交通运输工程学报

Cover and Contents of Vol.26, No.3, 2026

Cover Contents

2026, 26(3): .

FullText HTML (83) PDF (4599KB) (60)

Research review on autonomous detect and avoid technologies for unmanned aerial vehicles

TANG Xin-min, GU Jun-wei, ZHANG Kang, MIAO Ya

Abstract: The future complex low-altitude operational environment is characterized by diversity, high density, and high dynamics. To promote autonomous detect and avoid (DAA) systems of unmanned aerial vehicles (UAVs) to adapt to this, the basic functional architecture of the system was introduced based on core technical principles and standards. The applicability and limitations of existing standards in complex low-altitude environments were analyzed. The focus was put on four key technologies including target detection, dynamic tracking, conflict alerting, and avoidance decision-making. Domestic and foreign research achievements for each technology were elucidated. The characteristics of various technical means were compared from different dimensions. The development trends of future autonomous DAA systems were explored. Potential advancements in key processes such as target detection, tracking, early warning/alerting, and autonomous decision-making were discussed. Possible challenges and opportunities in this field were summarized. The research results show that target detection technologies are diverse. However, they are constrained by factors like power consumption, weight, and frequency band resources, making it difficult for them to meet the multi-scenario and fully autonomous detection needs of low-altitude UAVs. Tracking and alerting algorithms suffer from issues, including insufficient architectural versatility, weak generalization ability, and limited computing resources, hindering them from adapting to complex dynamic environments. Avoidance decision-making still relies mainly on traditional rules or non-cooperative optimization, leading to significant limitations in decision-making efficiency and flexibility in high-density UAV collaborative scenarios. In general, DAA technology serves as a core support for ensuring the safe operation of manned/unmanned mixed airspace. Future DAA systems will develop along four major trends, including multi-dimensional surveillance integrating cooperative and non-cooperative technologies, accurate tracking combining multi-target perception and intelligent learning, dynamic early warning with collaborative risk measurement and state deduction, and distributed collaborative autonomous decision-making mechanisms. A reference is then provided for the subsequent research and development, upgrading, and practical application of DAA systems.More>

2026, 26(3): 1-24. doi: 10.19818/j.cnki.1671-1637.2026.085

FullText HTML (177) PDF (3125KB) (86)

Intelligent flying cars: Driving future of urban air mobility

LIANG Jun, DAI Yu-xin, WANG Wen-sa, SHA Yang-yang, DU Xue-wen, CHEN Long

Abstract: The progress of IFC key technologies was systematically concluded from three major aspects: car body design and power systems, autonomous navigation and control technologies, and vehicle-road-cloud collaboration. The current development of diverse configurations such as ducted fans, folding wings, and modular structures, as well as energy architectures including hybrid power and fuel cell was summarized. Methods of autonomous navigation and control, including multi-sensor fusion, deep learning-based path planning, and robust control were mainly reviewed. The application of low-altitude communication networks, cooperative perception, and intelligent scheduling in integrated vehicle-road-cloud frameworks was discussed. Current challenges of IFC in configuration standardization, complex environment perception, cross-modal collaboration, and large-scale scheduling were analyzed, and potential research directions were proposed. Research results show that, driven by artificial intelligence, communication, and energy technologies, IFC technology is evolving from a single-point breakthrough stage dominated by configuration and power toward a comprehensive integration stage centered on intelligent control and system collaboration. By incorporating IFCs into urban integrated transportation systems, a predictable, manageable, and verifiable operational closed loop can be achieved at the transportation system level. Thus, a new pathway is provided for building efficient, safe, and sustainable future urban transportation systems.More>

2026, 26(3): 25-44. doi: 10.19818/j.cnki.1671-1637.2026.150

FullText HTML (128) PDF (2606KB) (60)

Research review on public acceptance of low-altitude unmanned aerial vehicle logistics

ZHANG Sheng-zhong, LIU Tai-li, ZENG Miao-hua, WU Shan

Abstract: To comprehensively understand the research progress in the field of public acceptance of low-altitude unmanned aerial vehicle (UAV) logistics and to promote the coordinated development of technological innovation and social adoption, this paper employed bibliometric analysis and systematic review methods to synthesize and summarize relevant studies. The analysis was conducted from three perspectives: literature distribution, theoretical models, and data analysis methods. Key factors influencing public acceptance of low-altitude UAV logistics were identified, and their impacts were analyzed. A research framework for this field was constructed, and limitations of existing studies as well as future research directions were proposed. Research results show that the number of publications in this area is generally increasing, and the research exhibits interdisciplinary characteristics. Theoretical models are primarily based on the Technology Acceptance Model (TAM) and its extensions, gradually integrating cognitive theories such as perceived risk and trust to form a comprehensive analytical framework. Quantitative methods, especially structural equation modeling (SEM), dominate current research, with a growing trend toward integration with machine learning and big data analytics. Key influencing factors include technical features, service functions, and environmental benefits in terms of drone functional attributes, as well as attitude, perceived ease of use, perceived usefulness, perceived risk, personal innovativeness, trust, and subjective norms in terms of public psychological cognition. However, in the context of the large-scale development of low-altitude UAV logistics, existing research still requires further advancement in developing theoretical frameworks tailored to the Chinese context, enriching data analysis and quantitative methodologies, and expanding decision-making research that incorporates public acceptance. These efforts aim to provide scientific insights and a solid foundation for industry policy formulation and operational optimization.More>

