Human-driven Vehicles Research Articles

Driving style recognition based on a Bayesian belief-renewing method

Developing connected and automated vehicle (CAV) technologies that can interact effectively with surrounding intelligent agents is a crucial challenge to cope with the near-future mixed-flow environment composed of CAVs and human-driven vehicles (HDVs). A fundamental problem to be solved is understanding and recognising driving styles, which can help CAVs reason and predict human driving behaviour. Motivated by it, this study designs a belief-renewing method to recognise the styles of different driving primitives. We first propose a hierarchical Dirichlet process hidden semi-Markov model (HDP-VAR-GP-HSMM) considering time-varying properties to extract driving primitives from natural driving behaviours. The extracted primitives are divided into two driving styles (aggressive or cautious) to build a training dataset with a multi-dimensional k-means clustering algorithm. Next, a Long Short-Term Memory (LSTM) neural network with the Bayesian posterior correction model is trained to recognise the driving styles of primitives. The recognition model can dynamically renew the belief of CAVs on target vehicles' driving styles to mitigate the negative effect caused by stochastic fluctuations. The proposed algorithm is tested using an open-source freeway merging zone dataset. Experimental results show that: (1) the proposed HDP-VAR-GP-HSMM can capture time-varying characteristics of driving primitives and achieves better performance compared with other nonparametric approaches, (2) the accuracy of recognition models with Bayesian belief-renewing method is higher than conventional models. Related approaches and findings in this study can be helpful in the decision-making and control of CAVs.

A field implementation of traffic speed harmonisation on the highway with long-tunnel clusters

Traffic Speed Harmonization (SH) is an efficient active traffic control strategy. However, previous studies rarely put it into field implementations. Moreover, they focus on a common scenario where no severe capacity reduction exists at the downstream bottleneck. This paper focuses on the highway with long-tunnel clusters. This scenario requires more on the SH controller, since more temporary bottlenecks are caused by more traffic accidents and there are not enough regulating spaces between tunnels. The proposed SH controller can efficiently manage these temporary bottlenecks, by guiding the Human-driven Vehicles (HVs) and controlling Connected and Automated Vehicles (CAVs). The proposed controller is evaluated by simulation and verified in field implementations. Results show that it can: (i) harmonize traffic speed for different traffic flow and CAV penetration rate; (ii) enhance fuel efficiency by 16.3%; (iii) enhance TTC by 337%; (iv) reduce speed fluctuations by 38.5% and the number of stops by 58.6%.

Enhancing Mixed Traffic Flow with Platoon Control and Lane Management for Connected and Autonomous Vehicles

As autonomous driving technology advances, connected and autonomous vehicles (CAVs) will coexist with human-driven vehicles (HDVs) for an extended period. The deployment of CAVs will alter traffic flow characteristics, making it crucial to investigate their impacts on mixed traffic. This study develops a hybrid control framework that integrates a platoon control strategy based on the “catch-up” mechanism with lane management for CAVs. The impacts of the proposed hybrid control framework on mixed traffic flow are evaluated through a series of macroscopic simulations, focusing on fundamental diagrams, traffic oscillations, and safety. The results illustrate a notable increase in road capacity with the rising market penetration rate (MPR) of CAVs, with significant improvements under the hybrid control framework, particularly at high MPRs. Additionally, traffic oscillations are mitigated, reducing shockwave propagation and enhancing efficiency under the hybrid control framework. Four surrogate safety measures, namely time to collision (TTC), criticality index function (CIF), deceleration rate to avoid a crash (DRAC), and total exposure time (TET), are utilized to evaluate traffic safety. The results indicate that collision risk is significantly reduced at high MPRs. The findings of this study provide valuable insights into the deployment of CAVs, using control strategies to improve mixed traffic flow operations.

The right-of-way assignment strategy under mixed traffic: considering drivers’ confidence in CAVs

With the development of communication and automated driving technology, connected and automated vehicles (CAVs) will coexist with human-driven vehicles (HDVs) for a long time. Can CAVs and HDVs share a lane or should there be a CAV-only lane? This paper considers drivers’ confidence in CAVs and establishes a car-following model for driving behaviour simulation. The setting condition for the CAV-only approach and the control measure without it are then proposed to optimize carbon emissions and traffic efficiency. The simulation results show that the best CAV permeability range and the corresponding impact of drivers’ confidence heterogeneity on this permeability range for setting the CAV-only approach are obtained. Moreover, the control measures that enable CAVs to gather at the beginning of the platoon help optimize the efficiency and emissions without the CAV-only approach. The results provide a reference for right-of-way assignment and collaborative control with different levels of CAV permeability and confidence.

