Mobile Robot Group Control With Acceleration-Based Position Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mobile object group control systems fail to accurately account for acceleration and deceleration, leading to deviations between estimated and actual positions of mobile objects, which can result in collisions and increased costs when multiple autonomous mobile robots operate in narrow areas.
Innovation Solution
A group control system and method that incorporates acceleration sensors and deep reinforcement learning to correct position estimation deviations by training a model with acceleration data, ensuring accurate position information and preventing collisions among autonomous mobile robots.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If acceleration and deceleration are not considered in position estimation, then the control system is simpler, but the position information becomes inaccurate leading to collisions
Solution Approach 1:
The patent introduces an acceleration sensor as an intermediary device that measures acceleration data, which is then fed into the deep reinforcement learning model. This mediator bridges the gap between simple control architecture and accurate position estimation by providing additional dynamic information without requiring complete redesign of the control system.
Solution Approach 2:
The patent changes the parameter set used for position estimation by incorporating acceleration data alongside position sensor data. The deep reinforcement learning model processes these changed parameters (acceleration and position) to generate corrected position estimates, thereby improving accuracy while maintaining a relatively simple control architecture.
2Reliability
If multiple autonomous mobile robots operate in narrow areas without considering acceleration data, then the layout can be simpler, but collisions occur and layout changes are required increasing costs
Solution Approach 1:
The system performs preliminary position correction by training a deep reinforcement learning model that predicts position deviations caused by acceleration and deceleration before actual collisions occur. This preliminary action allows the robots to maintain safe distances in narrow layouts without requiring costly layout modifications.
Solution Approach 2:
The patent implements feedback by continuously monitoring acceleration data from sensors and using it to correct position estimates in real-time. This feedback loop enables the control system to adapt to dynamic conditions and prevent collisions, maintaining reliability without changing the physical layout.
3Measurement precision
If deep reinforcement learning model is trained with acceleration data, then position estimation accuracy improves, but computational requirements and training complexity increase
Solution Approach 1:
The deep reinforcement learning model serves itself by learning from acceleration data patterns during training and then autonomously applying this learned knowledge to correct position estimates during operation. This self-service capability allows the system to handle the computational complexity internally without requiring external intervention or simplified model architectures.
Data Source
AI summary
The group control system controls a plurality of mobile objects capable of autonomously traveling in a predetermined area. The group control system includes a position information estimation unit that estimates position information of each mobile object, a route planning unit that creates a route plan of each mobile object based on the estimated position information, an acceleration data acquisition unit that acquires acceleration data acquired from the acceleration sensor, and a mobile object position acquisition unit that acquires an actual position of each mobile object using the position sensor, and learns the deep reinforcement learning model so as to correct a deviation amount between an actual position of the mobile object acquired using the position sensor and an estimated position of the mobile object based on the acquired acceleration data.