2026, 26(3): 45-59. doi: 10.19818/j.cnki.1671-1637.2026.151

FullText HTML (130) PDF (1527KB) (78)

UAV-guided multi-vehicle cooperative passage control on narrow and curved roads

ZHANG Qian, GUO Ge, WANG Yong-chuan

Abstract: In response to insufficient perception capability, path planning difficulty, and delayed cooperative response of ground vehicles in disaster emergency transportation scenarios involving dense traffic flow and narrow lanes, a predefined-time cooperative control method was proposed for ground vehicles guided by an unmanned aerial vehicle (UAV). The wide-area perception and path planning capability of the UAV was utilized to obtain guiding path information for ground traffic and transmit them to the leading vehicle via wireless communication. Bidirectional inter-vehicle communication was then employed to realize information sharing within the vehicle platoon. An extended look-ahead zero-initial coupled error dynamics based on an exponential spacing policy was designed to remove constraints in cooperative control design and effectively prevent error accumulation and cutting-corner behavior on curved roads. Based on the proposed error dynamics, a distributed vehicle controller was developed using backstepping control and the predefined-time stability lemma. Therefore, the predefined-time stability of individual vehicles, platoon mesh stability, and the existence of traffic flow stability was guaranteed. As a result, cooperative response efficiency and traffic smoothness were improved. Results show that the proposed method achieves accurate path tracking and fast convergence of cooperative errors within the predefined time of 5 s under various complex conditions, including narrow roads and curved paths. Evaluation results based on platoon mesh stability and traffic flow stability indicate that the proposed approach effectively suppresses disturbance propagation caused by information transmission and traffic congestion. Traffic safety and flow efficiency are thus significantly enhanced. Therefore, the proposed method demonstrates good engineering applicability and practical potential. The theoretical and technical support can be provided for disaster emergency transportation in intelligent transportation systems.More>

2026, 26(3): 60-74. doi: 10.19818/j.cnki.1671-1637.2026.152

FullText HTML (115) PDF (7418KB) (57)

Unmanned aerial vehicle cruise risk identification technology based on multi-source data and large models

MA Tao, WU Jun, TANG Fan-long, FAN Jian-wei, WANG Ning

Abstract: To identify complex risk events during the cruise of unmanned aerial vehicles (UAVs), the basic elements of UAV cruise risks were explored, and characteristic parameters required for prompt were specified. The implementation methods, architectures, and typical models of multimodal large models were analyzed, and a scheme for integrating multi-source data in the prompt generation model was proposed. By combining environmental perception, detection, identification and tracking methods, a prompt generation model integrating with macroscopic scene description, dynamic scene supplementation, and sudden risk detection was established. The extracted feature parameters were then integrated into the prompt. The UAV cruise risk identification and judgment were completed through DeepSeek's comprehensive analysis. Research results show that the three modules can quickly complete the identification of UAV cruise risks and obtain complete prompts. The static scene description based on the Owl-ViT model can effectively identify static obstacles during flight, with confidence exceeding 80%. The dynamic object capture based on the ByteTrack algorithm can quickly obtain dynamic information such as the distance, speed, and coordinates of flying birds and other UAVs. The sudden risk identification based on point clouds can capture point cloud obstacle information, including the distance, size, volume, and aspect ratio of the target, and can quickly detect obstacles that suddenly enter the safe area. The output results of DeepSeek generated by the prompt can detail the risk content and level during the cruise, and provide safety suggestions. The developed UAV cruise risk identification system can visualize the perception and identification data and determine the device and task information for the tasks, further assisting DeepSeek in risk judgment. The research results can provide effective technical support for risk identification during UAV cruise as well as safe and efficient flight.More>

2026, 26(3): 75-88. doi: 10.19818/j.cnki.1671-1637.2026.036

FullText HTML (146) PDF (4061KB) (69)

Collaborative location method for drone vertiport and truck parking point in urban logistics

