Torque Converter Quadratic Modeling for Maximum Powertrain Torque
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for quantifying the behavior of torque converters in electrified powertrains are either computationally expensive or lead to inaccurate performance due to reliance on complex models or empirical data, resulting in suboptimal powertrain performance.
Innovation Solution
A powertrain control system that utilizes empirical operation data to determine a quadratic polynomial representation of impeller speed for the torque converter across an intersection region, integrating torque converter characteristics with engine and motor torque capabilities, and optimizing the search for the intersection window from the highest speed breakpoint.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a complex torque converter model (e.g., neural network model) is generated and utilized by a controller, then measurement precision of torque converter behavior is improved, but device complexity and computational cost increase
Solution Approach 1:
The patent replaces complex, computationally expensive neural network models with simple quadratic polynomial equations that can be easily calculated by the controller. These simplified mathematical models achieve sufficient accuracy for torque converter behavior prediction without requiring complex computational resources, effectively using 'cheap' simple models instead of 'expensive' complex ones.
Solution Approach 2:
The patent transforms the torque converter model from a complex neural network structure into simple quadratic polynomial parameters (coefficients a, b, c). By changing the mathematical representation from complex nonlinear functions to simple quadratic equations with adjustable coefficients, the system achieves both computational efficiency and adequate measurement precision.
2Device complexity
If predetermined empirical data is used to estimate torque converter behavior, then device complexity is reduced, but measurement precision deteriorates resulting in decreased powertrain performance
Solution Approach 1:
The patent incorporates feedback mechanisms where the quadratic model coefficients are continuously adjusted based on actual torque converter performance data. The system compares predicted torque converter behavior with actual measurements and refines the quadratic equation parameters, ensuring the simplified model maintains high accuracy through continuous learning and adaptation.
Solution Approach 2:
The patent makes the torque converter model dynamic by allowing the quadratic polynomial coefficients to change based on operating conditions. Rather than using fixed empirical data, the system adapts the model parameters in real-time to match current torque converter behavior, maintaining precision across varying operational states while keeping the computational structure simple.
3Device complexity
If linear behavior assumption is used between breakpoints, then device complexity is reduced, but measurement precision of maximum torque capability deteriorates
Solution Approach 1:
The patent replaces linear interpolation between breakpoints with quadratic polynomial equations that capture the curved, nonlinear relationship between torque and speed. By using quadratic curves instead of straight lines, the model accurately represents the actual torque converter behavior including peak torque regions, while maintaining computational simplicity through the closed-form quadratic solution.
Data Source
AI summary
A powertrain control method for a vehicle includes determining, based on empirical operation data for a powertrain, maximum torques for the torque generating system at each of a plurality of breakpoints corresponding to different impeller speeds for a torque converter and different speeds of a torque generating system, identifying, between two particular breakpoints, (i) a linear intersection point between the maximum torque and the impeller speed for the torque converter and (ii) an intersection region between the two particular breakpoints, determining a quadratic polynomial representation of the impeller speed for the torque converter across the intersection region based on the empirical operation data for the torque converter, and utilizing the quadratic polynomial representation of the impeller speed for the torque converter across the intersection region for improved control of the powertrain.