Trajectory prediction and planning remain key challenges for autonomous vehicles, particularly in complex and dynamic environments. Existing methods, typically based on static safety metrics like time-to-collision, fail to account for the evolving nature of risk in real-world traffic. This article proposes a novel safety-aware trajectory prediction and planning (SA-TP2) model, which introduces an adaptive driver risk field to simulate human-like risk perception and decision-making. By dynamically modeling risk as a continuous variable, SA-TP2 adjusts vehicle trajectories in real time, accounting for interactions with other agents, road conditions, and environmental uncertainties. The model integrates imitation learning, rule-based strategies, and physics-informed neural networks to ensure safe, efficient, and human-compatible behavior. A Linformer-based architecture and temporal hypergraph convolution network are introduced to optimize computational efficiency, enabling real-time operation in resource-constrained environments. Experimental results on benchmark datasets including next generation simulation (NGSIM), highway drone dataset (HighD), Macao connected autonomous driving (MoCAD), and NuScenes demonstrate that SA-TP2 achieves the state-of-the-art performance in trajectory prediction. In addition, extensive closed-loop testing on the NuPlan and CommonRoad platforms further confirms that SA-TP2 outperforms existing baselines, paving the way for safer navigation of autonomous driving systems.

This article investigates the chance-constrained control problem for uncertain systems, with a focus on the distortion of signal transmission between the controller and the actuator caused by noisy and bandwidth-limited communication channels, and its impact on the system's control performance. Initially, a binary dynamic encoding mechanism (DEM) is employed to encode the system's amplitude-continuous signal into a finite-length binary string, aiming to mitigate the communication burden. In the DEM-based control scheme, a critical issue is that the control performance is seriously affected by the bit-flip error (BFE), which inevitably occurs during the transmission of binary data through a noisy channel. To address this problem, a novel active BFE-resistant controller is proposed to effectively accomplish the desired control task by thoroughly considering the dynamic coupling effects between the BFE and the DEM. Subsequently, a chance-constraint index is jointly considered to ensure the safe operation of the uncertain system under a guaranteed probability bound. Sufficient conditions are established for the existence of the active BFE-resistant controller such that the mean-square boundedness and the chance constraint index are ensured simultaneously. Finally, the validity of the proposed algorithm is verified by a simulation study targeting the remote control problem for autonomous ground vehicles.

This paper investigates the real-time distributed state estimation problem for autonomous vehicles in unknown environments. During the process of autonomous driving, the distributed information of the multi-vehicle system, composed of the ego vehicle and its surrounding vehicles, must be perceived and comprehended by the ego vehicle. This process is known as situational awareness, during which the exact state information of surrounding vehicles might not be available. Furthermore, the perception accuracy of the ego vehicle regarding its environment is significantly reduced by unknown-but-bounded noises and unknown inputs. To address this issue, an innovative situational awareness scheme, based on the distributed set-membership estimation method, is proposed in this paper. First, the estimation error is decoupled from unknown inputs by using a designed unknown input estimator, which significantly enhances state estimation performance. Next, the impact of sparse topology in a multi-vehicle system on the solution of distributed set-membership state estimator gain matrices is analyzed, and a sparse optimization method is employed to obtain satisfactory estimator gain matrices. The parameters affecting estimation performance are optimized using the Lagrange multiplier method. It is shown that the ultimate boundedness of the estimation error is ensured under the proposed state estimation scheme. Moreover, a new dynamic collision warning scheme for autonomous vehicles is established based on the designed real-time state estimation algorithm. Finally, the effectiveness of the proposed method is validated through a simulation example. (c) 2026 Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.

In this paper, the encryption-decryption-based set-membership filtering issue is considered for a class of networked two-dimensional systems with unknown-but-bounded noises. In order to preserve data privacy and reduce communication burden, an encryption-decryption mechanism is put forward based on the exclusive or logical operation technique and the one-time pad method, under which the data is encrypted into a ciphertext with finite bits before being transmitted through the networks. The aim of the addressed problem is to develop an encryption-decryption-based set-membership filter (EDSMF) that is capable of generating an ellipsoidal set that contains the true system state and ensuring the security performance. Sufficient conditions are established for the existence of the desired ellipsoidal set, and the corresponding EDSMF gains are obtained by applying the Lagrange multiplier method. Moreover, the uniform boundedness of the ellipsoidal shape-defining matrix is thoroughly studied. Finally, the effectiveness of the proposed filter design method is demonstrated by a practical example of a heat exchanger. (c) 2024 Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.

