LSTM RNN Sensing for HDV Detection in Partial VANET
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In vehicular ad hoc networks, human-driven vehicles (HDVs) without communication devices are not detectable by the system, which hinders data routing efficiency and interaction with connected and autonomous vehicles (CAVs) during dynamic platoon formations.
Innovation Solution
A sensing method using LSTM RNN-based modules in CAVs to infer the existence and location of HDVs by analyzing historical motion information, with confliction criteria for fusion of estimation results from multiple CAVs, and deploying this method within CAVs or roadside units to obtain HDV information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If CAVs are equipped with communication devices to exchange information, then information exchange capability is improved, but HDVs without communication devices remain undetectable
Solution Approach 1:
The patent introduces an intermediary sensing system that indirectly detects HDVs by monitoring their effects on CAVs. Instead of directly detecting HDVs, the system uses CAVs as mediators - observing how HDVs influence CAV motion states and using this information to infer HDV presence and location.
Solution Approach 2:
The patent replaces direct communication/detection mechanisms with indirect inference based on motion analysis. Instead of using communication devices to directly detect HDVs, the system substitutes mechanical motion observation and neural network-based inference to achieve HDV detection.
2Area of stationary object
If multiple CAVs are used to sense HDVs, then detection coverage is improved, but confliction in estimation results increases
Solution Approach 1:
The patent implements feedback mechanisms where estimation results from multiple CAVs are continuously compared and validated. The system uses confliction criteria to evaluate consistency and adjusts estimations based on feedback from multiple sources, improving reliability through iterative validation.
Solution Approach 2:
The patent merges estimation results from multiple CAVs using fusion techniques. By combining data from multiple sources and applying confliction criteria, the system integrates individual estimations into a more reliable collective assessment of HDV presence and location.
3Speed
If LSTM RNN modules are deployed in each CAV, then real-time sensing capability is improved, but system complexity increases
Solution Approach 1:
The patent employs universal LSTM RNN modules that can be deployed in any CAV. These multi-functional modules handle both existence sensing and location estimation tasks, reducing overall system complexity through standardization while maintaining real-time capability.
Solution Approach 2:
The patent segments the sensing system into modular LSTM RNN components that can be independently deployed in each CAV. This segmentation allows distributed real-time processing while keeping individual module complexity manageable through modular design.
Data Source
AI summary
A sensing method and a sensing device for HDVs (human driven vehicles) under a partial VANET environment are provided. According to the method, an existence sensing module and a location sensing module are constructed, based on a long-short-term-memory recurrent neural network, with utilizing historical information of motion states of a single CAV (connected and autonomous vehicle) as an input, existence and exact locations of surrounding HDVs of the CAV are outputted. The method is not only applicable to sensing the surrounding HDVs of the single CAV, but also the surrounding HDVs of the multiple CAVs. An estimation result of each CAV is firstly obtained, then the estimation results of the CAVs are checked with confliction criterion, according to checking results, the estimation results of the multiple CAVs are fused, and information about the surrounding HDVs of each CAV is finally outputted.