Behavior-Adaptive Path Control for Mobile Body Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional autonomous driving systems face challenges in accurately controlling mobile bodies due to the difficulty in collecting and processing diverse user behaviors, leading to inaccuracies in vehicle control, especially for infrequent behaviors.
Innovation Solution
A control device that stores multiple trained control models, each specifically trained for different user behaviors using machine learning, allowing for the selection and execution of the most appropriate path based on monitored user behavior, thereby improving control accuracy and simplifying the model structure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single trained control model is used for all user behaviors, then the model structure remains simple, but the accuracy of controlling the mobile body according to specific user behaviors deteriorates
Solution Approach 1:
The patent divides the single control model into multiple segmented control models, where each model is specifically trained for a particular user behavior type. This segmentation allows each model to specialize in recognizing and responding to specific behaviors, thereby improving control accuracy for each behavior category while maintaining manageable model complexity through modular organization.
2Measurement precision
If multiple trained control models are used for different user behaviors, then the accuracy of controlling the mobile body according to specific user behaviors is improved, but the device complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-training multiple control models for different user behaviors before actual operation. Each model is prepared in advance for specific behavior patterns, allowing the system to quickly select and apply the appropriate pre-trained model when a particular behavior is detected, thereby improving response accuracy without adding complex real-time processing requirements.
Solution Approach 2:
The system changes parameters by selecting different pre-trained models based on the detected user behavior type. Instead of using a single model with adjustable parameters, the system switches between models with fixed parameters optimized for specific behaviors, which simplifies the decision-making process and improves accuracy for each behavior category.
3Adaptability or versatility
If diverse user behaviors are collected and processed together, then the system can handle various behaviors, but the processing accuracy for infrequent behaviors deteriorates
Solution Approach 1:
By segmenting the control models according to specific user behavior types, the patent ensures that even infrequent behaviors receive dedicated model attention. Each segmented model is trained specifically on its corresponding behavior data, preventing dominant frequent behaviors from overwhelming the training process and ensuring accurate recognition and response for all behavior types including rare ones.
Data Source
AI summary
The control device according to one or more aspects of the present disclosure generates one or more paths by performing the calculation of one or more trained control models among the plurality of trained control models, monitors the behavior of the target user, selects one path from the generated one or more paths according to the result of monitoring the behavior of the target user, and controls the movement of the moving body according to the selected path. Each trained control model has acquired the ability to generate paths for controlling the movement of a mobile body related to the corresponding behavior of the user by machine learning.