Abstract:
Quadrotor unmanned aerial vehicles (UAVs) are playing important roles in cargo transportation, especially in complex and unknown environments. However, the underactuated ...Show MoreMetadata
Abstract:
Quadrotor unmanned aerial vehicles (UAVs) are playing important roles in cargo transportation, especially in complex and unknown environments. However, the underactuated nature of the quadrotor UAV transportation systems makes it difficult for control schemes design. Considering system parameter uncertainties, a novel nonlinear adaptive control law is designed via Immersion and Invariance (I&I) method. Besides, in order to pursue faster convergence and stronger robustness, a novel control structure based on deep reinforcement learning is presented. The RL agent, which is designed through twin delayed deep deterministic policy gradient (TD3) algorithm, generates optimal control gains for the I&I adaptive controller in real-time. The stability of the closed-loop system are guaranteed through Lyapunov technique and LaSalle’s invariance theorem. Finally, by comparing with the classical control scheme, the superior performance and stronger robustness of the proposed controller are validated through numerous simulations.
Date of Conference: 09-11 July 2022
Date Added to IEEE Xplore: 29 November 2022
ISBN Information:
Funding Agency:
Keywords assist with retrieval of results and provide a means to discovering other relevant content. Learn more.
- IEEE Keywords
- Index Terms
- Deep Learning ,
- Transport System ,
- Adaptive Control ,
- Deep Reinforcement Learning ,
- Control Strategy ,
- Shipment ,
- Control Structure ,
- Gain Control ,
- Unmanned Aerial Vehicles ,
- Adaptive Law ,
- Reinforcement Learning Agent ,
- Deterministic Policy Gradient ,
- Adaptive Control Law ,
- Policy Gradient Algorithm ,
- Time Step ,
- System Dynamics ,
- Test Group ,
- Control Design ,
- Control Problem ,
- Unknown Parameters ,
- Aerodynamic Coefficients ,
- Actor Network ,
- Convergence Of System ,
- Proportional-integral-derivative ,
- Goal Of This Section ,
- Critic Network ,
- Adaptive Control Scheme ,
- Position Error ,
- Equilibrium Point ,
- Positive Definite Matrix
Keywords assist with retrieval of results and provide a means to discovering other relevant content. Learn more.
- IEEE Keywords
- Index Terms
- Deep Learning ,
- Transport System ,
- Adaptive Control ,
- Deep Reinforcement Learning ,
- Control Strategy ,
- Shipment ,
- Control Structure ,
- Gain Control ,
- Unmanned Aerial Vehicles ,
- Adaptive Law ,
- Reinforcement Learning Agent ,
- Deterministic Policy Gradient ,
- Adaptive Control Law ,
- Policy Gradient Algorithm ,
- Time Step ,
- System Dynamics ,
- Test Group ,
- Control Design ,
- Control Problem ,
- Unknown Parameters ,
- Aerodynamic Coefficients ,
- Actor Network ,
- Convergence Of System ,
- Proportional-integral-derivative ,
- Goal Of This Section ,
- Critic Network ,
- Adaptive Control Scheme ,
- Position Error ,
- Equilibrium Point ,
- Positive Definite Matrix