Multi-Agent Formation Control Under Faults and Sensor Deception
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing formation maneuver control methods for nonlinear multi-agent systems, particularly in unmanned vehicles, are inadequate in addressing cyber-attacks and actuator faults, leading to instability and performance degradation.
Innovation Solution
A system utilizing reinforcement learning neural networks, specifically an identifier, actor, and critic radial basis function neural networks, to estimate and adjust vehicle movements, while compensating for actuator faults and deception attacks, ensuring stable leader-follower maneuvers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional formation maneuver control methods are used, then the system is simpler to implement, but the system becomes vulnerable to cyber-attacks and actuator faults leading to instability
Solution Approach 1:
The patent introduces an intermediary observer system that mediates between the control input and the plant dynamics. This observer estimates the state of the system and the attack signals, acting as a buffer that protects the formation control from direct exposure to deception attacks and actuator faults, thereby maintaining stability without requiring complete redesign of the control architecture
Solution Approach 2:
The patent implements feedback mechanisms where the observer continuously monitors the system state and attack signals, and adjusts the control input accordingly. The estimated attack signals are fed back to compensate for their effects, creating a closed-loop system that actively counteracts disturbances and maintains formation stability despite the presence of cyber-attacks and faults
2Reliability
If fault-tolerant control methods are implemented, then the system becomes more robust to actuator faults, but the control approach becomes less effective against deception attacks
Solution Approach 1:
The patent designs a universal control framework that serves multiple functions simultaneously: it provides fault-tolerant control for actuator faults while also offering protection against deception attacks. The observer-based compensation mechanism is versatile enough to handle both types of disturbances, eliminating the need for separate specialized control strategies for each threat type
3Adaptability or versatility
If reinforcement learning neural networks are used, then the system achieves better adaptability to disturbances, but the computational complexity and resource requirements increase
Solution Approach 1:
The patent employs lightweight neural network architectures that are computationally efficient and suitable for embedded systems. Rather than using complex deep learning models, the implementation uses simpler neural network structures that can be trained offline and deployed with minimal computational resources, making them suitable for real-time control applications on resource-constrained unmanned vehicles
Data Source
AI summary
A system, computer readable storage medium and method for controlling a trajectory of coordinated time-varying maneuvers of a geometric formation of unmanned vehicles is disclosed. The system includes unmanned vehicles, each configured with communication circuitry to communicate between the vehicles. A subset of the unmanned vehicles function as leader vehicles, with the remaining vehicles functioning as follower vehicles for leader-follower maneuvering. The system further includes an actuator suite configured to adjust the direction and orientation of each vehicle, a sensor suite for stabilization and navigation, and a flight controller for maintaining stable maneuvering, even in the presence of actuator faults and sensor deception attacks. Processing circuitry is configured with a reinforcement learning neural network that includes identifier, actor, and critic radial basis function neural networks to estimate movement, adjust control actions, and assess vehicle performance based on feedback signals, including corrupted signals from the sensor suite due to deception attacks.