Article
Automation & Control Systems
Marvin Jung, Paulo Renato da Costa Mendes, Magnus oennheim, Emil Gustavsson
Summary: The prediction model plays a vital role in MPC strategies as its accuracy directly impacts the quality of predictions and control performance. In cases where a model based on physical equations is not available or difficult to obtain all parameters, using black-box models within the MPC framework is an attractive alternative, as they only require input and output data. This paper discusses questions such as the feasibility of using LSTM as predictors, implementation methods, computation of derivatives, recommended solvers and tools, and ensuring real-time capability.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Engineering, Aerospace
Jacopo Guadagnini, Michele Lavagna, Paulo Rosa
Summary: This study aims to demonstrate the benefits and limitations of on-board guidance for reusable launch vehicles and explores different model predictive control-based guidance and control architectures. The focus is on implementing a sequential convexification model predictive control guidance algorithm to solve the powered descent guidance problem. The robustness and feasibility of real-time implementation are evaluated using an industrial simulation framework.
Article
Automation & Control Systems
Yu Yang, Hongze Xu, Xiuming Yao
Summary: A novel robust hierarchical multi-loop composite control scheme is proposed, which reduces the computational effort and stabilizes decoupled subsystems for trajectory tracking control of robotic manipulators subject to constraints and disturbances. The control scheme includes an inner loop based on inverse dynamics control and an external loop relying on disturbance-observer-based tube model predictive composite controllers. The effectiveness of the proposed scheme is demonstrated through simulation experiments on a PUMA 560 robotic manipulator.
JOURNAL OF THE FRANKLIN INSTITUTE-ENGINEERING AND APPLIED MATHEMATICS
(2023)
Article
Automation & Control Systems
Dae Jung Kim, Yong Woo Jeong, Chung Choo Chung
Summary: This article proposes a lateral vehicle trajectory planning and control algorithm using a model predictive control (MPC) scheme for an automated perpendicular parking system. The proposed method addresses the issues of state-dependent planning and undesirable steering maneuvers, and it is shown to achieve smooth and stable lateral vehicle motion even in tight parking space conditions.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2023)
Article
Engineering, Marine
Shengzheng Wang, Zhaoyang Sun, Qiumeng Yuan, Zhen Sun, Zhizheng Wu, Tsung-Hsuan Hsieh
Summary: This paper proposes a ship handling and motion model that considers disturbances of wind and currents, a trajectory prediction model based on LSTM neural network, and a nonlinear MPC model to control trajectory tracking, improving the ship's anti-disturbance capability and accuracy. Experiments demonstrate the outstanding performance and effectiveness of the proposed approach under hybrid disturbances of wind and currents.
Article
Engineering, Mechanical
Tiao Kang, Hui Peng, Xiaoyan Peng
Summary: In this paper, data-driven modeling techniques and deep learning methods are used to accurately capture the spatiotemporal features of a smooth nonlinear system. The LSTM-CNN-ARX model is developed by incorporating the strengths of LSTM for learning temporal characteristics and CNN for capturing spatial characteristics. Control comparison experiments on a water tank system show the effectiveness of the developed models and MPC methods.
Article
Engineering, Marine
Zhichao Hu, Junmin Mou, Linying Chen, Xuefei Jia, Pengfei Chen
Summary: In this paper, a novel smart tug escorting mode is proposed to achieve energy conservation while ensuring the safe navigation of the serviced ship. A green and smart framework is presented for energy-saving escorting. An escorting model is established to describe the dynamics of the tug and the relation between the tug and the serviced ship. Reference trajectories are planned collaboratively and the Model Predictive Control (MPC) scheme is adopted for tug trajectory tracking. Simulation results show the energy conservation potential of the proposed framework.
Article
Automation & Control Systems
Linhe Ge, Yang Zhao, Shouren Zhong, Zitong Shan, Konghui Guo
Summary: This paper proposes a split integration method to address the integration stability problem of autonomous driving motion control at low speeds, and significantly improve the efficiency of NMPC.
CONTROL ENGINEERING PRACTICE
(2023)
Article
Automation & Control Systems
Kaihui Wang, Wei Zou, Ruichen Ma, Yu Wang, Hu Su
Summary: This article presents a study on trajectory tracking control for an underactuated bionic underwater vehicle that is propelled by undulating fins. A novel dynamics model-based predictive control strategy is proposed to solve the challenging trajectory tracking control problem. Real-time precise tracking control is achieved with the integration of a practical sidesway compensator. Extensive simulations and real-world experiments demonstrate the reliability and superiority of the proposed approach.
IEEE-ASME TRANSACTIONS ON MECHATRONICS
(2023)
Article
Engineering, Industrial
Bertrand Galy, Laurent Giraud
Summary: This article provides a statistical review of mine hoists in use in Quebec, Canada and discusses the safety regulations and machinery safety standards applicable to mine hoists. It also presents a fault tree analysis and risk analysis for a generic drum hoist. The main conclusion is the need for a SIL 3 compliant command and supervision system to cover all hazardous situations in the operation of a mine hoist.
