Multi-Agent Autonomous Vehicle Control for Planning and Fast Response
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional autonomous vehicle control systems struggle to balance high processing demands with cost, weight, power consumption, and adaptability requirements, leading to inefficiencies and reliability issues in real-time decision making and control.
Innovation Solution
A multiple agent autonomous vehicle control system is proposed, bifurcating into a deep computing subsystem for long-range planning and a fast response subsystem for immediate actions, each optimized for different processing loads and frequencies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If high-powered data processing systems are configured to handle processing loads, then processing capability and reliability are improved, but cost, weight, and power requirements increase
Solution Approach 1:
The autonomous vehicle control system is divided into two separate agents: a deep computing agent for low-frequency, high-complexity tasks and a fast response agent for high-frequency, low-complexity tasks. This segmentation allows each agent to be optimized for its specific function, reducing overall power consumption while maintaining reliability.
Solution Approach 2:
The system changes the operating parameters of processing units by having the deep computing agent operate at lower frequency with higher computational depth, while the fast response agent operates at higher frequency with simpler computations. This parameter optimization reduces power consumption compared to a single high-powered processor.
2Productivity
If high-powered data processing systems are configured to handle processing loads, then processing capability is improved, but system weight increases
Solution Approach 1:
The control system is segmented into two specialized agents rather than using a single high-powered processor. This allows the use of smaller, more efficient processing units that collectively provide the necessary processing capability while reducing overall system weight.
3Productivity
If high-powered data processing systems are configured to handle processing loads, then processing capability is improved, but cost increases
Solution Approach 1:
The system uses multiple smaller processing agents instead of a single expensive high-powered processor. This segmentation approach reduces component costs while maintaining overall processing capability through coordinated operation of the agents.
4Device complexity
If conventional single-agent systems are used for autonomous vehicle control, then device complexity is reduced, but responsiveness to emergency situations deteriorates
Solution Approach 1:
The control system is divided into two agents with different response characteristics. The fast response agent specifically handles time-critical tasks with minimal processing delay, improving overall system responsiveness while the deep computing agent handles less time-sensitive tasks.
Solution Approach 2:
The fast response agent performs partial processing of control tasks, handling only the most critical aspects that require immediate response. This allows the system to respond to emergencies quickly without requiring the entire system to be optimized for maximum speed.
5Loss of information
If deep computing subsystem handles all computational requests, then comprehensive long-range planning is achieved, but response time to high-frequency requests increases
Solution Approach 1:
Computational tasks are segmented and distributed to appropriate agents based on their characteristics. The deep computing agent handles complex, long-range planning tasks, while the fast response agent handles time-critical requests, optimizing both comprehensiveness and response time.
Solution Approach 2:
The deep computing agent performs preliminary computation of long-range plans in advance, allowing the fast response agent to make quick adjustments without needing to recalculate entire plans. This preliminary action reduces response time for high-frequency requests.
Data Source
AI summary
A system and method for providing multiple agents for decision making, trajectory planning, and control for autonomous vehicles are disclosed. A particular embodiment includes: partitioning a multiple agent autonomous vehicle control module for an autonomous vehicle into a plurality of subsystem agents, the plurality of subsystem agents including a deep computing vehicle control subsystem and a fast response vehicle control subsystem; receiving a task request from a vehicle subsystem; determining if the task request is appropriate for the deep computing vehicle control subsystem or the fast response vehicle control subsystem based on content of the task request or a context of the autonomous vehicle; dispatching the task request to the deep computing vehicle control subsystem or the fast response vehicle control subsystem based on the determination; causing execution of the deep computing vehicle control subsystem or the fast response vehicle control subsystem by use of a data processor to produce a vehicle control output; and providing the vehicle control output to a vehicle control subsystem of the autonomous vehicle.