Control Of Autonomous Vehicles Research Articles

Provably Safe Design Methodology for Automatic Longitudinal Control of Autonomous Vehicles

Provably Safe Design Methodology for Automatic Longitudinal Control of Autonomous Vehicles

Optimal Tracking Control for Autonomous Vehicle With Prescribed Performance via Adaptive Dynamic Programming

Optimal Tracking Control for Autonomous Vehicle With Prescribed Performance via Adaptive Dynamic Programming

Fast Online Computation of MPC-based Integrated Decision Control for Autonomous Vehicles

Fast Online Computation of MPC-based Integrated Decision Control for Autonomous Vehicles

Efficient Path Planning and Tracking Control of Autonomous Vehicles in Park Scenarios Based on Sparrow Search and Stepwise Prediction

Efficient Path Planning and Tracking Control of Autonomous Vehicles in Park Scenarios Based on Sparrow Search and Stepwise Prediction

Motion control of autonomous vehicles after a tyre blow-out based on differential-flatness-MPC

Motion control of autonomous vehicles after a tyre blow-out based on differential-flatness-MPC

Learn Zero-Constraint-Violation Safe Policy in Model-Free Constrained Reinforcement Learning.

We focus on learning the zero-constraint-violation safe policy in model-free reinforcement learning (RL). Existing model-free RL studies mostly use the posterior penalty to penalize dangerous actions, which means they must experience the danger to learn from the danger. Therefore, they cannot learn a zero-violation safe policy even after convergence. To handle this problem, we leverage the safety-oriented energy functions to learn zero-constraint-violation safe policies and propose the safe set actor-critic (SSAC) algorithm. The energy function is designed to increase rapidly for potentially dangerous actions, locating the safe set on the action space. Therefore, we can identify the dangerous actions prior to taking them and achieve zero-constraint violation. Our major contributions are twofold. First, we use the data-driven methods to learn the energy function, which releases the requirement of known dynamics. Second, we formulate a constrained RL problem to solve the zero-violation policies. We prove that our Lagrangian-based constrained RL solutions converge to the constrained optimal zero-violation policies theoretically. The proposed algorithm is evaluated on the complex simulation environments and a hardware-in-loop (HIL) experiment with a real autonomous vehicle controller. Experimental results suggest that the converged policies in all environments achieve zero-constraint violation and comparable performance with model-based baseline.

Level 4 commercial autonomous vehicle control system transition to an open-source solution

Level 4 commercial autonomous vehicle control system transition to an open-source solution

A hybrid technique for an autonomous vehicle control system to enhance the vehicle robustness: a HBA-RBFNN technique

A hybrid technique for an autonomous vehicle control system to enhance the vehicle robustness: a HBA-RBFNN technique

Connected Autonomous Vehicle Control Strategy for Mandatory Lane-Changing At Intersection: A Cyber-Physical System Perspective

Connected Autonomous Vehicle Control Strategy for Mandatory Lane-Changing At Intersection: A Cyber-Physical System Perspective

Analysis of Autonomous Vehicle Control Algorithm Based on Different Models

As a matter of fact, the rise of autonomous vehicles (AVs) has necessitated the development of advanced control algorithms to ensure safe and efficient operation in recent years. With this in mind, the aim of this paper is to provide a comprehensive analysis of control algorithms for self-driving vehicles and to identify research opportunities for innovation by comparing the strengths and weaknesses of existing algorithms. On this basis, through a careful dissection of existing literature as well as case studies, this study provides an in-depth understanding of different control algorithms and a comparative analysis of performance metrics from real-world tests and simulations. To be specific, three models will be applied. According to the analysis, the results of this study emphasize the critical role of accurate vehicle dynamics in the development of complex automated driving control algorithms at the same time, paving the way for safer, more efficient as well as sustainable transportation solutions.

