Disturbance Observer for Electric Power Steering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing electric power steering (EPS) systems face challenges in accurately detecting and rejecting harmonic disturbances, such as steering wheel shake or nibble, due to mass imbalances or brake disc variations, which affects steering feel and requires complex calibration and computational resources, and existing methods like feedforward cancellation are not effective in all scenarios.
Innovation Solution
The system employs a disturbance rejection method that includes a baseline control torque signal and an auxiliary control torque signal generated by a disturbance rejection controller, which estimates and refines operating parameters to minimize errors and generate a harmonic correction torque, maintaining baseline performance and correcting system parameter changes through inherent feedback mechanisms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If feedforward cancellation is used to detect and counteract disturbances, then disturbance rejection capability is improved, but computational resources and memory requirements increase significantly
Solution Approach 1:
The patent replaces complex trigonometric calculations with a simplified observer-based estimation approach. Instead of using computationally intensive feedforward cancellation methods that require significant memory and processing power, the system uses a disturbance observer that estimates disturbances through algebraic operations and basic integration, dramatically reducing computational resource requirements while maintaining effective disturbance rejection
Solution Approach 2:
The patent extracts the disturbance estimation function from the complex feedforward cancellation framework and implements it as a separate disturbance observer module. This observer independently estimates disturbances and generates compensation torques, separating the disturbance rejection function from the main control loop and reducing overall computational burden
2Object-affected harmful factors
If high gain is used to increase disturbance rejection, then disturbance rejection is improved, but system stability deteriorates and sensitivity to parameter uncertainties increases
Solution Approach 1:
The patent implements a disturbance observer that continuously monitors system behavior and feeds back disturbance estimates to generate compensation torques. This feedback mechanism allows the system to achieve strong disturbance rejection with moderate gain by continuously adapting to actual system conditions, avoiding the instability issues associated with high open-loop gain feedforward cancellation
Solution Approach 2:
The disturbance observer performs preliminary estimation of disturbances before they significantly affect system performance. By proactively estimating and compensating for disturbances based on early system response indicators, the system achieves effective disturbance rejection without requiring high gain that would compromise stability
3Measurement precision
If filters are used to detect disturbances, then disturbance detection is improved, but steering feel deteriorates due to filtering out torque signals
Solution Approach 1:
The patent introduces a disturbance observer as an intermediary that indirectly estimates disturbances through system model-based calculations rather than directly filtering sensor signals. This observer uses motor current, voltage, and position data along with a system model to compute disturbance estimates, avoiding the need to filter out torque signals and preserving natural steering feel while maintaining accurate disturbance detection
Solution Approach 2:
The patent replaces traditional filter-based disturbance detection with model-based estimation. Instead of using filters that remove frequency components from sensor signals (affecting steering feel), the system uses a mathematical model to estimate disturbances from available sensor data, eliminating the trade-off between detection accuracy and steering feel
4Object-affected harmful factors
If feedforward cancellation is implemented, then disturbance rejection is improved, but system complexity increases due to non-linear system analysis requirements
Solution Approach 1:
The patent replaces non-linear feedforward cancellation algorithms with a linear disturbance observer framework. The observer uses linear system models and algebraic operations to estimate disturbances, avoiding the complex non-linear analysis required for feedforward cancellation and simplifying both implementation and performance verification
Data Source
AI summary
An electric motor of an electric power steering assembly is activated in response to a response signal. The response signal includes a baseline control torque signal generated by a baseline controller and an auxiliary control torque signal generated by a disturbance rejection controller. At least one operating parameter of the electric power steering assembly is estimated as a function of the baseline control torque signal. The at least one operating parameter is also measured. An error amount is determined as a function of the at least one measured operating parameter and the at least one estimated operating parameter. The disturbance rejection controller refines the at least one estimated operating parameter as a function of the error amount. The auxiliary control torque signal is generated to minimize the error amount.