Network-level traffic signal control is an effective way to increase throughput and reduce congestion. The max-pressure algorithm, known for maximizing network throughput, has been widely studied. However, it requires accurate queue length and turn ratio measurements, and its theoretical guarantee is limited to feasible demand (i.e., demand within the capacity region) under the assumption of infinite queue capacity. To overcome these limitations, this study proposes a distributed joint admission and signal control algorithm for finite-capacity networks with both connected and regular vehicles. By using feedback from connected vehicles, the algorithm estimates queue lengths and turn ratios, reducing reliance on precise measurements. It also adaptively adjusts input flow rates to prevent oversaturation and ensure demand feasibility, even under high-demand conditions, while optimizing signal phases to ensure analytic performance. Using a Lyapunov drift optimization approach, we analytically prove a$[O(1/V), O(V)]$tradeoff between throughput and delay and establish degradation bounds that quantify the impact of queue length estimation errors on network performance. Simulations in a network with 256 origin-destination pairs show up to a 16.3% increase in throughput and reduced delays, especially in high-demand settings. The method also demonstrates strong resilience to sudden demand changes and incidents, ensuring quick recovery.

Using the backstepping technique, this paper formulates innovative adaptive finite-time stabilizing controllers for uncertain nonlinear systems featuring nonuniform input quantization and asymmetric, time-varying output constraints. These novel controllers leverage the consistent characteristics of both hysteresis quantizers and logarithmic quantizers. Quantization errors, when consistent, become unbounded and contingent on control input, rendering them incompatible with the growth conditions of nonlinear systems. Consequently, the developed adaptive controllers eliminate the reliance on growth conditions, effectively addressing the impact of unbounded quantization errors on finite-time stability. This adaptability allows the controllers to function effectively with systems employing either hysteresis quantizers or logarithmic quantizers. The paper establishes the convergence of these controllers through the finite-time Lyapunov stability theorem. It also provides a comprehensive guideline for tuning settling time, enabling fine-grained control over finite-time convergence and adjustable tracking error performance. Additionally, the controllers rigorously maintain system output within predefined limits. Their effectiveness and low computational burden are demonstrated through three comparative numerical simulations and a practical simulation in collision-free trajectory tracking control of an autonomous vehicle platoon using the vehicle motion software CarSim. These simulations confirm the advanced performance of the adaptive controllers.Note to Practitioners—This paper introduces an innovative approach to control uncertain nonlinear systems encountering intricate input quantization and output constraints. Employing the sophisticated backstepping technique, the authors present adaptive finite-time-stabilizing controllers engineered to address nonuniform input quantization and asymmetric, time-varying output restrictions. What distinguishes these controllers is their reliance on the consistent behavior exhibited by hysteresis and logarithmic quantizers. This unique feature equips them to effectively counteract unbounded quantization errors influenced by control input. Most notably, these controllers eliminate the conventional growth conditions typically demanded by nonlinear systems. As a result, they extend their applicability to a broad spectrum of systems employing either hysteresis or logarithmic quantizers. The research also provides practitioners with a valuable guideline for precisely adjusting settling time. This enables the attainment of desired convergence rates while permitting adaptable tracking error performance. Additionally, these controllers guarantee that the system’s output adheres to predefined limits. The practical significance of this study is highlighted through three comparative numerical simulations and a real-world application simulation. This real-world simulation involves collision-free trajectory tracking control of an autonomous vehicle platoon, executed using the vehicle motion software CarSim. These simulations unequivocally demonstrate the effectiveness and low computational burden of the developed controllers, thereby establishing them as a valuable resource for practitioners facing complex control challenges in various domains.