LI Zhuo-lun, LU Jian, WANG Xue-rui, LI Shan

Abstract: To enhance the operational efficiency and service quality of the synchronized truck-drone delivery (STDD) mode and to address the facility layout problem under the STDD mode, a collaborative location method for drone vertiports and truck parking points was proposed. First, based on geographic information data, the three-dimensional urban space was discretized using the grid method. By integrating indicators such as obstacle distribution, noise impact, and pedestrian accessibility, the environmental characteristics of grid units were quantitatively analyzed. Then, by considering factors such as logistics delivery distance and user demand distribution, a multi-objective drone vertiport location model was established. Furthermore, combining conditions such as vertiport service relationships and drone performance, a multi-objective truck parking point service allocation model was constructed. Finally, the fuzzy C-means clustering algorithm was employed to spatially aggregate user demand points. By considering flight safety, noise impact, and pickup efficiency, a relatively optimal vertiport layout scheme was selected from the Pareto front. Based on the distribution data of urban public parking lots, the fuzzy C-means clustering and multi-objective multi-verse optimization algorithms were integrated to obtain the optimal service matching relationship between truck parking points and vertiports. The results show that with the increase in the number of buildings served by a single vertiport, the scale of vertiports shows a decreasing trend, while the average walking distance of residents shows an increasing trend. Notably, the clustering algorithm integrated with the location optimization strategy yields an average environmental score of 0.682 for drone vertiports, representing a 49.2% improvement compared to the clustering algorithm integrated with the proximity-based location strategy. Furthermore, the number of vertiports presents a positive correlation with the number of enabled truck parking points and a negative correlation with the nearest-neighbor parking point spacing. Compared with the traditional multi-objective multi-verse optimization algorithm, the proposed algorithm reduces the number of enabled parking points by 33%, decreases the service imbalance degree by 28.6%, and increases the mean value of nearest-neighbor spacing by 22.01%. The proposed method realizes the optimized location layout of drone vertiports and truck parking points, which can provide technical support for the construction of low-altitude logistics networks in smart cities.More>

2026, 26(3): 89-105. doi: 10.19818/j.cnki.1671-1637.2026.037

FullText HTML (118) PDF (45546KB) (67)

Throughput envelopment analysis of urban low-altitude vertiports based on multi-commodity flow model

CHANG Xin, TANG Yao, TANG Xin-min, GAO Jian-shu, YAO Zhi-hong

Abstract: To explore the takeoff and landing mode selection and influencing factors for throughput of urban low-altitude vertiports, a vertiport multi-commodity flow model was constructed. The impact of the number of touchdown and lift-off facilities (TLOF) and the number of boarding gates on the throughput of vertiports was systematically analyzed. The quantitative calculation of facility utilization rate and the generation of surface throughput envelope map were performed. On this basis, the optimal quantity configuration of support facilities was revealed. The performance differences under different takeoff and landing modes were compared to provide a scientific basis for optimizing the facility configuration of vertiports. The utilization efficiency of surface space was improved and the intensive design and operation of vertiports were achieved. The research results show that the reasonable range of the number of boarding gates in a single TLOF vertiport is 4 - 5. At this time, the utilization of TLOF is 92% and 100%, respectively. The utilization of boarding gates is 55% and 57%, respectively. The waiting rate of boarding gates is 7% and 12%, respectively. Compared with 4 boarding gates, 5 boarding gates provide higher utilization of facilities, but lead to significant departure queues. The independent takeoff and landing mode of multiple TLOF vertiports has the most significant improvement on the surface throughput, which is 57% and 135% higher than that of the single takeoff and landing mode among 2 - 3 TLOF vertiports considered. The influence of TLOF number on throughput is limited under the single takeoff and landing mode. The parallel takeoff and landing mode demonstrates the same throughput growth and maximum unbalanced approach growth as the independent takeoff and landing mode. However, the envelope area of the former is smaller than that of the latter. The research results can provide theoretical and methodological support for calculating the throughput of urban air traffic vertiports, designing infrastructure schemes, and making operational control decisions, thus promoting the standardized development of low-altitude infrastructure construction.More>

2026, 26(3): 106-117. doi: 10.19818/j.cnki.1671-1637.2026.087

FullText HTML (48) PDF (1443KB) (25)

Research on drone hub-rider collaborative delivery model under urban instant delivery conditions

YANG Yang, ZUO Bo-rui, SHANG Ke-qi

Abstract: To address the capacity bottlenecks of the traditional direct delivery model by riders in urban instant delivery and the challenges faced by direct drone delivery in high-density urban environments, a drone hub-rider collaborative delivery model was proposed. A three-stage process of "rider front-end pickup, drone inter-hub transportation, and rider terminal delivery" was adopted, and five extended models were derived. Based on the Logit model, a modal split model containing the dual attributes of economy and timeliness was constructed; the calculation formulas for the critical delivery distance when the utilities of the new model and existing models reached equilibrium were derived, and the competitiveness boundaries were evaluated. Monte Carlo simulation, numerical experiments based on the traveling salesman problem, and a queueing theory model were adopted to estimate key parameters such as the expected value of inter-hub distance, the enroute coefficient, and hub costs. A baseline scenario conforming to the characteristics of urban instant delivery in China was constructed, and the impacts of factors such as road network conditions and hub locations on the competitiveness of the delivery model were explored through sensitivity analysis. The results indicate that the critical delivery distance of the collaborative delivery model under the baseline scenario is 3.88 km, which is within the main service range of the direct rider delivery model. When the distance between the hub and supply/demand points is 2.57–4.32 km, and the delivery distance is 5.17–9.49 km, transfers are required due to the endurance limit of the drone, which leads to a discontinuous distribution of the advantage intervals.The critical delivery distances of the five extended models range from 1.57 to 3.88 km, presenting differentiated competitive characteristics.The demand-side self-pickup mode has the lowest critical delivery distance (1.57 km), which is suitable for customer self-pickup scenarios; the end-to-end direct delivery and fully autonomous delivery modes have the second lowest critical delivery distances, but both require supply-side self-equipped hubs, with the fully autonomous delivery mode additionally demanding a higher level of automation. In multi-order scenarios, the collaborative delivery model maintains a competitive advantage over a wider distance range for time-sensitive goods. The proposed method provides a new idea for solving the adaptability problem of drones in urban delivery and can provide a quantitative analysis tool for drone delivery network planning, hub location layout, and low-altitude economic policy formulation.More>

