Active Wheel Suspension Velocity Estimation Without High-Pass Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for determining displacement velocity signals for active wheel suspension control of land vehicles are inadequate, leading to suboptimal ride comfort due to inaccurate control force requirements and waveform distortion caused by high-pass filtering.
Innovation Solution
A system utilizing a Kalman filter with a mathematical motion model of the sprung mass, supplemented by a displacement measurement signal of 0 and high noise variance values, to generate a high-quality displacement velocity signal for active wheel suspension control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-pass filtering is applied to remove drift from integrated acceleration signal, then drift is reduced, but waveform distortion occurs and low-frequency components are removed
Solution Approach 1:
The patent introduces a virtual displacement measurement signal with value 0 as an intermediary to the Kalman filter. This virtual measurement allows the filter to operate without high-pass filtering, preserving low-frequency components while still achieving drift removal through the filter's inherent integration and prediction capabilities.
Solution Approach 2:
The patent replaces the conventional mechanical signal processing approach (integration followed by high-pass filtering) with a computational approach using a Kalman filter. This substitution eliminates the need for high-pass filtering while maintaining drift removal, as the Kalman filter naturally handles the integration without introducing waveform distortion.
2Measurement precision
If Kalman filter is used with mathematical vehicle model, then displacement velocity signal accuracy is improved, but computational load increases
Solution Approach 1:
The patent extracts and removes the complex mathematical vehicle model from the Kalman filter, retaining only the essential acceleration measurement and the virtual displacement measurement. This extraction maintains the accuracy benefits of Kalman filtering while eliminating the high computational load associated with complex vehicle dynamics models.
Solution Approach 2:
The patent uses a simplified, computationally inexpensive approach by employing a virtual displacement measurement signal with value 0 instead of complex vehicle models. This disposable-like simplification achieves the desired accuracy without the burden of maintaining and computing complex models.
3Measurement precision
If integration of acceleration signal is performed, then displacement velocity is obtained, but drift is introduced
Solution Approach 1:
The Kalman filter implements feedback by continuously comparing the predicted displacement velocity (from integration) with the virtual displacement measurement (value 0) and adjusting the estimate accordingly. This feedback mechanism naturally removes drift without requiring high-pass filtering, as the filter's prediction-correction cycle inherently compensates for integration errors.
Data Source
AI summary
A system for determining a displacement velocity signal for controlling an active wheel suspension of a land vehicle by open-loop and/or closed-loop control includes at least one Kalman filter, and at least one acceleration sensor arranged on a sprung mass of the land vehicle to sense a vertical acceleration of the sprung mass and to generate a corresponding acceleration signal supplied to the Kalman filter. The Kalman filter includes a mathematical motion model of the sprung mass, and input states of the Kalman filter include a vertical acceleration of the sprung mass, a vertical displacement velocity of the sprung mass, and a vertical displacement distance of the sprung mass. A displacement measurement signal having a value 0 is supplied continuously to the Kalman filter to determine the displacement velocity signal. Constant noise variance values of a measurement noise covariance matrix of the Kalman filter that are assigned to the displacement measurement signal are, in each case, set at one half of a maximum vertical displacement distance of the sprung mass.