Highway On-Ramp Truck Platooning Based on Deep Reinforcement Learning

The advancement of connected and automated trucks (CATs) offers a novel avenue for the freight industry to optimize fuel efficiency, enhance traffic flow, and bolster safety through platooning. Particularly at highway on-ramps, how to effectively form CAT platoons is a key research topic. Previous studies have not addressed the issue of platooning at on-ramps. Within mixed traffic conditions, the interference from human driven vehicle (HDV) notably complicates the platooning operation. In this process, the position, timing, and speed of CAT merging significantly affect energy consumption and traffic safety. Thus, this work introduces a hierarchical merging strategy with a dynamic mechanism for selecting optimal merging points. Its objective is to facilitate efficient autonomous CAT platooning at highway on-ramps while considering the interference of HDVs. Specifically, it uses a model-free deep reinforcement learning method that guides the platooning process by exploring optimal driving behaviors. It ensures the safety and efficiency of the CAT merging process. Furthermore, we incorporate a realistic vehicle dynamics model within our simulations. The proposed strategy can handle the variation of the CATs’ initial positions and speeds at on-ramps, as well as interference caused by HDVs at highway mainline. The efficacy of the proposed strategy has been verified through a series of simulation experiments. The findings indicate that our strategy is capable of effectively orchestrating CAT platooning maneuvers at the highway on-ramps.

Cooperative control of self-learning traffic signal and connected automated vehicles for safety and efficiency optimization at intersections.

Cooperative control of intersection signals and connected automated vehicles (CAVs) possess the potential for safety enhancement and congestion alleviation, facilitating the integration of CAVs into urban intelligent transportation systems. This research proposes an innovative deep reinforcement learning-based (DRL) cooperative control framework, including signal and speed modules, to dynamically adapt signal timing and CAV velocities for traffic safety and efficiency optimization. Among the DRL-based signal modules, a traffic state prediction model is merged with the current state to augment characteristics and the agent-learning process. A multi-objective reward function is designed to evaluate safety and efficiency using a traffic conflict prediction model and vehicle waiting time. The double deep Q network (DDQN) model is used to design the agent observing the traffic state, learning the optimal signal control policy, and then inputting the signal phase into the speed module. Based on the green duration analysis and constraints of mixed traffic flow of CAVs and human-driven vehicles, a speed planning model is constructed to optimize CAVs' speed and alter traffic state, which in turn affects the agent's next signal decisions. The proposed framework is tested at isolated intersections simulated by two real-world intersections in Changsha, China. The results reveal the superiority of the proposed method over DRL-based traffic signal control (DRL-TSC) in terms of coverage speed and computation time. Compared to actuated signal control, adaptive traffic signal control, and DRL-TSC, the proposed method significantly optimizes traffic safety and efficiency across diverse intersections, temporal spans, and traffic demands. Furthermore, the advantage of the proposed method substantially amplifies with the increased CAV penetration, regardless of the intersection types.

A speed optimization model for connected and autonomous vehicles at expressway tunnel entrance under mixed traffic environment.

Rear-end collisions frequently occurred in the entrance zone of expressway tunnel, necessitating enhanced traffic safety through speed guidance. However, existing speed optimization models mainly focus on urban signal-controlled intersections or expressway weaving zones, neglecting research on speed optimization in expressway tunnel entrances. This paper addresses this gap by proposing a framework for a speed guidance model in the entrance zone of expressway tunnels under a mixed traffic environment, comprising both Connected and Autonomous Vehicles (CAVs) and Human-driven Vehicles (HVs). Firstly, a CAV speed optimization model is established based on a shooting heuristic algorithm. The model targets the minimization of the weighted sum of the speed difference between adjacent vehicles and the time taken to reach the tunnel entrance. The model's constraints incorporate safe following distances, speed, and acceleration limits. For HVs, speed trajectories are determined using the Intelligent Driver Model (IDM). The CAV speed optimization model, represented as a mixed-integer nonlinear optimization problem, is solved using A Mathematical Programming Language (AMPL) and the BONMIN solver. Safety performance is evaluated using Time-to-Collision (TTC) and speed standard deviation (SD) metrics. Case study results show a significant decrease in SD as the CAV penetration rate increases, with a 58.38% reduction from 0% to 100%. The impact on SD and mean TTC is most pronounced when the CAV penetration rate is between 0% and 40%, compared to rates above 40%. The minimum TTC values at different CAV penetration rates consistently exceed the safety threshold TTC*, confirming the effectiveness of the proposed control method in enhanced safety. Sensitivity analysis further supports these findings.