2026, 26(3): 118-139. doi: 10.19818/j.cnki.1671-1637.2026.088

FullText HTML (92) PDF (37342KB) (40)

Multi-UAV cooperative path planning method integrating risk avoidance and combined strategy

YANG Zhao, QI Hong-biao, YU Yang-yang, GUO Yue-xiang, LI Jia-chen

Abstract: To address challenges such as safety, economic loss, social impact, and operational efficiency faced by unmanned aerial vehicles (UAVs) operating in complex urban environments during the development of the low-altitude economy, a multi-UAV cooperative path planning method integrating risk avoidance and a combined strategy was proposed. First, based on the real urban environment, a three-dimensional grid map was constructed, and multi-source risk information was integrated to establish a risk assessment and dynamic risk map model. Second, for a single UAV, the path planning algorithm was improved based on the dynamic risk map to guide the aircraft to actively avoid high-risk areas, achieving the synergistic minimization of total path risk and length. Third, to address the conflict problem among multiple UAVs, a priority calculation model was constructed by comprehensively considering indices such as conflict risk, path length, path risk, and remaining distance ratio, and a combined conflict resolution strategy was formulated to achieve multi-UAV cooperative path planning. Experimental results indicate that compared with the Dijkstra algorithm, ant colony algorithm, and particle swarm optimization algorithm, the proposed improved algorithm reduce the path risk by 6.59%, 25.94%, and 20.24%, respectively and decrease the path length by 9.80%, 11.94%, and 9.54%, respectively. In the multi-UAV cooperative planning, for scenarios with 5, 10, and 15 UAVs, the calculation times of the designed combined conflict resolution strategy is reduced by 24.56%, 27.42%, and 36.42%, respectively, compared with the path replanning strategy; the task durations is reduced by 2.83%, 3.29%, and 4.09%, respectively, compared with the starting point waiting strategy. The method can efficiently resolve conflicts between UAVs and finally generate cooperative flight paths with both safety and cost-effectiveness.More>

2026, 26(3): 140-158. doi: 10.19818/j.cnki.1671-1637.2026.089

FullText HTML (63) PDF (12676KB) (32)

MPC-based control strategy for conflict resolution between vehicles and aircraft on airport

ZHANG Hai-yan, ZHANG Jian, OUYANG Jie, YUAN Xun-ming

Abstract: To address the frequent conflicts between airport ground service vehicles and aircraft at intersections of taxiways and roadways, a model predictive control (MPC)-based conflict resolution method was proposed for airport autonomous vehicles and aircraft. The typical operation scenarios of ground vehicle-aircraft intersections on the airport were analyzed. The longitudinal dynamic constraints of vehicles and the safety boundary requirements of aircraft in conflict situations were clarified. On this basis, the period when the aircraft and its safety clearance passed through the conflict area was defined as a dynamic red light time window constraint. With the objectives of minimizing vehicle energy consumption and maximizing traffic efficiency, an MPC model for vehicle was established. A sequential quadratic programming method was employed to solve the nonlinear constrained optimization problem in a rolling manner, thereby generating optimal speed and acceleration control sequences for vehicles in real time. Simulation experiments were conducted using real-world scenario data from Tianjin Binhai International Airport. Multiple sets of random operating conditions were also designed for comparative analysis. Research results show that, with the proposed MPC strategy, the vehicle can avoid potential conflicts with aircraft. Compared with conventional human driving and idealized human driving, the average energy consumption is reduced by 14.81% and 14.27%, respectively, and the average travel time is shortened by 6.48% and 5.70%. The trajectory smoothness and control performance are both significantly superior. These findings indicate that the proposed conflict resolution control strategy not only enhances the safety and efficiency of airport ground vehicle operations but also provides practical theoretical support and technical reference for the future application and popularization of autonomous vehicles on airports.More>

2026, 26(3): 159-170. doi: 10.19818/j.cnki.1671-1637.2026.090

FullText HTML (68) PDF (5678KB) (20)

