Electronic Control Suspension Using Deep Learning Road Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional active suspension systems face challenges in accurately reflecting real-time road surface conditions, leading to suboptimal damping control and reduced riding comfort due to reliance on GPS-based methods that require frequent updates and do not account for the actual shape of the road surface.
Innovation Solution
A deep learning-based method for controlling electronic control suspension systems that uses a road surface classification model to determine optimal control values by collecting and processing location information, image data, and physical sensor data to identify obstacles and correct characteristic values in real-time, incorporating convolutional neural networks for image processing and radar data analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPS-based methods are used to identify road surface conditions, then road surface information can be obtained, but the system requires frequent updates and cannot accurately reflect the actual shape of the road surface
Solution Approach 1:
The patent replaces GPS-based road surface identification with a camera-based vision system. The camera captures images of the road surface, and image processing algorithms analyze the visual data to identify road surface conditions and obstacle shapes. This substitution eliminates the need for frequent GPS updates and provides more accurate real-time road surface information.
Solution Approach 2:
The patent introduces an image processing system as an intermediary between the camera and the suspension control system. This intermediary processes camera images to extract road surface characteristics, obstacle types, and shape information, providing accurate real-time data to the suspension controller without requiring direct GPS measurements.
2Reliability
If conventional suspension systems use GPS and camera data, then obstacle information can be obtained, but the damping control is suboptimal due to inaccurate road surface shape information
Solution Approach 1:
The patent performs preliminary image processing and road surface analysis before the vehicle reaches the obstacle. The system captures images, processes them to identify obstacle characteristics and road surface conditions, and pre-calculates optimal damping control parameters. This preliminary action ensures accurate and timely damping control when the vehicle encounters the obstacle.
Solution Approach 2:
The patent implements a feedback mechanism where the suspension control system continuously receives updated road surface and obstacle information from the camera and image processing system. Based on this feedback, the system adjusts damping parameters in real-time to optimize suspension performance as the vehicle approaches and encounters obstacles.
Data Source
AI summary
The present disclosure relates to a method and an apparatus for controlling an electronic control suspension using a deep learning-based road surface classification model. The method for controlling an electronic control suspension in a vehicle including a camera and a GPS receiver may include collecting location information of the vehicle using the GPS receiver while driving, identifying whether there is a previously generated road surface classification model corresponding to a front obstacle when the front obstacle is detected, determining a first control value based on a first characteristic value corresponding to the road surface classification model when there is the road surface classification model as a result of the identification, controlling the electronic control suspension with the determined first control value when entering the obstacle, and collecting new sensing data through a physical sensor, and correcting the first characteristic value based on the new sensing data.