331 results about "Dynamic vehicle" patented technology

The invention relates to the field of battery state evaluation, in particular to a new energy automobile battery health state evaluation method and system. The method comprises the following steps: acquiring a vehicle battery multi-mode state monitoring parameter and a historical vehicle endurance mileage log; performing battery state dynamic evolution on the vehicle battery multi-mode state monitoring parameters to obtain battery multi-mode state evolution characteristics; carrying out multi-time-point maximum endurance calculation on the historical vehicle endurance mileage log, carrying out time sequence trend evolution, and constructing a vehicle endurance evolution trend chart; performing nonlinear attenuation trend analysis on the vehicle endurance evolution trend graph, and performing battery aging mechanism evolution analysis so as to construct a battery state aging evolution knowledge graph; according to the battery state aging evolution knowledge graph, environment change influence analysis is carried out on the vehicle endurance evolution trend graph, dynamic vehicle state modeling is carried out, and a dynamic vehicle twinborn model is constructed. According to the invention, efficient and accurate battery health state evaluation is realized.

The invention relates to the technical field of data processing, in particular to a data model processing method and system. The method comprises the following steps: acquiring multi-modal vehicle-mounted sensing data, and performing vehicle environment sensing to obtain dynamic vehicle environment sensing data; behavior pattern modeling is carried out through a vehicle-mounted sensing network, and a driver behavior prediction model is obtained; performing driving scene risk assessment based on the driver behavior prediction model and the dynamic vehicle environment perception data to obtain driving scene risk assessment data; performing vehicle multi-level early warning strategy analysis according to the driving scene risk assessment data to obtain a vehicle early warning instruction set; performing driver behavior prediction through a vehicle-mounted sensing network to obtain real-time early-warning driver behavior prediction data; and carrying out vehicle emergency intervention measure automatic decision making on the real-time driving early warning scene simulation data to obtain an emergency intervention decision execution instruction set. According to the invention, the driving safety and the response efficiency can be improved.

Systems and methods are provided herein for implementing dynamic, scalable platoon formations for traveling vehicles. Such platoons may be formed and modified on the fly by vehicles that have varying sensing or autonomy capabilities. Platoons may be negotiated among available vehicles, to designate roles including global platoon leader, local leader(s), and platoon member(s). While the platoons are in place and the vehicles traveling in coordination, platoon members may depart the platoon, and the remaining members may re-allocate roles among the remaining members.

The invention discloses a dynamic vehicle scheduling intelligent decision-making system based on big data, and relates to the technical field of intelligent traffic and logistics scheduling, and the system comprises a data collection module, an event analysis module, a knowledge graph construction module, a hypergraph modeling module, a constraint processing engine, an optimization decision-making module, a strategy verification module and an output interaction module. The system has the advantages that multi-source data such as government announcement texts and social media information are acquired in real time through the data acquisition module, event key information is extracted through the event analysis module by utilizing a natural language processing technology, and an event knowledge graph of an incidence relation is constructed through the knowledge graph construction module; the strategy verification module simulates and verifies the strategy effect through the digital twinning technology and iteratively optimizes the strategy effect, the whole process does not need manual intervention to adjust rules, the limitation that a traditional system depends on manual processing is broken through, and the problems that dynamic strategy adjustment is time-consuming, labor-consuming and error-prone in an extreme scene are solved.

The invention relates to the field of sensor activation, in particular to a tire pressure sensor activation method and device, equipment and a storage medium. The method comprises the following steps: acquiring vehicle running state monitoring parameters and vehicle driving monitoring images in real time based on a multi-sensor array; vehicle vibration time-frequency analysis is carried out on the vehicle running state monitoring parameters, vehicle vibration situation mining is carried out, and a vehicle vibration situation curve is constructed; road object recognition is carried out on the vehicle driving monitoring image based on the vehicle vibration situation curve, real-time road condition scene structure rendering is carried out, and a real-time road condition scene model is constructed; obtaining a historical behavior log of the vehicle; carrying out multi-time-point vehicle position calculation on the vehicle historical behavior log, and carrying out dynamic vehicle movement track reconstruction so as to construct a historical vehicle dynamic track; and performing dynamic behavior feature analysis on the historical vehicle dynamic trajectory, and constructing a personalized driving portrait of the vehicle owner. According to the invention, efficient and rapid tire pressure sensor activation is realized.