This article is concerned with the secure state estimation problem for artificial neural networks (ANNs) subject to unknown-but-bounded noises, where sensors and the remote estimator are connected via open and bandwidth-limited communication networks. Using the encoding-decoding mechanism (EDM) and the Paillier encryption technique, a novel homomorphic encryption scheme (HES) is introduced, which aims to ensure the secure transmission of measurement information within communication networks that are constrained by bandwidth. Under this encoding-decoding-based HES, the data being transmitted can be encrypted into ciphertexts comprising finite bits. The emphasis of this research is placed on the development of a secure set-membership state estimation algorithm, which allows for the computation of estimates using encrypted data without the need for decryption, thereby ensuring data security throughout the entire estimation process. Taking into account the unknown-but-bounded noises, the underlying ANN, and the adopted HES, sufficient conditions are determined for the existence of the desired ellipsoidal set. The related secure state estimator gains are then derived by addressing optimization problems using the Lagrange multiplier method. Lastly, an example is presented to verify the effectiveness of the proposed secure state estimation approach.

In mixed autonomous driving environments, accurately predicting the future trajectories of surrounding vehicles is crucial for the safe operation of autonomous vehicles (AVs). In driving scenarios, a vehicle’s trajectory is determined by the decision-making process of human drivers. However, existing models primarily focus on the inherent statistical patterns in the data, often neglecting the critical aspect of understanding the decision-making processes of human drivers. This oversight results in models that fail to capture the true intentions of human drivers, leading to suboptimal performance in long-term trajectory prediction. To address this limitation, we introduce a Cognitive-Informed Transformer (CITF) that incorporates a cognitive concept, Perceived Safety, to interpret drivers’ decision-making mechanisms. Perceived Safety encapsulates the varying risk tolerances across drivers with different driving behaviors. Specifically, we develop a Perceived Safety-aware Module that includes a Quantitative Safety Assessment for measuring the subject risk levels within scenarios, and Driver Behavior Profiling for characterizing driver behaviors. Furthermore, we present a novel module, Leanformer, designed to capture social interactions among vehicles. CITF demonstrates significant performance improvements on three well-established datasets. In terms of long-term prediction, it surpasses existing benchmarks by 12.0% on the NGSIM, 28.2% on the HighD, and 20.8% on the MoCAD dataset. Additionally, its robustness in scenarios with limited or missing data is evident, surpassing most state-of-the-art (SOTA) baselines, and paving the way for real-world applications.

Autonomous vehicles (AVs) rely on accurate trajectory prediction of surrounding vehicles to ensure the safety of both passengers and other road users. Trajectory prediction spans both short-term and long-term horizons, each requiring distinct considerations: short-term predictions rely on accurately capturing the vehicle’s dynamics, while long-term predictions rely on accurately modeling the interaction patterns within the environment. However current approaches, either physics-based or learning-based models, always ignore these distinct considerations, making them struggle to find the optimal prediction for both short-term and long-term horizon. In this paper, we introduce the Dynamics-Enhanced Learning MOdel (DEMO), a novel approach that combines a physics-based Vehicle Dynamics Model with advanced deep learning algorithms. DEMO employs a two-stage architecture, featuring a Dynamics Learning Stage and an Interaction Learning Stage, where the former stage focuses on capturing vehicle motion dynamics and the latter focuses on modeling interaction. By capitalizing on the respective strengths of both methods, DEMO facilitates multi-horizon predictions for future trajectories. Experimental results on the Next Generation Simulation (NGSIM), Macau Connected Autonomous Driving (MoCAD), Highway Drone (HighD), and nuScenes datasets demonstrate that DEMO outperforms state-of-the-art (SOTA) baselines in both short-term and long-term prediction horizons.

