Vehicle Control Unit Using Deep Learning for Route Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles rely solely on local environment sensing, lacking the capability to consider learned driving experiences and behaviors, which limits their operational efficiency and adaptability beyond immediate surroundings.
Innovation Solution
A vehicle control unit processes driving map data to produce vehicle operational data, using a deep neural network trained on temporal data sets that include vehicle sensor data, enabling the vehicle to traverse routes by recognizing obstacles and making decisions based on learned experiences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If autonomous vehicles rely solely on local environment sensing, then real-time response capability is maintained, but operational efficiency and adaptability beyond immediate surroundings deteriorate
Solution Approach 1:
The patent segments the decision-making process into two distinct parts: (1) real-time local environment sensing and obstacle detection, and (2) strategic route planning using pre-processed map data and learned driving experiences. This segmentation allows the vehicle to maintain fast real-time response while simultaneously incorporating comprehensive learned experiences for improved operational efficiency.
Solution Approach 2:
The system performs preliminary processing of map data and training of deep neural networks offline before actual vehicle operation. By pre-processing route information and pre-training decision-making models with extensive driving experiences, the vehicle can quickly access and apply learned knowledge during real-time operation without compromising response speed.
2Adaptability or versatility
If comprehensive map data and learned experiences are integrated into real-time processing, then adaptability improves, but computational complexity and processing time increase
Solution Approach 1:
The patent performs extensive data processing, map analysis, and neural network training in advance, before the vehicle needs to make real-time decisions. By completing the computationally intensive tasks of processing comprehensive map data and training deep neural networks offline, the system reduces the computational burden during actual vehicle operation while maintaining high adaptability.
Solution Approach 2:
The system introduces an intermediary layer (the trained deep neural network and pre-processed map data structure) that mediates between the complex input data and real-time decision requirements. This intermediary has already performed the heavy computational work of integrating comprehensive map data and learned experiences, presenting simplified, pre-processed information to the real-time control system.
3Measurement precision
If forward sensing and local environment recognition are used, then immediate obstacle detection is achieved, but strategic route planning and learned behavior application are limited
Solution Approach 1:
The patent divides the autonomous driving system into specialized modules: one dedicated to precise real-time obstacle detection through forward sensing, and another dedicated to strategic route planning using comprehensive map data and learned driving experiences. This functional segmentation allows each module to excel at its specific task without compromise.
Solution Approach 2:
The system merges the outputs of local environment sensing with pre-processed map data and learned driving behaviors through a trained deep neural network. By combining real-time sensor data with strategically planned route information and learned experiences, the system achieves both precise obstacle detection and intelligent strategic decision-making.
Data Source
AI summary
A method and device for effecting vehicle control is presented. The method includes receiving driving map data, which includes vehicle target data relative to a vehicle location data. The driving map data is processed to produce desired vehicle operational data, the desired vehicle operational data facilitates a vehicle to traverse a travel route. From the desired vehicle operational data vehicle, corresponding actuator control data is produced, and transmitted to effect vehicle control in either of the autonomous or driver-assisted modes.