Influence of information flow topology and maximum platoon size on mixed traffic stability

Beyond platooning with pure connected and automated vehicles (CAVs), mixed platooning with both CAVs and human-driven vehicles (HDVs) emerges as a practical formation pattern in mixed traffic flow. This paper investigates how the two types of information flow topology (IFT) as “looking ahead” or “looking behind”, alongside the maximum size of mixed platoons, influence the stability of the entire traffic flow. Precisely, the mixed traffic system is partitioned into three subsystems: independent HDVs, independent CAVs, and mixed platoons, and a dynamical modeling framework is established under two types of specific IFT: Multi-Predecessor Following (MPF) for “looking ahead” and Multi-Successor Leading (MSL) for “looking behind”. Then, a unified string stability evaluation method is proposed, which captures the role of IFT and maximum platoon size, the latter being associated with the emergence probability of different subsystems and CAV penetration rates. Practical considerations, such as the heterogeneity of vehicle dynamics and the absence of HDV communication capabilities, are also incorporated. Extensive numerical analysis and nonlinear traffic simulations reveal that “looking behind” shows superior performance in mitigating traffic perturbations, especially at low penetration rates, compared with the prevailing choice of “looking ahead”. In addition, increasing the platoon size could further enhance traffic flow stability. These results provide practical insights into the design of IFT and platoon size for CAV cooperation in future mixed traffic environments.

An optimization-free approximation Framework for Connected and Automated Vehicles Eco-Trajectory Planning Under limited computing capacity

The trajectory planning problem (TPP) has become increasingly crucial in the research of next-generation transportation systems, but it presents challenges due to the non-linearity its constraints. One specific case within TPP, namely the Eco-trajectory Planning Problem (EPP), poses even greater computational difficulties due to its nonlinear, high-order, and non-convex objective function. This paper proposes an optimization-free framework to address the eco-trajectory planning problem of connected and automated vehicles (CAVs) under a limited computing capacity scenario. The framework consists of an offline module and an online module. In the offline module, an optimal eco-trajectory batch is constructed by solving a sequence of simplified optimization problems to minimize fuel consumption, considering various initial and terminal system states. Each candidate trajectory in the batch yields the lowest fuel consumption subject to a specific travel time from the vehicle entry to the departure from the intersection. In the online module, an approximation framework is proposed, which greatly improves the computational efficiency of planning and only suffers from a limited extent of optimality losses through a batch-based selection process because optimization and calculation are pre-computed in the offline module. Numerical tests are presented and discussed to evaluate the computational quality and efficiency of the optimization-free approximation framework under a mixed-traffic flow environment that incorporates human-driving vehicles (HDV) and connected and automated vehicles (CAVs) with different market penetration rates (MPR) and the stochasticity of HDVs.

Estimation of mixed traffic flow including the impact of bus bay in the connected environment

The incorporation of connectivity and automation is crucial to the development of urban transportation, with connected and automated buses (CABs) anticipated to enhance network efficiency, however, bus stops, particularly bus bays, will substantially impact surrounding traffic, resulting in lane changes and additional delays. A novel analytical model is derived to capture the dynamics of the mixed traffic flow, which includes CABs, CAVs (connected and automated vehicles), and HVs (human-driven vehicles), according to the fundamental relationship between speed and density. This model separately accounts for the average headway considering platooning intensity, the increase in density due to lane changes, and the additional delay incurred by CABs entering and exiting the bus bay. Furthermore, a simulation model is constructed, wherein control rules for longitude movements and lane-changing maneuvers are defined according to different driving mechanisms. The simulation results are used to propose mutual verification for the accuracy of the proposed analytical model. Furthermore, sensitivity analyses are conducted to assess the influence of various parameters, including the proportion of left-turning buses, the number of arriving CABs, the proportion of stopping buses, and the location of bus bay.