A security control problem is addressed for a kind of networked Markov jump system subject to two-side stochastic denial-of-service (DoS) attacks, in which the attack occurrence situation obeys the Markov chain. To resist the attack effects, a redundant channel protocol is adopted, wherein measurement/control signals are sent to the controller/actuator via the redundant channel when the primary channel suffers from cyber-attacks. According to the above redundant channel protocol, both measurement and input models are established to represent the received signals by the controller and actuator, respectively. Then, a novel two-layer mapping strategy via logical operations is proposed to describe attack occurrences and system jumping. These facilitate the design of a security sliding mode controller, under which the exponential mean-square stability of the networked Markov jump systems subject to DoS attacks and the corresponding sufficient conditions are derived. Eventually, the simulation results via the DC chopper circuit system are provided to illustrate the proposed redundant-channel-based security control method. Note to Practitioners —This paper is motivated by the attack resistance problem in networked control systems. With the revolutionary evolution of wireless communication technology, in many practical systems, such as smart power grids, unmanned aerial vehicles (UAVs), and unmanned surface vehicles (USVs), the physical devices including sensors, actuators, controllers, and other intelligent devices are interconnected and transmitted the data through a network infrastructure. The inevitable DoS attacks may cause reliability degradation of data transmission, which in practice results in shaky grid frequency, failed UAV formation, and inaccurate USV detection. Distinguished from the previous investigation results from a passive perspective, this paper proposes a redundant-channel-based method to solve the reliability degradation problem caused by DoS attacks from a proactive perspective. The developed redundant channel method against Markov-chain-based DoS attacks can improve the reliability of data transmission, and the proposed novel two-layer mapping strategy via logical operations can increase the freedom of security controller design. A DC chopper circuit is used to verify the effectiveness of the proposed novel control method and comparison simulation results show the advantages of the redundant channel method. How to improve the reliability of vulnerable channels is a knotty problem in communication. With the groundbreaking research of this paper, against other attacks, such as deception attacks, reply attacks, and frequency-and duration-constrained DoS attacks, the attack resistance problems via proactively strengthening the reliability of data transmission should be further explored in the future.

Accurately and safely predicting the trajectories of surrounding vehicles is essential for fully realizing autonomous driving (AD). This paper presents the Human-Like Trajectory Prediction model (HLTP++), which emulates human cognitive processes to improve trajectory prediction in AD. HLTP++ incorporates a novel teacher-student knowledge distillation framework. The “teacher” model, equipped with an adaptive visual sector, mimics the dynamic allocation of attention human drivers exhibit based on factors like spatial orientation, proximity, and driving speed. On the other hand, the “student” model focuses on real-time interaction and human decision-making, drawing parallels to the human memory storage mechanism. Furthermore, we improve the model’s efficiency by introducing a new Fourier Adaptive Spike Neural Network (FA-SNN), allowing for faster and more precise predictions with fewer parameters. Evaluated using the NGSIM, HighD, and MoCAD benchmarks, HLTP++ demonstrates superior performance compared to existing models, which reduces the predicted trajectory error with over 11% on the NGSIM dataset and 25% on the HighD datasets. Moreover, HLTP++ demonstrates strong adaptability in challenging environments with incomplete input data. This marks a significant stride in the journey towards fully AD systems.

This article is concerned with the secure collision-free tracking control problem for autonomous vehicles with uncertainties, where system signals are transmitted through constrained communication networks. In such open and uncertain environments, the control performance of vehicles is seriously affected by privacy leakage, non-Gaussian noise, and obstacles. The aim of this research is to propose a tracking control scheme that ensures security, mean-square boundedness, and collision-free performance concurrently. Initially, to safeguard the privacy of the transmitted data and to enable secure tracking control, a dynamic encoding-based ElGamal encryption mechanism is introduced, which is further embedded in the design of the observer-based tracking controller. Subsequently, a collision-free chance-constrained index is proposed for achieving real-time obstacle avoidance by comprehensively considering the influence of stochastic noises. A thorough analysis is conducted to examine the impact of non-Gaussian noise and unmeasurable states on the performance of collision-free tracking control. Sufficient conditions are derived to guarantee the desired performance, and the corresponding control inputs are obtained by solving certain optimization problems subject to chance constraints. Finally, an illustrative example is provided to validate the effectiveness of the proposed secure collision-free tracking controller.