Method of Autonomous Vehicle Control Using Simplified Reference Models and Regulators

This study contains a conceptual solution of the control method for an autonomous vehicle based on a simplified reference model and controllers. The whole issue is considered on the example of the execution phase of the lane change manoeuvre, in which the well-known 4WS (Four-Wheel-Steering) bicycle model and Kalman regulators based on the LQR (Linear-Quadratic Regulator) technique were used. The overall developed control mechatronic system was subjected to simulation studies in Matlab&Simulink. The simulation results showed correct functioning of the entire mechatronic control system and allowed us to determine further research directions. The developed control method can also find application in military vehicles.

Tracking control of AUV via novel soft actor-critic and suboptimal demonstrations

Tracking control for autonomous underwater vehicles (AUVs) faces multifaceted challenges, making the acquisition of optimal demonstrations a daunting task. The suboptimal demonstrations mean less tracking accuracy. To address the issue of learning from suboptimal demonstrations, this paper proposes a model-free reinforcement learning (RL) method. Our approach utilizes suboptimal demonstrations to obtain an initial controller, which is iteratively refined during training. Given the suboptimal characteristics, demonstrations will be removed from the replay buffer upon reaching capacity. Building upon the soft actor-critic (SAC), our approach integrates a Recurrent Neural Network (RNN) into the policy network to capture the relationship between states and actions. Moreover, we introduce logarithmic and cosine functions to the reward function for enhancing the training effectiveness. Finally, we validate the effectiveness of the proposed Initialize Controller from Demonstrations (ICfD) algorithm through simulations with two reference trajectories. We provide a definition for tracking success. The success rates of ICfD in the two reference trajectories are 95.60% and 94.05%, respectively, surpassing the state-of-the-art RL method SACfD (80.03% 90.55%). The average one-step distance errors of ICfD are 1.20 m and 0.76 m, respectively, significantly lower than the S-plane controller (9.725 m 8.325 m). Besides, we evaluate the generalization of the ICfD controller in different scenarios.

Adversarial deep reinforcement learning based robust depth tracking control for underactuated autonomous underwater vehicle

In this paper, an adversarial deep reinforcement learning-based control method is proposed to address the issue of robust depth tracking of an underactuated autonomous underwater vehicle in the presence of intrinsic coupled dynamics and external disturbances. First, long-short-term-memory neural network is presented to memorize and predict the changes in the state of vehicle, and a cascaded multilayer perception projects the output into action space of vehicle. Subsequently, adversarial deep reinforcement learning scheme is applied to the training of control agent by introducing an adversary which counteracts the control behavior, whereby the agent is enabled to learn the control strategy in different distributions of state transition. For evaluation of the performance, a control agent is pre-trained in simulation environment based on the reliable digital model of a real vehicle, and the simulation environment is paced at one iteration per second to align with real-time operations to ensure the portability of training result. Furthermore, the training cost is also extremely reduced. Finally, experiments are conducted with time-varying disturbances to further prove the feasibility of the proposed learning-based control scheme on a prototype of underwater vehicle in towing tank. Moreover, comparative experiment results show the better robustness performance of the learning-based control agent than that of classic line-of-sight based proportional–integral–derivative and adaptive line-of-sight based proportional–integral–derivative controllers in different scenarios.

Preface: Workshop “NGC of AV: Navigation, Guidance and Control of Autonomous Vehicles”

Preface: Workshop “NGC of AV: Navigation, Guidance and Control of Autonomous Vehicles”

Preface: Workshop “NGC of AV: Navigation, Guidance and Control of Autonomous Vehicles”

Preface: Workshop “NGC of AV: Navigation, Guidance and Control of Autonomous Vehicles”

Trajectory tracking control for underactuated autonomous vehicles via adaptive dynamic programming

In this paper, the trajectory tracking control problem for underactuated autonomous vehicles moving in 2-D space is studied. Firstly, the trajectory tracking error dynamics with the equivalent point being the origin are given. Since both model and input uncertainties of vehicles are considered, a robust control problem is formulated by giving the extended trajectory tracking error dynamics. Through showing that a robust control law is equivalent to an optimal control law for a nominal system under some conditions, the robust control problem is transformed into an equivalent optimal control problem and a cost function composed of uncertainty bounds and discount factor is constructed. In order to solve the optimal control problem, a policy iteration-based ADP algorithm is proposed, in which two neural networks, called critic and actor networks, are applied respectively to describe the unknown value function and control law. The weights of the two networks are tuned through the gradient descent method at each iteration and the optimality and convergence properties are provided in the sequel. At last, simulation results are given to illustrate the validity of the proposed algorithm.