UAV-assisted jam-absorption driving strategy for on-ramp weaving sections

LIU Yue, LIANG Guo-hua, CHEN Zi-yu, TIAN Xin, CHEN Yi-xin, MENG Xiao-yang

Abstract: Self-exciting and propagating stop-and-go waves are easily generated in expressway entrance ramp weaving sections, affecting traffic efficiency and energy consumption. Therefore, for a future application scenario of coordination between unmanned aerial vehicles (UAVs) and connected autonomous vehicle (CAV), a UAV-assisted jam-absorption driving strategy for entrance ramp weaving sections was proposed and validated. A multi-scenario comparative evaluation was conducted under the condition of a low CAV penetration rate. Based on four preset schemes, namely, UAV observation only as the baseline; traditional vehicle-road coordination jam-absorption; adaptive dynamic control jam-absorption; and UAV and CAV coordinated jam-absorption, an integrated process was constructed, and identification, prediction, and control were completed sequentially. Multiple UAVs were used to continuously observe the weaving sections to identify the significant speed-drop region and its propagation direction, determining the spatial location and movement trend of the stop-and-go waves. The time range for the stop-and-go wave to pass a key upstream location was calculated to form an arrival time window, and the control trigger time and the target speed were determined accordingly. Within the arrival time window, CAVs satisfying communication and safety constraints were selected from the traffic flow. A gentle speed stabilization control was applied on them to allow these vehicles to decelerate slightly before entering the weaving section. Their speed gradually recovered after passing through. Throughout the entire process, safety distance, acceleration and deceleration limits, and a speed recovery threshold were set to ensure feasibility and safety. The analysis results show that, in an entrance ramp weaving section scenario constructed on an open-source microscopic traffic simulation platform, compared to the UAV observation-only scenario, the average travel time is reduced from 65.78 s to 63.71 s, a decrease of 3.1%, by the UAV and CAV coordinated jam-absorption. Wave-level indicators show that congestion severity is reduced, and the speed distribution is shifted upward overall. At a 2% penetration rate, the trigger coverage rate and the vehicle selection success rate remain stable. Under the same demand and disturbance intensity, its wave-suppression benefit is superior to the traditional vehicle-road coordination scheme with a 5% penetration rate, and also to the adaptive dynamic control strategy scheme with a 2% penetration rate. Implementable jam-absorption governance can be achieved through high-angle observation provided by UAVs and speed stabilization intervention by CAVs under conditions of low penetration rate and light roadside infrastructure. The approach is applicable to expressway entrance ramp weaving sections and possesses the potential for coordinated application with variable speed limits and ramp metering.More>

2026, 26(3): 171-184. doi: 10.19818/j.cnki.1671-1637.2026.091

FullText HTML (71) PDF (2134KB) (26)

Real-time 3D conflict resolution method for low-altitude heterogeneous aircraft based on multi-agent proximal policy optimization

CHEN Yun-xiang, GOU Ming, ZHANG Jian-ping, LU Wei-ning, TANG Kai, ZHANG Guang-yuan

Abstract: In response to the real-time three-dimensional conflict resolution for low-altitude heterogeneous aircraft, a rapidly developing operational scenario was studied, including shared airspace operations between medium-to-large fixed-wing aircraft and light small multi-rotor unmanned aerial vehicles (UAVs). A multi-agent proximal policy optimization (MAPPO)-based method was proposed with a centralized training and decentralized execution framework. Based on the operational characteristics of the two types of aircraft, a real-time three-dimensional conflict resolution strategy was established to allow fixed-wing aircraft to maintain stable flight while multi-rotor UAVs perform avoidance maneuvers. A multi-dimensional reward function was designed, taking into account collision avoidance, mission efficiency, priority, and smoothness. A priority mechanism was introduced to ensure the mission priority of fixed-wing aircraft and encourage proactive avoidance by multi-rotor UAVs. Simulation results show that baseline tests involving 5, 10, 20, and 30 light small multi-rotor UAVs all achieve a mission success rate of over 92%, with computational overhead ranging from 0.16 to 0.36 min, average conflict resolution time between 0.28 and 1.76 s, and flight conflict proportions between 0.95% and 2.18%. Through optimization of the state space, action space, and reward function, the proposed method reduces conflict resolution time by 2.25 s and improves mission success rate by 2% compared to existing methods. A foundation is thus laid for further research on the integrated operation of low-altitude heterogeneous aircraft in wide-area scenarios.More>

2026, 26(3): 185-197. doi: 10.19818/j.cnki.1671-1637.2026.092

FullText HTML (66) PDF (5560KB) (30)

Autonomous avoidance decision-making and control method for eVTOL aircraft under non-cooperative differential games

