mechanism provides robustness to component variations.
Solution Approach 2:
The patent applies segmentation by analyzing richness dynamics independently for each cylinder using the Kalman filter. This allows the system to capture individual cylinder variations and dynamics that would be averaged out in a global mapping approach, while the feedback mechanism provides robustness to component dispersions through continuous correction.
4Measurement precision
If a Kalman filter is applied at transient operating points, then continuity and precision are improved, but probe saturation causes incorrect drift identification
Solution Approach 1:
The patent uses feedback through a monitoring mechanism that detects probe saturation conditions and provides feedback to the Kalman filter algorithm. When saturation is detected, the system adjusts its behavior to avoid incorrect drift identification, ensuring reliability is maintained while preserving precision during normal operation through continuous feedback correction.
Solution Approach 2:
The patent applies preliminary anti-action by implementing a saturation monitoring mechanism that prevents incorrect drift identification before it occurs. The system proactively detects saturation conditions and takes corrective action (such as suspending drift updates or adjusting filter parameters) to prevent the propagation of erroneous information, maintaining both precision and reliability.
Data Source
Figure 1~2b
Figure 3
AI summary
The method involves calculating a simple model of richness (Phi-Calc) using flow value of fuel injected (qinj) in a cylinder, measuring intake air flow (Qair) and storing quantized wealth drifts operating points. Correction is applied to a simple model of richness from a drift wealth stored in the mapping. A quantified drift is stored in operating points when the engine is in transition state. A value of drift memorized in the mapping is retimed, where the drift value memorized in the mapping is carried out by application of a Kalman filter in a point of operation in the transitory mode.