In the context of autonomous driving, acquiring the trajectories of surrounding vehicles in advance by autonomous vehicles is a crucial factor in ensuring high-level road safety. While trajectory prediction methods ba...
详细信息
In the context of autonomous driving, acquiring the trajectories of surrounding vehicles in advance by autonomous vehicles is a crucial factor in ensuring high-level road safety. While trajectory prediction methods based on deep learning have achieved promising results, these data-driven models lack interpretability and transparency, making their reliable use a significant challenge. In this paper, firstly, an intention-aware spatial-temporal attention network-based trajectory prediction model is constructed, which considers the coupling of driving intention and the interaction with surrounding vehicles, extracts important feature information of vehicles in both temporal and spatial dimensions. Secondly, a vehicle trajectory prediction method via the integration of data-driven and knowledge-guided is proposed, considering both hard and soft constraints. A hard constraint of vehicle kinematics is incorporated into the intention-aware spatial-temporal attention network prediction model to generate physically feasible predicted trajectories and to make this part of the network structure have a human-understandable physical meaning. In addition, by leveraging knowledge related to traffic rules, an auxiliary loss function based on knowledge constraint penalties is designed as a soft constraint to optimize the training of the model and improve the interpretability of the training process. Finally, the proposed model is experimentally evaluated on the datasets and the prediction results are analyzed in terms of reliability and accuracy. The experimental results demonstrate that knowledge guidance effectively enhance the reliability and interpretability of the prediction, and improve the accuracy of long-term trajectory prediction.
In this brief, we present a learning-based tracking controller based on Gaussian processes (GPs) for a fault-tolerant hexarotor in a recovery maneuver. In particular, we use GPs to estimate certain uncertainties that ...
详细信息
In this brief, we present a learning-based tracking controller based on Gaussian processes (GPs) for a fault-tolerant hexarotor in a recovery maneuver. In particular, we use GPs to estimate certain uncertainties that appear in a hexacopter vehicle with the ability to reconfigure its rotors to compensate for failures. The rotor's reconfiguration introduces disturbances that make the dynamic model of the vehicle differ from the nominal model. The control algorithm is designed to learn and compensate for the amount of modeling uncertainties after a failure in the control allocation reconfiguration by using GP as a learning-based model for the predictions. In particular, the presented approach guarantees a probabilistic bounded tracking error with high probability. The performance of the learning-based fault-tolerant controller is evaluated by experimental tests with a hexarotor unmanned aerial vehicle (UAV).
In order to effectively reduce the fluctuation of blade loads during independent pitch change of large wind turbines, a real-time correction model predictive control (MPC) control strategy for independent pitch contro...
详细信息
High-dimensional chaotic systems possess complex structural characteristics, which have potential research value in fields like encrypted communication and chaos control. The inverse problem of parameters in high-dime...
详细信息
Lately, nonlinear model predictive control (NMPC) has been successfully applied to (semi-) autonomous driving problems and has proven to be a very promising technique. However, accurate control models for real vehicle...
详细信息
Lately, nonlinear model predictive control (NMPC) has been successfully applied to (semi-) autonomous driving problems and has proven to be a very promising technique. However, accurate control models for real vehicles could require costly and time-demanding specific measurements. To address this problem, the exploitation of system data to complement or derive the prediction model of the NMPC has been explored, employing learning dynamics approaches within learning-based NMPC (LbNMPC). Its application to the automotive field has focused on discrete gray-box modeling, in which a nominal dynamics model is enhanced by the data-driven component. In this manuscript, we present an LbNMPC controller for a real go-kart based on a continuous black-box model of the accelerations obtained by Gaussian processes (GP). We show the effectiveness of the proposed approach by testing the controller on a real go-kart vehicle, highlighting the approximation steps required to get an exploitable GP model on a real-time application.
To address the challenges of sensor fusion and safety risk prediction, contemporary closed-loop autonomous driving neural networks leveraging imitation learning typically require a substantial volume of parameters and...
详细信息
Based on the graph theory, multi-UAV form a formation through the leader-follower method. The control command is sent to the leader UAV, and the formation of UAVs is led by the leader to the target location and avoids...
详细信息
ISBN:
(纸本)9798350321050
Based on the graph theory, multi-UAV form a formation through the leader-follower method. The control command is sent to the leader UAV, and the formation of UAVs is led by the leader to the target location and avoids obstacles through an improved artificial potential field method, and the formation is controlled by a consistency protocol to keep the formation stable during the flight. If the leader UAV is balanced by forces at a certain place when it does not reach the target point, a random disturbance is triggered to make it jump out of the local minimum. Through the formation consistency obstacle avoidance algorithm, the formation of UAVs is kept constant during the obstacle avoidance process to achieve formation position consistency control. Finally, the effectiveness of the algorithm is verified by MATLAB simulation.
Distribution systems have limited observability, as they were a passive grid to consume power. Nowadays, increasing distributed energy resources turns individual customers into "generators," and two-way powe...
详细信息
Distribution systems have limited observability, as they were a passive grid to consume power. Nowadays, increasing distributed energy resources turns individual customers into "generators," and two-way power flow between customers makes the grid prone to power outages. This calls for new control methods with performance guarantees in the presence of limited system information. However, limited system information makes it difficult to employ model-based control, making performance guarantees difficult. To gain information about the model, active learning methods propose to disturb the system consistently to learn the nonlinearity. The exploration process also introduces uncertainty for further outages. To address the issue of frequent perturbation, we propose to disturb the system with decreasing frequency by minimizing exploration. Based on such a proposal, we superposed the design with a physical kernel to embed system non-linearity from power flow equations. These designs lead to a highly robust adaptive online policy, which reduces the perturbation gradually but monotonically based on the optimal control guarantee. For extensive validation, we test our controller on various ieee test systems, including the 4-bus, 13-bus, 30-bus, and 123-bus grids, with different penetrations of renewables, various set-ups of meters, and diversified regulators. Numerical results show significantly improved voltage control with limited perturbation compared to those of the state-of-the-art data-driven methods.
This article concentrates on adaptive tracking control of strict-feedback uncertain nonlinear systems with an event-based learning scheme. A novel neural network (NN) learning law is proposed to design the adaptive co...
详细信息
This article concentrates on adaptive tracking control of strict-feedback uncertain nonlinear systems with an event-based learning scheme. A novel neural network (NN) learning law is proposed to design the adaptive control scheme. The NN weights information driven by the prediction-error-based control process is intermittently transmitted in the event-triggered context to the NN learning law mainly for signal tracking. The online stored sampled data of NN driven by the tracking error are utilized in the event context to update the learning law. With the adaptive control and NN learning law updated via the event-triggered communication, the improvements of NN learning capability, tracking performance, and system computing resource saving are guaranteed. In addition, it is proved that the minimum time interval for triggering errors of the two types of events is bounded and the Zeno behavior is strictly excluded. Finally, simulation results illustrate the effectiveness and good performance of the proposed control method.
This paper investigates the containment control problem of a class of quadrotor unmanned aerial vehicle (QUAV) systems under directed topologies. Firstly, an integral chain model with unknown control directions is con...
详细信息
暂无评论