Automatic Transmission Gearshift Calibration With Model-Based Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for calibrating and adapting gearshift controllers in automatic transmissions require extensive effort and are inefficient, especially for transmissions with higher speeds, as they rely heavily on design-of-experiments approaches that cannot adapt to system changes over time due to wear and use, leading to increased calibration needs and complex rule-based adaptations.
Innovation Solution
A model-based learning method that automates the calibration and adaptation of gearshift controllers by performing a sequence of gearshifts, acquiring data from speed sensors, averaging it, and updating control parameters using learning controllers, reducing the number of gearshifts required for calibration and enabling adaptation during vehicle operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If DoE-based calibration methods are used for transmissions with more speeds (8, 9, 10 speeds), then the calibration can be performed systematically, but the number of required gearshifts and calibration labels increases dramatically (e.g., 22,000 labels for 10-speed vs. 800 for 4-speed)
Solution Approach 1:
The patent segments the gearshift control into multiple distinct phases (fill phase, torque phase, inertia phase), each with its own control parameters. This segmentation allows the calibration process to focus on specific phases independently, reducing the total number of calibration labels needed while maintaining comprehensive coverage of all gearshift conditions.
Solution Approach 2:
The patent performs preliminary system identification to develop a dynamic model of the transmission system before calibration. This preliminary modeling step captures the system's behavior characteristics, allowing the subsequent calibration process to be more efficient and require fewer test points, thereby reducing calibration time and resource requirements.
2Reliability
If DoE-based calibration methods are used, then initial calibration can be performed, but the method cannot adapt to system changes over time due to wear and use, requiring additional calibration effort for adaptive routines
Solution Approach 1:
The patent implements a feedback mechanism where the dynamic model is continuously updated based on actual system performance data collected during operation. This feedback loop enables the system to adapt to wear and changes over time automatically, eliminating the need for separate adaptive calibration routines while maintaining high reliability.
Solution Approach 2:
The patent transitions from static DoE-based calibration to a dynamic model-based approach where the transmission model adapts in real-time to system changes. This dynamic capability allows the system to automatically adjust to wear and operational variations, providing both reliability and adaptability without requiring additional calibration efforts.
3Adaptability or versatility
If rule-based adaptive policies are used for adaptation, then some adaptation capability is provided, but the rules become complex and the effectiveness is limited
Solution Approach 1:
The patent replaces complex rule-based adaptive policies with a physics-based dynamic model that naturally captures transmission behavior. This model-based approach substitutes the need for numerous complex control rules with a unified mathematical representation, reducing control complexity while maintaining or improving adaptation effectiveness.
Data Source
AI summary
Methods for automated calibration and adaption of a gearshift controller (39) are disclosed. In one aspect, the method automates calibration a gearshift controller (39) for controlling a sequence of gearshifts in either a stepped automatic transmission equipped with at least one speed sensor mounted on a dynamometer (42) or an automotive vehicle mounted on a dynamometer (42), where the dynamometer (42) is electronically controlled by a dynamometer controller (43). Each gearshift in the sequence includes a first phase, a second phase, . . . and an Nth phase. The gearshift controller (39) includes (initial values of) a first phase control parameters set, a second phase control parameters set, . . . and an Nth phase control parameters set for each gearshift in the sequence that are updated using a first phase learning controller, a second phase learning controller, . . . and an N*11 phase learning controller respectively.