With the rapid development of V2V and V2I communication technologies and autonomous control systems, autonomous vehicles (AVs) are gaining increasing popularity. Small-spacing AV platoons offer advantages such as enhanced road capacity and energy efficiency. However, in non-ideal communication environments, packet loss can cause partial loss of trajectory information, resulting in discontinuous tracking. This may induce significant transients and trigger actuator saturation, aggravating traffic disturbances. In bidirectional platoons, where control signals propagate in both directions, the impact of such disruptions is further amplified due to mutual vehicle interdependence. This paper addresses these challenges by considering asymmetric actuator saturation, discontinuous tracking trajectories, and non-zero initial spacing errors in bidirectional AV platoons. A continuous control law is designed based on coupled sliding mode control, and Lyapunov stability theory is employed to ensure both trajectory tracking stability and string stability. Our contributions include the development of a modified spacing policy that not only eliminates large transients and string instability caused by non-zero initial spacing errors but also ensures rapid convergence to the desired spacing within a finite and adjustable time frame. Furthermore, a variant sigmoid function is introduced to actively smooth the discontinuous tracking trajectories, thereby reducing communication demands and suppressing transients. An auxiliary system is also designed to manage actuator saturation effectively, ensuring provable stability and fully leveraging actuator capabilities. Results demonstrate that the control strategy achieves both trajectory tracking stability and string stability, while also enabling rapid tracking performance and maintaining small spacing errors by making full use of actuator potential.

Urban public transportation is a complex and open system integral to urban mobility. Its operation is often disrupted by various random factors, necessitating robust scheduling solutions. This study develops a bus robust scheduling model based on mixed-integer linear programming to enhance system resilience. First, an arterial signal coordination model is proposed for mixed traffic environments, enabling autonomous public transport vehicles to traverse intersections without stopping. Second, a demand-deterministic bus scheduling model is constructed, integrating timetables, trajectories, and origin-destination transfer schemes to balance passenger waiting time fairness and efficiency. Third, to address stochastic passenger demand during actual operations, a robust bus scheduling model is developed by incorporating robust constraints. Numerical experiments demonstrate that the demand-deterministic model generates optimal scheduling schemes when passenger demand remains within bus capacity. However, when passenger demand exceeds capacity, the demand-deterministic model becomes infeasible. In such scenarios, the robust scheduling model produces feasible schemes, albeit with reduced optimization, and its robustness can be tuned by adjusting model parameters. Additionally, practical management insights are provided for real-world applications.

Human-like visual perception systems are indispensable and vital components of human-like autonomous vehicles. In the real driving environment, there is much unlabeled information and the total number of categories of information is uncertain. While human brains are adept at processing such information, current methods are not. Thus, this study presented a novel hybrid unsupervised deep learning method to model the information processing mechanism of the driver’s visual perception. The proposed approach (CAE-SOM) was a neuron-triggered method, which leveraged the virtues of a convolutional autoencoder (CAE) and a self-organizing map (SOM) neural network. The CAE mimicked the hierarchical structures of the driver’s visual system to extract the high-level features, whilst the SOM neural network simulated the working principle of human brain neurons during the information judgment process to perform unsupervised clustering. The CAE-SOM method was built by using a dataset with eight common types of objects in road environments, and then it was tested on a public dataset LabelMe. The results showed that the CAE-SOM method performed well with an average accuracy of 90%. Compared with current unsupervised methods, the CAE-SOM model could improve the accuracy by nearly 10%. Compared with current supervised methods, this new model was still competitive, and its accuracy was close to the highest one. More importantly, the CAE-SOM model could reduce the cost of human labeling work in an unsupervised way and handle data from new categories that had never appeared. The outcomes could contribute to the visual algorithm optimization and safety improvement for autonomous vehicles.

Autonomous vehicle (AV) platoons, especially those with the bidirectional communication topology, have significant practical value, as they not only increase link capacity and reduce vehicle energy consumption, but also reduce the consumption of communication resources. Small gaps between AVs in a platoon easily lead to emergency braking or even collisions between consecutive AVs. This paper applies barrier Lyapunov functions to collision avoidance between AVs in a bidirectional platoon during trajectory tracking. Based on backstepping technique, an adaptive collision-free platoon trajectory tracking control algorithm is developed to distributedly design control laws for each AV in the platoon. The control algorithm does not need to introduce additional car-following models to simulate AV driving, and only needs to integrate the position trajectories of consecutive AVs to avoid inter-vehicle collisions. Two sign functions are introduced into the control laws of each AV to ensure strong string stability for bidirectional AV platoons. Moreover, uncertainties and external disturbances in vehicle motion are effectively compensated by introducing adaptation laws. Strong string stability is rigorously proved. CarSIM-based comparison simulations verify the effectiveness of the proposed control algorithm in avoiding inter-vehicle collisions, compensating for uncertainties in vehicle motion, and suppressing the amplification of spacing errors along the platoon.