Average consensus time-varying formation control of underactuated autonomous underwater vehicles based on event triggering decoupling mechanism

This paper focuses on the challenge of time-varying formation control for underactuated autonomous underwater vehicles (AUVs). To tackle this challenge, an average consensus time-varying formation control based on event triggering decoupling mechanism (ETDM-ACTVF) is proposed. Firstly, the average consensus time-varying formation (ACTVF) based on the affine transformation are introduced, and the control objective considering the dynamic characteristics of underactuated AUVs is proposed. The trajectory tracking controller is designed by the back stepping method. Secondly, an event triggering decoupling mechanism (ETDM) achieves the state decoupling to constrain the uncontrollable state. Moreover, a dynamic event triggering mechanism is employed to reduce the execution frequency of actuators. Finally, the excellent performance of the proposed algorithm in addressing both the underactuated AUV time-varying formation control and decoupling control problems is demonstrated through simulation results.

Risk-informed longitudinal control in autonomous vehicles: A safety potential field modeling approach

The advent of autonomous vehicles (AVs) has complicated traffic flow and raised safety concerns, there is a pressing need to enhance traffic efficiency alongside ensuring robust safety. This paper proposes a dynamic longitudinal control strategy for AVs, based on real-time accident risk assessments. Data from traffic flow, road characteristics, and weather conditions are analyzed using a random forest algorithm, producing a real-time accident risk identification model through a support vector machine. In addition, the Perceived Risk Field Model (PRFM), a car-following model calibrated using the NGSIM dataset, considers vehicle distance, velocity, relative velocity, and acceleration. These models underpin our proposed longitudinal control strategies, tailored to variable accident risk levels. Under the normal risk control strategy, the PRFM outperforms the Intelligent Driver Model (IDM) and the Cooperative Adaptive Cruise Control Model (CACC) in stability and driving comfort. For high accident risk scenarios, our approach shows improved collision avoidance and traffic oscillation mitigation. Simulations indicate that the high accident risk control strategy offers better safety and stability than the normal risk strategy. Furthermore, traffic flow diagrams suggest an increase in road capacity as AV penetration rates rise, pointing to the promising efficiency of our proposed solution. This research provides a robust framework for AV operation, ensuring optimal traffic safety and efficiency.

Recent Advances in Motion Planning and Control of Autonomous Vehicles

An autonomous vehicle operates without human intervention, marking advancements in navigating structured urban roads and unstructured environments [...]

Adaptive fixed-time path following cooperative control for autonomous surface vehicles based on path-dependent constraints

This paper introduces a fixed-time adaptive path-following cooperative control approach for autonomous surface vehicles (ASVs), aimed at improving navigation accuracy and safety in practical scenarios. The scheme ensures that ASVs accurately follow the reference path during navigation while satisfying position and heading angle constraints. To address the position constraint challenge, an adapted version of the universal barrier Lyapunov function is employed for path-dependent constraint function. Furthermore, a novel form of path-dependent constraint function is introduced, relying on path parameters instead of direct time variables. A new adaptive backstepping technique is also developed to tackle fixed-time convergence issues, mitigate the complexity surge in controller design, and account for the model’s uncertain dynamics. The study demonstrates that this control scheme enables the ASV’s position and heading errors to converge near the equilibrium point at a fixed time rate while consistently satisfying path-dependent constraints. The efficacy of the proposed scheme is further verified through simulation and experimental studies.