ZHAO Xin-yi, WANG Yan-tao, ZHAO Yi-fei

Abstract: To ensure the safe autonomous operation of electric vertical take-off and landing (eVTOL) aircraft in urban airspace, an autonomous avoidance decision-making and control method for eVTOL aircrafts under non-cooperative games was proposed for non-cooperative aircraft intrusion scenarios. Optimal control models were constructed respectively for aircraft with different flight goals and avoidance intentions, and continuous and hybrid action spaces were adopted to express control inputs. A multi-aircraft decision-making model based on non-cooperative differential game theory was established to characterize the avoidance or priority-taking behaviors of aircraft in conflict scenarios. An event-triggered mechanism was integrated with a model predictive control framework, and a rolling optimization process of trajectory prediction, conflict detection, optimization calculation, and control execution was adopted to solve the single-step optimal control command for aircraft in real time. An iterative best response algorithm was adopted to progressively approach the Nash equilibrium solution of the non-cooperative differential game to improve online computational efficiency. Based on the proposed models and algorithms, autonomous avoidance simulation experiments of eVTOL aircraft were conducted under head-on, same-direction, and crossing conflict scenarios. Simulation results show that when the intruding aircraft is predicted to have no avoidance intention, the avoidance effect is better. The composite maneuver strategy of "speed adjustment+direction adjustment+altitude adjustment" improves avoidance safety by 32%, increases avoidance efficiency by 53%, and reduces maximum deviation distance by 88%. The average computation time per step of the game optimization algorithm based on iterative best response is less than 0.3 s, and the response speed is fast. The proposed autonomous avoidance decision-making and control method enables eVTOL aircraft to rapidly generate optimal control strategies when facing non-cooperative target conflicts, thereby achieving safe and efficient autonomous avoidance.More>

2026, 26(3): 198-214. doi: 10.19818/j.cnki.1671-1637.2026.093

FullText HTML (42) PDF (6996KB) (20)

A dynamic collision avoidance method for UAVs using value iteration

WEI Zhi-qiang, AN Xin

Abstract: A Markov decision process (MDP) optimization model based on value iteration was proposed for the needs of autonomous conflict resolution of unmanned aerial vehicles. A value iteration-based dynamic collision avoidance model was first constructed to achieve real-time safe collision avoidance. To address the complexity and uncertainty of airspace, a refined state space was formulated, incorporating parameters such as relative altitude between two aircraft, vertical speeds of ownship and intruder, historical actions, and time. A multi-factor dynamic cost function was designed to integrate conflict risk and time to closest approach for action judgement, thereby reducing unnecessary maneuvers during collision avoidance. An adaptive two-layer probabilistic fusion mechanism was introduced to address the vulnerability of traditional deterministic decision-making in complex dynamic environments and improve decision robustness. The results indicate that the proposed dynamic collision avoidance method can achieve safe collision avoidance in three conflict scenarios considering only dynamic intruders, and the final vertical relative heights between two aircraft are 152.5, 188.0, and 143.7 m, respectively. In the mixed conflict scenario considering both static obstacles and dynamic intruders, the minimum vertical relative height between the ownship and static obstacles is 174.7 m, and the vertical relative height between two aircraft is 230.7 m, which ensures the safe flight of unmanned aerial vehicle. Compared with the dynamic window approach (DWA) method, after the ownship executes the collision avoidance strategy based on value iteration in four scenarios, the average excessive altitude adjustment is reduced by 62.4%, and the average number of unnecessary action switches is reduced by 88%. It is indicated that the proposed dynamic programming method based on value iteration is feasible to solve the Markov decision process problem in collision avoidance scenarios, and the unmanned aerial vehicle can achieve safe collision avoidance.More>

2026, 26(3): 215-227. doi: 10.19818/j.cnki.1671-1637.2026.153

FullText HTML (56) PDF (3196KB) (23)

Low-altitude planar air route network planning method based on gradient optimization

LI Jie, SHEN Di, YU Fu-ping, GUO Yi-duo

Abstract: To address the problems of imperfect structure and immature planning method of low-altitude air route networks, the structure of low-altitude air route networks was designed and an innovative planning technology was proposed. In terms of structure, an overall "three-layer airspace" framework was constructed (the bottom layer is the logistics air route network, the middle layer is the commuter air route network, and the upper layer is the emergency/public air route network). A single-layer bidirectional air route design was adopted for the internal structure, with intersections referring to the overpass model to achieve omnidirectional non-waiting traffic. In terms of planning method, an innovative two-stage technology based on gradient optimization was proposed. In the global planning stage, a path planning algorithm was used to generate routes, the "scanning volume" method was adopted to identify turning points and intersection points, and density-based spatial clustering of applications with noise (DBSCAN) was employed to merge them to form an initial network. In the local optimization stage, the problem was transformed into the crossing waypoints location problem (CWLP), 9 movement directions were set to construct a direction matrix, and the total length of the air route network, the total length of routes, and safety constraints were taken as the objective function. The position of intersection points was iteratively optimized by dynamically updating the transition probability. With a certain area in Shanghai as the simulation scenario and 150 m as the standard for defining high-rise buildings, 12 take-off and landing points and 34 routes were set for verification. The results show that a conflict-free low-altitude air route network is successfully planned, with the number of iterations reduced by 66% compared with existing methods. 85.3% of the route length change rates are between -10% and 2%, with no excessive extension. The node layout is regular. The airspace occupancy rate and conflict risk are significantly reduced. The structural framework and planning technology provide a feasible scheme for the practical construction of urban low-altitude air route networks. The planning research of trunk air route networks can be expanded in the future.More>

2026, 26(3): 228-243. doi: 10.19818/j.cnki.1671-1637.2026.154

FullText HTML (70) PDF (24093KB) (23)

Construction method of urban low-altitude risk map based on block 3D Gaussian splatting

