Multi-Vehicle Motion Planning for Mixed-Traffic Conflict Zones
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for connected and automated vehicles (CAVs) lack an effective global multi-vehicle decision-making and motion planning solution that can handle mixed traffic scenarios with both autonomous and human-driven vehicles, particularly in dynamic environments with intersections and merging points, where safety and efficiency are compromised due to incomplete information and real-time control challenges.
Innovation Solution
A mixed-integer optimization-based global multi-vehicle decision-making system that computes target velocities and times for vehicles in interconnected conflict zones, incorporating real-time vehicle routing information and safety constraints, using V2X communication to coordinate both CAVs and non-controlled vehicles (NCVs), with modules implemented both on board and in infrastructure to optimize travel time, energy efficiency, and safety.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a global multi-vehicle decision-making system is implemented to coordinate CAVs in mixed traffic, then safety and efficiency are improved, but system complexity and computational requirements increase
Solution Approach 1:
The system segments the transportation network into multiple conflict zones (intersections and merging points) and processes vehicle coordination decisions in a decentralized manner across these zones. Each conflict zone has its own optimization problem, allowing parallel computation and reducing overall system complexity while maintaining global safety through coordinated timing information exchange.
Solution Approach 2:
The system introduces timing information as an intermediary variable that mediates between CAVs and NCVs at conflict zones. This timing information acts as a coordination signal that enables safe interaction without requiring direct control of NCVs, thereby improving safety while avoiding the complexity of controlling all vehicles in the network.
2Productivity
If real-time coordination of multiple vehicles is performed using optimization methods, then travel time and energy efficiency are improved, but computational time and processing requirements increase
Solution Approach 1:
The optimization problem is segmented into smaller sub-problems, one for each conflict zone, rather than solving a single large-scale optimization problem for the entire network. This segmentation enables parallel computation across zones and reduces the computational burden, allowing real-time solutions while maintaining productivity improvements.
Solution Approach 2:
The system performs preliminary computation of timing information and coordination decisions for each conflict zone before vehicles actually reach them. By pre-computing optimal trajectories and timing signals, the system reduces on-the-fly computational requirements and enables faster real-time response while improving overall travel time efficiency.
3Adaptability or versatility
If the system handles mixed traffic with both CAVs and NCVs, then adaptability to real-world scenarios is improved, but control authority and coordination capability are reduced
Solution Approach 1:
The system uses timing information as an intermediary to extend coordination capability to NCVs without requiring direct control authority over them. This timing information mediates the interaction between CAVs and NCVs at conflict zones, enabling the system to handle mixed traffic adaptably while maintaining clear distinctions between controlled and uncontrolled vehicles.
Solution Approach 2:
The system implements feedback mechanisms where timing information and coordination decisions are continuously updated based on the actual behavior of both CAVs and NCVs. This feedback allows the system to adapt to real-world mixed traffic scenarios while maintaining safety, even though full control authority over NCVs is not available.
4Measurement precision
If comprehensive vehicle information and routing data are collected for optimization, then decision-making accuracy is improved, but information processing load and communication requirements increase
Solution Approach 1:
The system segments information processing by conflict zone, collecting and processing only the routing data and vehicle states relevant to each specific zone. This segmented approach maintains high decision-making accuracy for each local optimization problem while reducing the overall information processing load compared to centralized network-wide processing.
Solution Approach 2:
The system applies local quality by processing information with different levels of detail appropriate to each conflict zone's specific requirements. Not all vehicles or all information are processed uniformly across the entire network, but rather with localized precision tailored to each zone's coordination needs, reducing overall information processing load while maintaining necessary accuracy.
Data Source
AI summary
A global multi-vehicle decision making system is disclosed for providing real-time motion planning and coordination of one or multiple connected and automated and/or semi-automated vehicles (CAVs) in an interconnected traffic network that includes one or multiple non-controlled vehicles (NCVs), one or multiple conflict zones and one or multiple conflict-free road segments. The system includes a receiver configured to acquire infrastructure sensing signals, at least one memory configured to store map and programs, and at least one processor configured to perform steps of formulating a global mixed-integer programming (MIP) problem using the infrastructure sensing signals, computing a motion plan for each CAV and each NCV in the traffic network by solving the global MIP problem, computing an optimal sequence of entering/exiting times and a sequence of average velocities for each CAV and each NCV, and computing a velocity profile and/or one or multiple planned stops for each CAV.