Mixed-Autonomy Platoon Control Using Headway-Based Indirect Steering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing traffic control systems face inefficiencies when manually operated vehicles join a platoon of autonomous vehicles, disrupting the formation and stability of the platoon, as existing methods struggle to effectively control and model manually operated vehicles, leading to increased traffic congestion and reduced energy efficiency.
Innovation Solution
A system and method for controlling mixed vehicle platoons, using a deep reinforcement learning (DRL) controller to indirectly influence uncontrolled vehicles by setting constraints on headway and speed, allowing controlled vehicles to maintain a platoon formation with both autonomous and manually operated vehicles, utilizing a headway-based model to map target headways to target speeds and actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If manually operated vehicles are allowed to join the platoon, then the platoon can accommodate mixed traffic types, but the platoon stability and control efficiency deteriorate
Solution Approach 1:
The patent introduces a behavior model as an intermediary layer between the control system and manually operated vehicles. This behavioral model predicts and captures the dynamics of manually operated vehicles, allowing the control system to indirectly influence them through controlled vehicles without requiring direct control of each manual vehicle, thus maintaining platoon stability while accommodating mixed traffic types
Solution Approach 2:
The system enables manually operated vehicles to self-regulate their platoon participation through indirect control mechanisms. By using controlled vehicles to create environmental cues (such as spacing and speed patterns), the system allows manual drivers to self-adjust their behavior to match platoon dynamics, reducing the need for direct intervention while maintaining stability
2Stability of the object's composition
If all vehicles are equipped with CACC, then platoon control homogeneity is improved, but the complexity and cost of the system increases
Solution Approach 1:
The patent creates a universal control framework that works with both CACC-equipped vehicles and manually operated vehicles. The behavior model and indirect control mechanisms allow the system to achieve homogeneous platoon control without requiring every vehicle to have identical CACC equipment, thus reducing overall system complexity while maintaining control homogeneity
Solution Approach 2:
Instead of requiring all vehicles to have advanced CACC systems, the patent inverts the approach by equipping only controlled vehicles with CACC and using them to indirectly control manually operated vehicles. This reversal reduces the number of complex systems needed while achieving the same control homogeneity goals
3Productivity
If manually operated vehicles are removed from the platoon, then platoon efficiency is improved, but the difficulty of detecting and measuring manual vehicle behavior prevents effective removal
Solution Approach 1:
The patent replaces direct mechanical control of manually operated vehicles with a behavioral modeling approach. Instead of attempting to directly measure and control each manual driver's actions, the system uses computational behavior models to predict and indirectly influence manual vehicle behavior, making the system more efficient while overcoming measurement difficulties
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
A system for direct and indirect control of mixed-autonomy vehicles receives a traffic state of a group of mixed-autonomy vehicles traveling in the same direction, wherein the group of mixed-autonomy vehicles includes controlled vehicles willing to participate in a platoon formation and at least one uncontrolled vehicle, and wherein the traffic state is indicative of a state of each vehicle in the group, submit the traffic state into a parameterized function trained to transform the traffic state into target headways for the mixed-autonomy vehicles to produce the target headways, and submit the target headways to a headway-based model configured to map the target headways to target speeds of the mixed-autonomy vehicles to produce the target speeds. The system determines and transmits control commands to the controlled vehicle based on one or combination of the target headways and the target speeds.