ZHANG Jian-ping, WANG Zhi-yuan, ZHANG Guang-yuan, LUO Chuang, CHEN Yun-xiang

Abstract: An urban low-altitude three-dimensional (3D) risk mapping method was proposed to achieve unified modelling and representation of multiple risk factors, including terrain, human, and meteorological factors. A large-area unmanned aerial vehicle (UAV) image reconstruction approach based on block 3D Gaussian Splatting (3DGS) was investigated to ensure geometric and photometric continuity. A digital surface model (DSM), a digital elevation model (DEM), and morphometric indices were derived. Population density, layered wind speed, wind shear, and terrain factors were then aligned within a unified voxel grid. A 3D risk field was established using logarithmic linear pooling and quality-map weighted fusion, and risk levels were formed using quantile-based thresholds. Two-dimensional (2D) heat maps and 3D volume-rendering visualizations were generated. Experiments were conducted in a 15 km² area of a new district in Chengdu. Research results indicate that the block-consistency reconstruction strategy effectively reduces reconstruction seams and geometric steps, increases the peak signal-to-noise ratio (PSNR) from 27.8 dB to 28.6 dB, and produces smoother distributions of the derived terrain indices. The complementary effects of terrain, human, and meteorological factors are evident, with a reasonable fused risk distribution. The probability that the core area risk exceeds the acceptable threshold is 7.4%, which is significantly higher than 1.5% in the peripheral area. Both 2D slices and 3D volume rendering show spatial connectivity. The comprehensive risks of the urban core area at 30, 80, 120, 200, and 300 m are 0.68, 0.61, 0.53, 0.56, and 0.57, respectively. Ablation experiments further confirm the importance of modules such as inter-block consistency, curvature features, scene masks, wind shear, and probability calibration.More>

2026, 26(3): 244-260. doi: 10.19818/j.cnki.1671-1637.2026.155

FullText HTML (61) PDF (8413KB) (31)

Small object detection for UAV using bio-inspired spiking neural network

WANG Si-qi, LIU Jiang, SRIGRAROM Sutthiphong, XIANG Cheng, KHOO Boo Cheong, CAI Bai-gen

Abstract: To address the problem that small object detection methods for unmanned aerial vehicles struggle to achieve an effective trade-off between detection accuracy and computational complexity, which hinders practical deployment, a SpikeSOD model based on the bio-inspired spiking neural network (SNN) framework was proposed for small object detection of unmanned aerial vehicles. The model scheme was improved based on the YOLOv8 object detection model. A leaky integrate and multi fire (LIMF) neuron inspired by the multi-synaptic structure of biological neurons was introduced to reduce the spike quantization error and alleviate the small object information loss problem exacerbated by the sparsity of the SNN model. A lightweight spiking feature enhancement module inspired by the lateral inhibition mechanism of biological visual neurons was used to enhance the perception ability of the backbone network for the local information of small objects and their surrounding environment. A spiking feature fusion module was adopted to enhance the multi-scale feature fusion and complementary representation abilities of the neck network. The proposed model was validated on the unmanned aerial vehicle dataset VisDrone-DET2019. The results indicate that compared with the baseline YOLOv8n model, the proposed SpikeSOD model increases the mean average precision by 22.7%, reduces the parameter amount by 16.7%, and decreases the energy consumption by 12.3 mJ, among which the improvement in small object detection performance is particularly significant. The designed LIMF neuron increases the mean average precision by 8.8% compared with the most competitive neuron and effectively alleviates the limitations of traditional SNN models in small object information processing. Compared with existing object detection models, the SpikeSOD model achieves an effective balance among three key indicators, i.e., detection accuracy, lightweight design, and low power consumption, and has significant feasibility and application potential for practical deployment on unmanned aerial vehicle platforms.More>

2026, 26(3): 261-275. doi: 10.19818/j.cnki.1671-1637.2026.156

FullText HTML (97) PDF (3357KB) (26)

Reliability-oriented unmanned aerial vehicle nest location optimization method for smart city management inspection

GAO Feng, YU Bin

Abstract: To enable automatic inspection by unmanned aerial vehicles for urban management events based on fixed nests and to mitigate the impact of nest and unmanned aerial vehicle failures on efficiency and stability, the reliability-oriented fixed nest location-allocation problem with multi-level backup mechanisms was investigated. The differentiated inspection frequencies of task points and service radius constraints of fixed nests were considered, and a mixed integer programming model was formulated with the objective of minimizing the total cost of nest construction and operation. A hybrid algorithm based on Lagrangian relaxation was proposed. The original problem was decomposed into two subproblems, i.e., nest location and multi-level task allocation, by relaxing the location-allocation coupling constraints, and they were solved exactly to obtain a tight lower bound. A coverage gain-driven location repair algorithm was designed to generate feasible upper bounds. An upper bound improvement algorithm based on neighborhood search was proposed to accelerate convergence. Research results show that, for small-scale and medium-scale instances, the proposed algorithm reduces computation time by 57.56%–88.86% compared with Gurobi, while producing high-quality solutions for large-scale cases within short runtimes. Multi-level redundancy significantly reduces system costs. In the case of Zhongshan District of Dalian, the three-level redundancy configuration reduces the total cost from 723 600 CNY to 437 200 CNY, a reduction of approximately 39.59%. The marginal benefits of configuring nests with more than three levels of redundancy diminish significantly. As the nest service radius increases, total and construction costs decline and then stabilize, with inspection costs remaining nearly unchanged. Unit cost of nest procurement is positively correlated with total, construction, inspection, and manual inspection costs and negatively correlated with the number of nests. Unmanned aerial vehicle unit flight cost shows a near-linear positive correlation with total, construction, and inspection costs but has no significant impact on manual inspection cost.More>