Article
Engineering, Civil
Zhen-Ni Li, Xing-Hui Huang, Tong Mu, Jiao Wang
Summary: An attention-based LSTM model is proposed for lane change behavior prediction in highways, showing higher accuracy and better interpretability than other models. A crash risk prediction model based on Time-To-Collision is also proposed, further verifying the effectiveness of the method.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Engineering, Civil
Xinyou Lin, Yunliang Tang, Binhao Zhou
Summary: An improved model predictive control (MPC) based path tracking strategy is proposed to enhance tracking accuracy and adaptability in different velocities, utilizing cosine similarity for online updating and fuzzy control for horizon factor determination. The weighting factors of prediction and control horizons are discussed for improved adaptability. The strategy shows good adaptability and satisfactory tracking accuracy in various velocities, validated through a real prototype vehicle test.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Yang Liang, Zhishuai Yin, Linzhen Nie, Yuanxin Ba
Summary: This paper proposes a shared controller for safe and human-friendly cooperative driving based on predictive risk assessment enabled by Digital Twin technologies. A fine-grained digital replica of the driving scene is created in the digital world, and spatial-temporal interactive features are used to predict future trajectories of neighboring vehicles. These trajectories are integrated into the risk distribution to construct predictive risk fields, and a novel shared controller is designed to minimize the driving risk while honoring the driver's commands in a smooth and minimal-intervention manner.
IEEE TRANSACTIONS ON INTELLIGENT VEHICLES
(2023)
Article
Engineering, Aerospace
Da Huo, Li Dai, Runqi Chai, Ruochen Xue, Yuanqing Xia
Summary: In this article, a collision-free model predictive trajectory tracking control algorithm for UAVs in environments with both static and dynamic obstacles is proposed. Collision avoidance is ensured by obtaining outer polyhedral approximations of each interval of the dynamic obstacles trajectories and optimizing a plane to separate the polyhedra and the UAV's trajectory. The algorithm combines computationally efficient collision-free constraints and physical constraints to formulate a model predictive control optimization problem with a tailored terminal constraint set, which can be solved by a standard nonlinear programming solver. Control theoretic properties, including recursive feasibility, collision avoidance guarantee, and closed-loop stability, are established. The efficacy of the proposed algorithm is successfully evaluated through simulation in a multiobstacle environment.
IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS
(2023)
Article
Computer Science, Information Systems
Sibang Liu, Kuili Liu, Zhen Zhong, Jinghan Yi, Hamdulah Aliev
Summary: This paper proposes a new path tracking strategy for controlling various types of robots in known and unknown environments. By using model prediction control and gain-scheduled control law, the non-holonomic mobile robot is able to overcome kinematic disturbances. The simulation results demonstrate the feasibility of this strategy.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Jingda Wu, Zhiyu Huang, Chen Lv
Summary: In this paper, a novel uncertainty-aware model-based RL method is proposed to improve the learning efficiency and performance in autonomous driving scenarios. By establishing an action-conditioned ensemble model with uncertainty assessment capability, and developing an uncertainty-aware model-based RL method based on adaptive truncation approach, virtual interactions between the agent and environment model are provided to enhance RL's learning efficiency and performance. Validation results demonstrate that the proposed method outperforms the model-free RL approach in terms of learning efficiency, and the model-based approach in terms of both efficiency and performance, indicating its feasibility and effectiveness.
IEEE TRANSACTIONS ON INTELLIGENT VEHICLES
(2023)
Article
Computer Science, Artificial Intelligence
Xiaoyu Mo, Chen Lv
Summary: Recent advances in DTPI enable high-fidelity virtual representation of the physical world for intelligent prediction and decision-making in autonomous vehicles and intelligent transportation systems. This study investigates trajectory-prediction-enabled motion planning using deep neural networks and explores the impacts of historical states and future motions on planning performance.
IEEE TRANSACTIONS ON INTELLIGENT VEHICLES
(2023)
Letter
Automation & Control Systems
Xiangkun He, Chen Lv
IEEE-CAA JOURNAL OF AUTOMATICA SINICA
(2023)
Article
Robotics
Xiaoyu Mo, Yang Xing, Haochen Liu, Chen Lv
Summary: Predicting the future motions of neighboring agents is crucial for autonomous vehicles to navigate complex scenarios. Our proposed map-adaptive predictor can predict a variable number of future trajectories based on the number of lanes with candidate centerlines (CCLs). It incorporates three types of predictions, including single CCL-guided future motions, scene-reasoning prediction, and motion-maintaining prediction, through a single graph operation. By utilizing a hierarchical graph representation of the driving scene, our method achieves map-adaptive prediction and outperforms strong baselines in experiments on real-world driving datasets.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2023)
Article
Computer Science, Artificial Intelligence
Long Chen, Yuchen Li, Chao Huang, Bai Li, Yang Xing, Daxin Tian, Li Li, Zhongxu Hu, Xiaoxiang Na, Zixuan Li, Siyu Teng, Chen Lv, Jinjun Wang, Dongpu Cao, Nanning Zheng, Fei-Yue Wang
Summary: Interest in autonomous driving and intelligent vehicles is growing rapidly due to their convenience, safety, and economic benefits. However, existing surveys are limited in scope and lack systematic summaries and future research directions.