This paper is concerned with the robust set-membership fusion estimation problem for multi-sensor systems under the influence of uncertain parameters and unknown-but-bounded noises. The sensor nodes and the remote fusion estimator are connected through bandwidth-constrained communication networks. To address this challenge, a dynamic encoding-based data compression mechanism is introduced, enabling the data to be efficiently encoded using a reduced number of bits. By thoroughly analyzing the impact of uncertain parameters, unknown-but-bounded noises, and data distortions on estimation performance, sufficient conditions are derived to ensure that the estimation error remains within the bounded ellipsoidal set. Moreover, the weight parameters are calculated by solving constrained optimization problems. Finally, the effectiveness of the proposed estimation algorithm is verified through a maneuvering target tracking scenario.

This paper is concerned with the containment control problem for multi-unmanned aerial vehicle (multi-UAV) systems with data transmission among agents during a limited-bandwidth network. A good deal of data transmitted during a limited-bandwidth network may result in data collision, which brings negative effects on controller design and containment tracking implementation. To reduce the network load during data transmission, a consensus condition and a triggering subsidiary condition are first proposed to realize the containment tracking with H∞ performance by using the triggered states. Then a self-triggered scheme is established under a multi-agent framework, in which the next triggering instant is calculated beforehand by using the past states. Combined with the self-triggered scheme, a containment controller is obtained via solving the derived consensus condition, which ensures that follower UAVs are driven into a convex set surrounded by leader UAVs. Finally, the validity of the self-triggered containment control method is proved through a simulation experience.

This paper is concerned with the security sliding mode control problem for Markov switching systems suffering from both physical and communication constraints, that is, the actuator saturation and injection attack. With the help of the learning capability of neural networks, the nonlinear upper bound of injection attacks is analyzed and estimated, and the security control strategy based on neural networks is proposed to improve the reliability of control. Considering the transient performance implementation, a time-dependent sliding mode controller is proposed to guarantee the reachability of the specified sliding surface within a given finite time, in which the adaptive rules of neural network parameters are designed to reduce the effects of the considered attack. The arrival and sliding stages are, respectively, analyzed and sufficient conditions are achieved to guarantee the random finite-time boundedness of the closed-loop system. Eventually, a simulation example is given, combined with a genetic algorithm, to illustrate the validity of the proposed adaptive security control method.

This article investigates the sliding mode control problem based on dynamic encoding-decoding over lossy networks with Round-Robin (RR) protocol. Firstly, considering non-encoded transmission, the sliding mode controller gain and the sliding mode region are acquired over in case of lossy networks under RR protocol. Then, considering encoded transmission, state-dependent online adjustment rules are introduced, and a dynamic-quantization-based encoding-decoding scheme is designed. Pre-designed sliding mode controller is proven to be feasible even with limited encoding length. However, due to inevitable encoding errors, the convergence domain and sliding mode region of the closed-loop system both increase. Finally, the effectiveness of this method is verified by a numerical example.

This article addresses the privacy-preserving control issue for two-dimensional systems with probabilistic constraints. According to the exclusive or logical operation and the dynamic coding-decoding rule, a privacy-preserving mechanism (PPM) is developed, under which the transmitted data is efficiently compressed and encrypted into a ciphertext with finite bits. A PPM-based controller is designed that simultaneously guarantees a prescribed probabilistic constraint, mean-square boundedness, and privacy performance. Mathematical techniques, including mathematical induction, Chebyshev inequality, and matrix analysis, are employed to establish sufficient conditions for the presence of the desired controller gains. Additionally, the privacy and secrecy performance of the PPM is analyzed and simulation examples are presented to showcase the efficacy of the proposed controller design method.