Complexity Evaluation of Test Scenarios for Autonomous Vehicle Safety Validation Using Information Theory

The validation of autonomous vehicles remains a vexing challenge for the automotive industry’s goal of fully autonomous driving. The systematic hierarchization of the test scenarios would provide valuable insights for the development, testing, and verification of autonomous vehicles, enabling nuanced performance evaluations based on scenario complexity. In this paper, an information entropy-based quantification method is proposed to evaluate the complexity of autonomous vehicle validation scenarios. The proposed method addresses the dynamic uncertainties within driving scenarios in a comprehensive way which includes the unpredictability of dynamic agents such as autonomous vehicles, human-driven vehicles, and pedestrians. The numerical complexity calculation of the approach and the ranking of the scenarios are presented through sample scenarios. To automate processes and assist with the calculations, a novel software tool with a user-friendly interface is developed. The performance of the approach is also evaluated through six example driving scenarios, then through extensive simulation using an open-source microscopic traffic simulator. The performance evaluation results confirm the numerical classification and demonstrate the method’s adaptability to diverse scenarios with a comparison of complexity calculation ranking to the ratio of collision, near collision, and normal operation tests observed during simulation testing. The proposed quantification method contributes to the improvement of autonomous vehicle validation procedures by addressing the multifaceted nature of scenario complexities. Beyond advancing the field of validation, the approach also aligns with the broad and active drive of the industry for the widespread deployment of fully autonomous driving.

Traffic oscillation mitigation with physics-enhanced residual learning (PERL)-based predictive control

Real-time vehicle prediction is crucial in autonomous driving technology, as it allows adjustments to be made in advance to the driver or the vehicle, enabling them to take smoother driving actions to avoid potential collisions. This study proposes a physics-enhanced residual learning (PERL)-based predictive control method to mitigate traffic oscillation in the mixed traffic environment of connected and automated vehicles (CAVs) and human-driven vehicles (HDVs). The introduced model includes a prediction model and a CAV controller. The prediction model is responsible for forecasting the future behavior of the preceding vehicle on the basis of the behavior of preceding vehicles. This PERL model combines physical information (i.e., traffic wave properties) with data-driven features extracted from deep learning techniques, thereby precisely predicting the behavior of the preceding vehicle, especially speed fluctuations, to allow sufficient time for the vehicle/driver to respond to these speed fluctuations. For the CAV controller, we employ a model predictive control (MPC) model that considers the dynamics of the CAV and its following vehicles, improving safety and comfort for the entire platoon. The proposed model is applied to an autonomous driving vehicle through vehicle-in-the-loop (ViL) and compared with real driving data and three benchmark models. The experimental results validate the proposed method in terms of damping traffic oscillation and enhancing the safety and fuel efficiency of the CAV and the following vehicles in mixed traffic in the presence of uncertain human-driven vehicle dynamics and actuator lag.

SIMULATING VEHICLE-TO-VEHICLE COMMUNICATION AT ROUNDABOUTS

Smart roads integrate advanced technology to enhance safety, effectiveness, and eco-friendliness, revolutionizing transportation systems. Within this framework, connectivity and automation work together to improve road mobility and energy efficiency. However, there are still uncertainties regarding their effect on road safety and traffic operations, as well as assessment methods, especially for intersections and roundabouts. Navigating roundabouts involves complex decision-making influenced by cooperative and competitive interactions between human-driven vehicles and connected and autonomous driving vehicles (CADVs), affecting gap-acceptance patterns. This research employs Aimsun Next software to simulate the rising market penetration percentages of CADVs by examining assumption-based behavior on single-lane roundabouts. It integrates CADV-based capacity modification factors from the Highway Capacity Manual 2022 and compares adapted capacity curves for CADVs, with simulated capacities for model calibration. Critical model parameters affecting CADVs; ability to enhance roundabout safety and throughput are identified, with a focus on the transition towards cooperative driving in the context of smart roads.

A Three-Stage Cellular Automata Model of Complex Large Roundabout Traffic Flow, with a Flow-Efficiency- and Safety-Enhancing Strategy.