2026, 26(3): 276-290. doi: 10.19818/j.cnki.1671-1637.2026.157

FullText HTML (75) PDF (7429KB) (26)

Rotated object detection using low-altitude UAVs for open-pit mines with adaptive fine-tuning

GAO Ming, CHEN Xin, JIANG Shuo, HU Man-jiang, QIN Hong-mao, BIAN You-gang

Abstract: To achieve full-scene real-time visual perception for open-pit mines in low-altitude three-dimensional transportation systems, an adaptive fine-tuning detection method (AFTDet) was proposed for unmanned aerial vehicle (UAV)-based rotated object detection. To address the significant pose variation of mining vehicles observed from UAV perspectives, an adaptive spatial regression loss function was designed to optimize angle learning and improve the rotated bounding box regression accuracy for high-aspect-ratio targets. A fine-tuned non-maximum suppression algorithm was proposed to leverage spatial information from overlapping detection boxes and enhance prediction accuracy through weighted fusion of localization parameter differences. The open-pit mine rotated object detection dataset (MineR) was constructed, comprising 4 540 rotated annotated samples covering passenger vehicles, small excavators, loaders, and dump trucks. AFTDet was validated on both the public remote sensing dataset DOTAv1.0 and the self-built MineR dataset. The results demonstrate that AFTDet achieves 78.61% AP₅₀ and 55.45% AP₇₅ on DOTAv1.0, representing improvements of 0.47% and 1.80% respectively over the baseline model RTMDet-R-m. On the MineR dataset, it achieves 76.25% AP₅₀ and 44.38% AP₇₅, with improvements of 1.06% and 3.62% over the baseline. Ablation studies indicate that the adaptive label assignment strategy improves AP₅₀ by 0.99% and AP₇₅ by 2.50%, while the fine-tuned non-maximum suppression further improves AP₇₅ by 1.09%. The detection speed reaches 50.5 frames·s^-1 with parameters maintained at 2.467×10⁷. The adaptive fine-tuning detection method significantly enhances pose estimation performance for rotated objects, particularly improving the detection recall of large-aspect-ratio mining vehicles, providing effective technical support for UAV visual perception in low-altitude three-dimensional transportation systems while maintaining real-time detection capabilities. In addition, it promotes the development of intelligent monitoring and scheduling systems for open-pit mines.More>

2026, 26(3): 291-302. doi: 10.19818/j.cnki.1671-1637.2026.158

FullText HTML (44) PDF (26664KB) (29)

Collaborative control method for multi-UAV search trajectory in high-rise building emergency rescue

CHEN De-qi, ZHANG Zi-she, ZHANG Wen-hui, WANG Xian-bin

Abstract: To address the issues of low learning efficiency and poor strategy robustness in multi-UAV systems during their collaborative emergency search task for high-rise buildings, caused by insufficient experience in close-range inter-agent collision and formation reconfiguration, a multi-agent deep deterministic policy gradient model integrated with prioritized experience replay (PER-MADDPG) was proposed. A UAV swarm simulation environment with a six-DOF dynamic model was established. The multi-UAV collaborative search task was formulated as a multi-agent Markov decision process, and a hierarchical reward function integrating individual trajectory tracking, energy constraints, team formation keeping, and collision avoidance requirements was designed. A centralized critic network was used to calculate the temporal difference errors of the team's joint actions. The joint experiences were quantified, and the prioritized sampling was implemented. The algorithm was guided to focus on high-value sparse collaborative samples. As a result, the convergence to robust collaborative policies was accelerated. Experimental results show that a 98% task success rate is achieved by the PER-MADDPG algorithm, 15.3% higher than the baseline MADDPG algorithm, and the inter-agent collision rate is reduced from 8% to 1%. In terms of collaboration and control accuracy, the average formation error is decreased from 0.07 m to 0.03 m, and the average trajectory tracking error is lowered from 0.12 m to 0.05 m. In the scalability tests on four-UAV and six-UAV formations, the performance degradation caused by physical space congestion is effectively overcome, demonstrating superior robustness to baseline algorithms. The established PER-MADDPG can effectively balance individual control accuracy and team collaboration stability, enhancing search efficiency in high-rise building emergency rescue.More>

2026, 26(3): 303-316. doi: 10.19818/j.cnki.1671-1637.2026.159

FullText HTML (57) PDF (7383KB) (23)

User Center

Journal Info

Index byMore>

Current Issue
2026 Vol. 26, No. 3

Search Type

Search Journal

LinksMore>