DCT Clutch Torque Control Using Observer-Based Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Dual clutch transmissions (DCTs) with dry clutches face challenges in accurately controlling torque-stroke curve characteristics during shift stage changes, leading to potential shift shocks and engine flare due to incomplete learning of these characteristics.
Innovation Solution
A clutch torque control method that determines whether a power-on downshift is initiated and corrects basic control torque using observer torque calculated by a torque observer, adding or subtracting a correction value based on a function relation between the observer torque and basic control torque to stabilize and improve shift control within a real shift range.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If the latest TS curve characteristics are used to control the dry clutch during shift stage changes, then the control response is fast, but the accuracy of torque control deteriorates due to incomplete learning
Solution Approach 1:
The system performs preliminary learning of TS curve characteristics during periods when the clutch is not actively engaged in shifting. The learned characteristics are stored and then applied during subsequent shift operations, allowing the system to prepare accurate control data in advance without compromising real-time control speed.
Solution Approach 2:
The control system dynamically selects between using pre-learned TS curve characteristics and real-time sensor feedback based on the current operating conditions. During active shifting, the system uses the stored learned characteristics for stable torque control, while continuously updating the learning database for future improvements.
2Measurement precision
If the TS curve characteristics are learned frequently, then the accuracy of torque control improves, but the system complexity and computational load increase
Solution Approach 1:
Instead of performing complete TS curve learning at every possible opportunity, the system performs partial learning updates only when specific conditions are met (e.g., during idle periods, when sensor data quality is high, or when significant deviations are detected). This reduces computational burden while maintaining sufficient accuracy.
Solution Approach 2:
The control system automatically manages its own learning process by monitoring its own performance and initiating learning updates only when necessary. The system self-regulates the learning frequency and intensity based on detected performance degradation or operational conditions, eliminating the need for external intervention or complex scheduling.
3Stability of the object's composition
If the actuator is controlled based on stored TS curve characteristics, then the shift stage change is stable, but engine flare or shift shock occurs when the characteristics are outdated
Solution Approach 1:
The system continuously monitors actual shift performance and engine response during clutch operations. When deviations from expected behavior are detected (indicating outdated TS characteristics), the system triggers corrective learning updates. This feedback loop ensures that the stored characteristics remain accurate without requiring constant relearning.
Solution Approach 2:
The system applies compensatory adjustments to the control torque based on predicted errors from outdated TS characteristics. Before the actual shift occurs, the system calculates correction factors based on the age and reliability of the stored characteristics, cushioning against potential engine flare or shift shock.
Data Source
AI summary
A clutch torque control method for a dual clutch transmission (DCT) vehicle may include a shift initiation determining step of determining whether power-on downshift in which a driver steps on an accelerator pedal to change a current shift stage to a lower shift stage is initiated, and a torque correcting step of correcting basic control torque according to torque-stroke (TS) curve characteristics for controlling a disengagement-side clutch within a real shift range in which a number of rotations of an engine is changed with observer torque calculated by a torque observer when the power-on downshift is initiated, and determining the corrected basic control torque into control torque of the disengagement-side clutch.