Intelligent transportation systems (ITSs) present new opportunities for enhanced traffic management by leveraging advanced driving behavior sensors and real-time information exchange via vehicle-based and cloud-vehicle communication technologies. Specifically, onboard sensors can effectively detect whether human-driven vehicles are adhering to traffic management directives. However, the formulation and validation of effective strategies for vehicle implementation rely on accurate driving behavior models and reliable model-based testing; in this paper, we focus on large roundabouts as the research scenario. To address this, we proposed the Three-Stage Cellular Automata (TSCA) model based on empirical observations, dividing the vehicle journey over roundabouts into three stages: entrance, following, and exit. Furthermore, four optimization strategies were developed based on empirical observations and simulation results, using the traffic efficiency, delay time, and dangerous interaction frequency as key evaluation indicators. Numerical tests reveal that dangerous interactions and delays primarily occurred when the roundabout Road Occupancy Rate (ρ) ranged from 0.12 to 0.24, during which times the vehicle speed also decreased rapidly. Among the strategies, the Path Selection Based on Road Occupancy Rate Recognition Strategy (Simulation 4) demonstrated the best overall performance, increasing the traffic efficiency by 15.65% while reducing the delay time, dangerous interactions, and frequency by 6.50%, 28.32%, and 38.03%, respectively. Additionally, the Entrance Facility Optimization Strategy (Simulation 1) reduced the delay time by 6.90%. While space-based optimization strategies had a more moderate overall impact, they significantly improved the local traffic efficiency at the roundabout by approximately 25.04%. Our findings hold significant practical value, particularly with the support of onboard sensors, which can effectively detect non-compliance and provide real-time warnings to guide drivers in adhering to the prescribed traffic management strategies.

Real-time executable platoon formation approach using hierarchical cooperative motion planning framework

While connected and automated vehicle (CAV) platooning holds promise for enhancing traffic efficiency and reducing energy consumption, we still lack efficient algorithms for guiding the local movements of CAVs to form a platoon on a road due to significant computational and control challenges. This study addresses this gap by designing a real-time executable Hierarchical and Recursive Platoon Formation (HR-PF) framework tailored to mixed flow traffic conditions that encompass both Human-Driven Vehicles (HDV) and CAVs. The HR-PF framework comprises three hierarchical mathematical models (modules) designed to optimize platoon formation while considering both macroscopic traffic conditions and microscopic traffic safety. Module-I formulates a mixed integer quadratic program to determine the timing, location, and state of platoon formation. It is further extended to a mixed integer nonlinear program so that we can also select the optimal size of the target platoon. Module-II designs a Hybrid State-Lattice Motion planner to generate optimal trajectory references for CAVs to approach the target platoon state, ensuring microscopic traffic safety. Module-III develops longitudinal and lateral controllers to enable CAVs to track trajectory references accurately. These models function recursively at varying frequencies to balance mathematical rigor with practical application. Numerical experiments demonstrate that HR-PF facilitates efficient platoon formation in real-time across diverse traffic scenarios and road geometries while sustaining traffic efficiency. Furthermore, the performance of platoon formation is affected by surrounding traffic density and CAV penetrations, with prompt formation observed under LOS C and D traffic environments and slightly more traffic impacts under LOS E and F. These findings provide robust support for exploring advanced platoon formation and platooning strategies for CAVs under complicated traffic environments.

FAIRLANE: A multi-agent approach to priority lane management in diverse traffic composition

The rise of autonomous driving technologies prompts a reevaluation of traditional urban traffic control and lane management. Dedicated lanes for connected and autonomous vehicles (CAVs), with intermittent access for other vehicles, have been proven to enhance road capacity and reduce underutilization moderately. However, this assumes all non-CAVs are smart vehicles, which is different from the current baseline for the street vehicle mix. Presently, our streets feature a mix of CAVs, smart vehicles, and human-driven vehicles, and the research on dedicated lanes using the realistic mixed traffic environment is missing. In this paper, we investigate the enhancement of road utilization using realistic mixed traffic combinations and identify the penetration rate of CAVs and smart vehicles necessary to improve baseline utilization. Previous studies have focused on lane-management strategies in single-vehicle settings, neglecting the interaction of CAVs with neighboring CAVs and smart vehicles. Therefore, we propose a multi-agent reinforcement learning-based framework to facilitate fair utilization of priority lanes, considering driving comfort, traffic efficiency, and safety during lane-changing. Through multiple experiments on a realistic network, our results demonstrate that the proposed framework significantly improves traffic efficiency, particularly when the penetration rate of CAVs is below 40% and Semi-Autonomous Vehicles (SAVs) constitute 50% of the remaining vehicles. The framework outperforms traditional lane management strategies, reducing mean waiting time and increasing average speed. This study provides nuanced information on different vehicle penetration rates, enabling more informed decisions on when to install priority lanes. This highlights the importance of considering mixed traffic environments in designing autonomous vehicle infrastructure and sets the stage for future advancements in urban traffic management.

