The goal of infrared and visible image fusion (TVIF) is to integrate the unique advantages of both modalities to achieve a more comprehensive understanding of a scene. However, existing methods struggle to effectively handle modal disparities, resulting in visual degradation of the details and prominent targets of the fused images. To address these challenges, we introduce PromptFusion, a prompt-based approach that harmoniously combines multi-modality images under the guidance of semantic prompts. Firstly, to better characterize the features of different modalities, a contourlet autoencoder is designed to separate and extract the high-/low-frequency components of different modalities, thereby improving the extraction of fine details and textures. We also introduce a prompt learning mechanism using positive and negative prompts, leveraging Vision-Language Models to improve the fusion model's understanding and identification of targets in multi-modality images, leading to improved performance in downstream tasks. Furthermore, we employ bi-level asymptotic convergence optimization. This approach simplifies the intricate non-singleton non-convex bi-level problem into a series of convergent and differentiable single optimization problems that can be effectively resolved through gradient descent. Our approach advances the state-of-the-art, delivering superior fusion quality and boosting the performance of related downstream tasks. Project page: https://github.com/hey-it-s-me/Prompt-Fusion.

As the application of unmanned aerial vehicle (UAV) become increasingly widespread in various industries, its positioning in Global Navigation Satellite System (GNSS) denied environments plays an indispensable role in certain scenarios such as dense woods and enclosed underground environment. However, there are several existing defects of the conventional positioning method for UAV in GNSS-denied environments, such as the error accumulation and poor long-term accuracy in Inertial Navigation Systems and requirement for sufficient light and high computing power in vision-based localization. Therefore, a novel positioning method for UAV in GNSS-denied environments based on mechanical antenna (MA) is proposed in this work, which consists of MA installed on UAV to generated low-frequency (LF) magnetic signal, the three-dimensional magnetic field sensor in ground base station to receive signal, and the corresponding positioning algorithm based on particle swarm optimization. EM signals in LF bands is applied in positioning, therefore, this method has high propagation stability and anti-interference due to the characteristics of LF bands. Furthermore, because MA technology can greatly reduce the size and power consumption of the LF transmitting system, the LF signal used for positioning can be generated by a portable MA installed on UAV. Theoretical analysis and positioning experiments based on this method are carried out in detail. According to the results, the positioning method proposed in this work is of great feasibility with a mean error < 0.45 m in measurement, which will provide an alternative instrumentation for the positioning of UAV in GNSS-denied environments and can be used in a variety of industrial scenarios with complex electromagnetic environments in the future.

Active suspension systems (ASSs) are crucial for realizing the improvement demand of ride comfort in autonomous vehicles (AVs). This paper presents a novel integrated robust preview control (PC) algorithm to further improve the ride comfort by effectively handling actuator delay. An augmented ASS model is first established to consider actuator delay. Then, to obtain the time-domain feedforward disturbance, updating the local spatial road dataset from sensing systems by low-pass filtering and introducing speed for conversion. A predictive control framework that utilizes preview road disturbances is developed to achieve better ride comfort. And assuming parameterized disturbances, it is transformed into a consistent quadratic programming (QP) problem for real-time application in ASS. Furthermore, an innovative design of combined observer provides accurate and robust observations under multiple uncertainties, with which the algorithm can effectively compensate for actuator delay under different road conditions. Finally, the bounded-input bounded-output (BIBO) stability of the algorithm is well demonstrated. Simulation and bench test results indicate that the proposed algorithm can not only significantly improve road holding and ride comfort, but also realize a robust observation of the augmented ASS state. In addition, the solving time of the bench test satisfies the real-time requirements of the algorithm.

This paper deals with robust cooperative control of connected and automated vehicles (CAVs) with special consideration of the coupled safety inter-vehicle distance constraints. To address the hard constraints caused by the vehicle physical limitations and the coupled safety constraints among adjacent vehicles explicitly, the distributed model predictive control (DMPC) problem for CAVs is formulated. By virtue of the Lagrangian multiplier method and the dual decomposition technique, the DMPC problem subject to coupled constraints is recast into a distributed dual variable consensus optimization problem comprised of a series of subsystems with local copies of the dual variables, which is then cooperatively solved in parallel based on the alternating direction method of multiplier (ADMM) under the distributed information exchange mechanism. In addition, the disturbance is dynamically mitigated by the feedback control, and its influence on the vehicle status is quantified by the associated robust positively invariant set. Robust cooperative control protocols are proposed to regulate the cooperative operations of CAVs, and the recursive feasibility and the closed-loop stability are theoretically proved. Numerical experiments are carried out to illustrate the effectiveness of the proposed method.

Numerous missions in both civil and military fields involve the pursuit-evasion problem of vehicles. With vertical take-off and landing capability, the intelligent air–ground vehicle expands its feasible path to 3-D space, which has great advantages in the pursuit. This vehicle requires adequate path planning to obtain an optimal 3-D path and further improve the pursuit efficiency. The planning process of the air–ground vehicle currently faces the technical difficulties of acquiring the proper takeoff timing and position while optimizing the planning trajectory, especially in a complex environment with dense obstacles. To solve the above issues, a game-learning-based smooth path planning strategy for the intelligent air–ground vehicle considering mode switching is proposed in this article. First, a new reward function of the $Q$ -learning algorithm, considering the influence of flight obstacle crossing parameters, is presented to explore the short forward track distance. Second, in the update rule, the pursuit-evasion game acts in the mode switching decisions. During interactive learning between the vehicle and environment, this game constantly updates the Nash equilibrium solutions for mode switching and gets a series of switching decisions of the pursuer vehicle (ego vehicle). Third, a double-yaw correction for path smoothing modification is proposed to reduce turning points and avoid local path deviations. This modification provides heuristic information for the exploration of the environment, which significantly speeds up the convergence speed of the algorithm. Finally, the proposed strategy is verified on a 1000 m*1000 m map with 0–200 m obstacle height. Results show that this strategy is effective to decrease the 253-m distance compared with the traditional reinforcement learning algorithm and has a faster convergence speed. The number of trajectory direction changes is 36% less than that of the game-learning algorithm only considering mode switching. The unreasonable large angle turns are eliminated.

Energy management strategy (EMS) plays a vital role in improving the fuel economy of plug-in hybrid electric vehicle (PHEV). By virtue of excellent real-time performance, deterministic rule-based (DRB) method is widely introduced into EMS for the control of actual PHEV. However, fixed parameters are usually used as thresholds in traditional DBR control, which makes it difficult for PHEV to achieve excellent fuel economy. To solve this problem, relevant parameters need to be optimized, but the resulting time-consuming and complex process is an obstacle for practical application of this scheme. Nowadays, the emergence of technologies, such as wireless communication, remote monitoring and so on, has gave birth to the concept of cyber-physical system (CPS). It provides an opportunity to optimize parameters of DRB EMS. Motivated by this, this paper proposes a cyber physical energy optimization control design for PHEVs. Among them, DRB control is designed to allocate power tasks for the engine and electric motor (EM), according to state of the charge (SOC) of battery and demand power of vehicle. Moreover, to further improve performance of EMS, an enhanced firework algorithm (EFWA) is firstly proposed to optimize parameters of controller. Compared with original algorithm, a novel selection mechanism for non-CF is introduced in EFWA. It makes the excellent parameters could be obtained in a shorter time, which is more suitable for the complex optimization of EMS. Finally, the effectiveness of proposed EMS is verified and evaluated. The results show that it improves the fuel economy of PHEV by 10% and 12% over that using the unoptimized rule-based EMS, under the China typical urban driving cycle and real-world driving cycle, respectively.