Swarm Vehicle Anomaly Detection Using Inertial Sensor Channels
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting anomalies in swarm systems of vehicles, such as drones, are inadequate for real-time monitoring as the number of vehicles increases, posing safety risks due to potential malfunctions and crashes.
Innovation Solution
A method and apparatus using machine learning models, including supervised, semi-supervised, and unsupervised techniques, to analyze multi-channel inertial sensor data from vehicles in a swarm system, identifying abnormal states and channels in real-time through training and reconstruction error analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If direct visual inspection is used to detect anomalies in vehicles, then detection capability is maintained, but real-time monitoring becomes impossible as the number of vehicles increases
Solution Approach 1:
The patent replaces manual visual inspection with an automated machine learning-based detection system. The system collects movement data from multiple vehicles, trains learning models to recognize abnormal patterns, and automatically identifies anomalies without human intervention, thereby maintaining detection capability while enabling real-time monitoring of large-scale vehicle swarms.
Solution Approach 2:
The patent introduces movement data as an intermediary between the vehicle's operational state and the anomaly detection process. By collecting and analyzing movement data from sensors, the system creates a measurable representation of vehicle behavior that can be processed automatically, bridging the gap between physical vehicle states and digital detection capabilities.
2Adaptability or versatility
If the number of vehicles in the swarm system increases, then system capability is enhanced, but real-time anomaly monitoring becomes nearly impossible
Solution Approach 1:
The patent segments the anomaly detection process into distinct components: data collection from individual vehicles, training data generation, learning model development, and anomaly identification. This segmentation allows the system to handle large numbers of vehicles by processing data through standardized stages, reducing the overall complexity despite increased system scale.
Solution Approach 2:
The patent transforms physical vehicle movement characteristics into standardized data parameters that can be processed computationally. By converting movement data into training datasets and model inputs, the system changes the form of information from physical observations to digital parameters, enabling automated analysis regardless of swarm size.
3Measurement precision
If machine learning models are trained using multi-channel inertial sensor data, then detection accuracy is improved, but data processing complexity increases
Solution Approach 1:
The patent performs preliminary actions by pre-processing movement data into training datasets before model deployment. The system collects raw movement data, processes it into structured training formats, and trains learning models in advance, so that during actual operation, the trained models can quickly and accurately identify anomalies without requiring complex real-time processing of raw multi-channel data.
Data Source
AI summary
A method for detecting anomalies in a swarm system comprises: collecting first movement data from multiple vehicles moving as a swarm in a first scenario; generating first training data based on positioning data and second training data based on multi-channel inertial sensor data from the first movement data; training a first learning model using the first training data and multiple second learning models using the second training data for each vehicle; receiving real-time second movement data from vehicles moving as a swarm in a second scenario; generating first input data based on positioning data from the second movement data; inputting the first input data into the first learning model to detect abnormal vehicles in real-time; generating second input data for abnormal vehicles based on inertial sensor data from the second movement data; and inputting the second input data into the corresponding second learning model to identify abnormal channels in the inertial measurement unit of abnormal vehicles.