Intersection decision making for autonomous vehicles based on improved PPO algorithm

AbstractThe deployment of autonomous vehicles (AVs) in complex urban environments faces numerous challenges, especially at intersections where they coexist with human‐driven vehicles (HVs), resulting in increased safety risks. In response, this study proposes an improved control strategy based on the Proximal Policy Optimization (PPO) algorithm, specifically designed for hybrid intersections, known as MSA‐PPO. First, the Self‐Attention Mechanism (SAM) is introduced into the algorithmic framework to quickly identify the surrounding vehicles with a greater impact on the ego vehicle from different perspectives, accelerating data processing and improving decision quality. Second, an invalid action masking mechanism is adopted to reduce the action space, ensuring actions are only selected from feasible sets, thereby enhancing decision efficiency. Finally, comparative and ablation experiments in hybrid intersection simulation environments of varying complexity are conducted to validate the algorithm's effectiveness. The results show that the improved algorithm converges faster, achieves higher decision accuracy, and demonstrates the highest speed levels during driving compared to other baseline algorithms.

Car following trajectory planning of CAVs: An improved APF model with considering the stochasticity of HDVs

This paper considered the stochasticity of Human-Driven Vehicles (HDVs) and proposed an improved Artificial Potential Field (APF) method for car-following trajectory planning of Connected and automated vehicles (CAVs) based on real mixed traffic flow experiment. Firstly, the heterogeneity between HDVs and CAVs was considered to determine the type of attractive field. Then to adapt the APF model to dynamic traffic environments, the field functions were improved by incorporating the speed and position differences. In addition, taking into account both the vehicle itself and its impact on traffic flow, Grey Wolf Optimization-Chaos (GWO-C) was proposed to calibrate parameters, which helps avoid local optima. Furthermore, the proposed model was compared with experimental data and original APF method. The results show that the proposed APF improves the speed, jerk, and speed standard deviation of the platoon. Finally, the impact of different CAVs’ market penetration rates (MPR) on the stability and fundamental diagram were explored. It was found that traffic stability and capacity can be enhanced by CAVs, with a more significant impact at higher MPR.

A hierarchical intersection system control framework in mixed traffic conditions

Signalized intersections play a crucial role in urban road system and are a significant source of vehicle delays. Conventional intersection control methods struggle to cope with increasingly challenging traffic conditions. Fortunately, the development of artificial intelligence(AI), information communication technology(ICT) and connected and automated vehicle(CAV) technologies has brought new opportunities to enhance traffic performance at intersections. Based on these technologies, in the mixed traffic environment of CAVs and connected human-driven vehicles (CHVs), this paper proposes a hierarchical intersection system control framework, which adopts intersection level distributed control and corridor level centralized control. At the intersection level, considering the impact of CAV-dedicated lanes and using the CAV platoon and “1 + n” platoon consisting of one leading CAV and n following CHVs as control units, the mixed integer quadratic programming problem and optimal control problem are formulated to collaboratively optimize signal timing, lane settings and vehicle trajectories at isolated signalized intersections. At the corridor level, based on the optimization results of the intersection level and the critical path information, the critical path performance function and multi-stage optimal trajectory control formula are constructed to optimize the signal offsets and vehicle trajectories, which are passed to the intersection level, and ultimately minimize vehicle delays in the corridor. Simulation experiments and sensitivity analysis demonstrate the effectiveness and stability of the proposed framework in reducing vehicle delay and improving fuel consumption.

Safety-aware human-lead vehicle platooning by proactively reacting to uncertain human behaving

Human-Lead Cooperative Adaptive Cruise Control (HL-CACC) is regarded as a promising vehicle platooning technology in real-world implementation. By utilizing a Human-driven Vehicle (HV) as the platoon leader, HL-CACC reduces the cost and enhances the reliability of perception and decision-making. However, state-of-the-art HL-CACC technology still has a great limitation on driving safety due to the lack of considering the leading human driver’s uncertain behavior. In this study, a HL-CACC controller is designed based on Stochastic Model Predictive Control (SMPC). It is enabled to predict the driving intention of the leading Connected Human-Driven Vehicle (CHV). The proposed controller has the following features: (i) enhanced perceived safety in oscillating traffic; (ii) guaranteed safety against hard brakes; (iii) computational efficiency for real-time implementation. The proposed controller is evaluated on a PreScan&Simulink simulation platform. Real vehicle trajectory data is collected for the calibration of the simulation. Results reveal that the proposed controller: (i) improves perceived safety by 19.17 % in oscillating traffic; (ii) enhances actual safety by 7.76 % against hard brakes; (iii) is confirmed with string stability. The computation time is approximately 3.2 ms when running on a laptop equipped with an Intel i5-13500H CPU. This indicates the proposed controller is ready for real-time implementation.