In this article, the event-triggered cost-guaranteed control problem is investigated for a class of shift-varying linear repetitive processes (LRPs) with multiplicative noises under probabilistic constraints. First, in order to relieve the communication burden and improve the communication efficiency, the event-triggered mechanism is exploited for the shift-varying LRPs over the communication network of limited bandwidth. By constructing a novel event generator function, the order of the event triggering sequence is identified on two independent directions. Then, the probabilistic constraint is enforced onto the shift-varying LRPs under the influence of the event-triggered mechanism. With the aid of the Chi-square distribution and the completing-the-square technique, a controller design algorithm is proposed to ensure that both the probabilistic constraints and the quadratic cost index are satisfied by using the actual information at the triggering instants. Sufficient conditions are derived to guarantee the existence of event-based controllers, and the controller gains are obtained at each step by solving a series of matrix inequalities. Finally, an illustrative example is given to demonstrate the effectiveness of the proposed controller design method.

In this paper, the neural-network-based filtering problem is investigated for multi-sensor systems with dynamic encoding mechanisms. The sensor nodes and the remote filter are connected through bandwidth-constrained communication networks. To alleviate the communication burden, a novel dynamic encoding-based data compression-decompression mechanism is proposed so as to encode the data into a limited number of bits. Then, with the aid of the neural network learning method, a neural-network-based set-membership filter is developed for estimating the system states. Sufficient conditions are obtained to ensure that the filtering error remains within the bounded ellipsoidal set. In addition, the neural network tuning parameters and the filter gains are calculated by solving constrained optimization problems. Finally, the effectiveness of the proposed filtering algorithm is verified through a scenario of maneuvering target tracking using multiple unmanned aerial vehicles.

In this article, the adaptive neural-network-based (NN-based) set-membership state estimation problem is studied for a class of nonlinear systems subject to bit rate constraints and unknown-but-bounded noises. The measurement output signals are transmitted from sensors to a remote estimator via a bit rate constrained communication channel. To relieve the communication burden and ameliorate the state estimation accuracy, a bit rate allocation mechanism is put forward for the sensor nodes by solving a constrained optimization problem. Subsequently, through the NN learning method, an NN-based set-membership estimator is designed to determine an ellipsoidal set that contains the system state, where the proposed estimator relies upon a prediction-correction structure. With the help of the mathematical induction technique and the set theory, sufficient conditions are obtained to ensure the existence of both the adaptive tuning parameters and the set-membership estimators, and then, the corresponding parameters and estimator gains are calculated by solving a set of optimization problems. In addition, the monotonicity of the upper bound on the squared estimation error with respect to the bit rate and the convergence of the NN weight are analyzed, respectively. Finally, an illustrative example is given to demonstrate the effectiveness of the proposed state estimation algorithm.

In this article, the simultaneous state and fault estimation problem is investigated for a class of nonlinear 2-D shift-varying systems, where the sensors and the estimator are connected via a communication network of limited bandwidth. With the purpose of relieving the communication burden and enhancing the transmission security, a new encoding-decoding mechanism is put forward so as to encode the transmitted data with a finite number of bits. The aim of the addressed problem is to develop a neural-network (NN)-based set-membership estimator for jointly estimating the system states and the faults, where the estimation errors are guaranteed to reside within an optimized ellipsoidal set. With the aid of the mathematical induction technique and certain convex optimization approaches, sufficient conditions are derived for the existence of the desired set-membership estimator, and the estimator gains and the NN tuning scalars are then presented in terms of the solutions to a set of optimization problems subject to ellipsoidal constraints. Finally, an illustrative example is given to demonstrate the effectiveness of the proposed estimator design method.

In this paper, the set-membership filtering problem is investigated for a class of two-dimensional (2-D) shift-varying systems described by the general Fornasini-Marchesini second model with a dynamic event-triggered mechanism. The sensors are communicated with the remote filter through a communication network of limited bandwidth . In order to improve the efficiency of utilizing the network resources, a dynamic event-triggered mechanism is exploited for the 2-D shift-varying system, and a set-membership filter is then designed for 2-D systems subject to unknown-but-bounded noises so that the actual system state is guaranteed to reside within an optimized ellipsoidal set. By means of the mathematical induction technique and the set theory, sufficient conditions are derived to ensure the existence of the desired set-membership filter. Furthermore, the filter gains are obtained by solving a set of optimization problems subject to ellipsoidal constraints. Finally, an illustrative example is given to demonstrate the effectiveness of the proposed filter design method.