IEEE TRANSACTIONS ON INTELLIGENT VEHICLES
(2023)
Article
Engineering, Civil
Wenbo Shao, Yanchao Xu, Jun Li, Chen Lv, Weida Wang, Hong Wang
Summary: This study aims to address the uncertainty and lack of explainability in autonomous driving by exploring the impact of traffic environment on prediction algorithms. The study proposes a trajectory prediction framework with epistemic uncertainty estimation ability to output high uncertainty when facing unforeseeable or unknown scenarios. The results indicate that deep ensemble-based methods have advantages in improving robustness while estimating epistemic uncertainty.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Engineering, Civil
Zhiyu Huang, Haochen Liu, Jingda Wu, Chen Lv
Summary: Making safe and human-like decisions in autonomous driving systems is crucial, and this study proposes a predictive behavior planning framework that learns from human driving data. The framework includes a behavior generation module, a conditional motion prediction network, and a scoring module using inverse reinforcement learning. Comprehensive experiments on real-world urban driving dataset validate the effectiveness of this framework in predicting future trajectories and selecting human-like driving plans.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Xiangkun He, Zhongxu Hu, Haohan Yang, Chen Lv
Summary: In this paper, a novel constrained multi-objective reinforcement learning algorithm is proposed for personalized end-to-end robotic control with continuous actions. The approach trains a single model using constraint design and a comprehensive index to achieve optimal policies based on user-specified preferences.
Article
Engineering, Electrical & Electronic
Lin Yang, Mohammad Zaidi Ariffin, Baichuan Lou, Chen Lv, Domenico Campolo
Summary: This study proposes a motion planning framework with bilevel optimization for robotic contact-rich insertion tasks. It integrates Dynamic Movement Primitives (DMPs) to parameterize motion trajectories, Black-Box Optimization (BBO) to improve contact-rich insertion policy with hydroelastic contact model, and simulated variability to account for visual uncertainty in the real world.
Review
Transportation Science & Technology
Yan Wang, Henglai Wei, Lie Yang, Binbin Hu, Chen Lv
Summary: Precise vehicle state and surrounding traffic information are crucial for decision-making and dynamic control of intelligent connected vehicles. This study investigates the recent research progress in state estimation techniques, focusing on the concept of a vehicle neighborhood system that describes the state of multiple traffic elements. The review presents various state estimation methods and discusses their strengths, drawbacks, and future research directions.
SAE INTERNATIONAL JOURNAL OF VEHICLE DYNAMICS STABILITY AND NVH
(2023)
Article
Engineering, Civil
Tianci Yang, Carlos Murguia, Chen Lv
Summary: Cooperative Adaptive Cruise Control (CACC) is a technology that allows vehicles on the highway to form tightly-coupled platoons by exchanging inter-vehicle data through wireless communication networks. It improves traffic throughput and safety, while reducing energy consumption. However, the increased vehicle connectivity brings new security challenges, as adversaries can exploit the network to disrupt platooning performance or cause collisions. This manuscript proposes a novel anomaly detection scheme using real-time sensor/network data and mathematical models, but it acknowledges the limitations due to modeling uncertainties, network effects, and noise.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Zhiyu Huang, Haochen Liu, Jingda Wu, Chen Lv
Summary: To achieve safe and socially compliant autonomous driving, this study proposes a differentiable integrated prediction and planning (DIPP) framework that addresses the issues of separating prediction and planning modules and specifying the cost function. It utilizes a differentiable nonlinear optimizer to optimize the trajectory for the AV based on predicted trajectories of surrounding agents, making all operations differentiable. The framework is trained on real-world driving data and outperforms baseline methods in both open-loop and closed-loop tests.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Engineering, Civil
Xiaoyu Mo, Haochen Liu, Zhiyu Huang, Xiuxian Li, Chen Lv
Summary: This work proposes a novel map-adaptive multimodal trajectory predictor that predicts future traffic behavior in complex driving environments. The predictor is derived by training an intention-aware unimodal trajectory predictor and linking driving modalities, driver's intentions, and a vehicle's candidate centerlines. The proposed predictor offers a faster and more cost-effective alternative compared to traditional multimodal predictors and has demonstrated comparable or superior performance in specific applications.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Engineering, Civil
Wenhui Huang, Yanxin Zhou, Xiangkun He, Chen Lv
Summary: This study proposes a novel Goal-guided Transformer-enabled reinforcement learning approach that utilizes physical goal states as input to improve data efficiency and achieve efficient autonomous navigation. The results demonstrate that this method significantly enhances data efficiency and outperforms other state-of-the-art baselines in terms of multiple metrics.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Engineering, Electrical & Electronic
Xiangkun He, Chen Lv
Summary: This paper presents an observation-robust reinforcement learning approach to address perception uncertainties in autonomous driving for safe decision making. It trains an adversarial agent to generate optimal attacks on observations and develops an observation-robust actor-critic approach to ensure the changes in policies remain within a certain bound.
AUTOMOTIVE INNOVATION
(2023)