Vehicle Lateral Velocity Estimation Using LSTM for ESP Stability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in accurately estimating the lateral velocity of a vehicle, especially when the vehicle is driving on curved roads, due to the dependency between longitudinal and lateral velocities, and the difficulty in directly measuring lateral velocity.
Innovation Solution
An apparatus and method utilizing a long short-term memory (LSTM) model in a multiple input single output (MISO) type to estimate the lateral velocity of a vehicle, by obtaining longitudinal velocity, lateral acceleration, and yaw rate, and applying these inputs to the LSTM model to improve accuracy and stability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If lateral velocity is estimated based on longitudinal velocity, vehicle specifications, and tire characteristics, then the estimation can be performed, but accuracy deteriorates when slip angle increases due to tire nonlinearity and yaw behavior
Solution Approach 1:
The patent introduces an intermediary sensor system (GPS receiver and inertial measurement unit) that directly measures lateral velocity components rather than calculating them from longitudinal velocity and tire models. This intermediary measurement approach bypasses the nonlinear tire characteristic problems that occur at high slip angles, providing more reliable lateral velocity data for ESP control.
2Device complexity
If conventional sensors are used for vehicle velocity measurement, then the system is simple, but accuracy deteriorates during wheel slip conditions
Solution Approach 1:
The patent merges multiple sensor systems (wheel speed sensors, GPS receiver, and inertial measurement unit) into an integrated velocity measurement system. This combination allows the system to cross-validate measurements and maintain accuracy during wheel slip conditions where conventional wheel speed sensors alone would provide erroneous data.
Solution Approach 2:
The inertial measurement unit serves as an intermediary that provides independent velocity measurements not dependent on wheel rotation. This intermediary measurement source remains reliable during wheel slip when the wheels are rotating at speeds that do not reflect actual vehicle motion.
3Measurement precision
If LSTM model with multiple inputs is used for lateral velocity estimation, then estimation accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent performs preliminary processing of input data (longitudinal velocity, lateral acceleration, yaw rate) before feeding them to the LSTM model. This preliminary action includes data filtering, normalization, and feature selection, which reduces the complexity of the LSTM computation while maintaining estimation accuracy by providing pre-processed, high-quality inputs.
Data Source
AI summary
An apparatus for estimating a lateral velocity of a vehicle and a method for the same are provided to estimate the lateral velocity of the vehicle with higher accuracy, and to be applied to an electronic stability program (ESP) device to improve the driving stability of the vehicle, by including a long short-term memory (LSTM) model trained to estimate the lateral velocity of the vehicle, obtaining a longitudinal velocity, a lateral acceleration, and a yaw rate of the vehicle, and estimating the lateral velocity of the vehicle, which corresponds to the longitudinal velocity, the lateral acceleration, and the yaw rate of the vehicle, based on the LSTM model.