The invention relates to the technical field of intelligent traffic supervision, and discloses a vehicle tarpaulin coverage real-time detection method based on multi-modal feature fusion, and the method comprises the following steps: S1, constructing a bimodal input data stream; s2, the bimodal images are preprocessed respectively; s3, extracting features; s4, dynamically fusing the bimodal features through an adaptive feature fusion module; s5, inputting the mixed feature map into a lightweight convolution detection network, and outputting a segmentation mask; s6, motion trail compensation is carried out on the dynamic vehicle; by designing a multi-modal adaptive feature fusion mechanism, the robust detection performance in a complex environment is improved: the fusion weight is dynamically adjusted based on the environment illumination and the temperature gradient, so that the system automatically strengthens effective modal features under extreme conditions such as strong light, night, rain and fog and the like; by introducing a dynamic motion compensation framework, the problem of motion fuzzy interference of a running vehicle is effectively solved, and the boundary precision of dynamic vehicle tarpaulin coverage detection is improved in a breakthrough manner.

The invention relates to the technical field of sensor upgrading, in particular to a remote upgrading method and system for a tire pressure sensor. The method comprises the following steps: acquiring real-time state data of a tire pressure sensor; carrying out outlier parameter elimination on the real-time state data of the tire pressure sensor, and carrying out state diagnosis mining, so as to generate a state diagnosis report of the tire pressure sensor; performing upgrade demand prediction based on the tire pressure sensor state diagnosis report, and performing priority upgrade analysis to extract a plurality of upgradable sensors; performing byte-by-byte difference identification on the plurality of upgradable sensors based on preset latest version information, and performing incremental data encapsulation to generate a plurality of incremental upgrading data blocks; acquiring a historical running log of the vehicle; performing dynamic vehicle using behavior analysis and vehicle idle time period prediction on the vehicle historical operation log to generate a plurality of idle prediction time periods; the upgrading efficiency and the safety performance of the tire pressure sensor are improved.

A method for a shared control, dynamic driving system is described. The method includes determining a vehicle command requested by a vehicle operator of an ego vehicle. The method also includes predicting the ego vehicle entering an unstable, controllable operating range if the vehicle command is performed. The method further includes adjusting the vehicle command to maintain control of the ego vehicle in the unstable, controllable operating range. The method also includes performing an adjusted vehicle command to operate the ego vehicle in the unstable, controllable operating range while maintaining recovery stability to transition to a safe operating range.

The invention relates to the technical field of solid waste sorting scheduling, in particular to a real-time monitoring optimization method and device for a solid waste collection, transportation and sorting ERP system. The method comprises the following steps: obtaining a real-time solid waste sorting log based on a solid waste collection, transportation and sorting ERP system; solid waste carrying vehicle tracking is carried out on the real-time solid waste sorting log, and a carrying track distribution visual view is constructed; acquiring an omnibearing monitoring image of a real-time sorting site; equipment sorting behavior visual identification is carried out on the real-time sorting site omnibearing monitoring image, dynamic sorting evolution is carried out, and a dynamic sorting twinborn model is constructed; carrying out multi-time-point solid waste sorting simulation on the dynamic sorting twinborn model, carrying out sorting load demand prediction, and generating a sorting load demand prediction value; and carrying out vehicle load calculation on the sorting load demand prediction value one by one, and then carrying out dynamic vehicle scheduling decision making to obtain a dynamic vehicle scheduling strategy. According to the invention, efficient and real-time solid waste sorting scheduling optimization is realized.

The invention provides a road vehicle real-time twinning method and system based on a monocular camera, which can sense the change of a multi-target vehicle in a dynamic scene in real time by combining target detection, target tracking and depth estimation technologies, and accurately map the position and the feature of the multi-target vehicle to a virtual twinning scene, so that the real-time twinning of the multi-target vehicle is realized. And the defect of dynamic vehicle twinning support in the existing road twinning scene is overcome. The vehicle features are extracted by adopting a visual question and answer large model, dependence on specific task data and scenes is reduced, generative reasoning and multi-round interaction are supported, problems can be dynamically adjusted, and a new scene can be quickly adapted.

The invention discloses a dynamic vehicle path optimization method based on online deep reinforcement learning. The method mainly solves the problems that an existing dynamic vehicle path optimization method is poor in dynamic change adaptive capacity and poor in algorithm real-time performance. The implementation scheme is as follows: 1) generating a data set; 2) constructing a decision model; 3) constructing and training a strategy network model and a value function estimation network model; and 4) solving a dynamic vehicle path problem with a time window by using the trained strategy network model. According to the method, a partially observable Markov decision model is established according to a time sequence of delivery clients of a motorcade, and a corresponding strategy network and a value function estimation network are designed to optimize a vehicle path. The strategy model obtained through training can quickly give the dynamic vehicle optimal path, and can adjust the planned path on line according to the dynamic change of related conditions, so that the solution quality is ensured, the solution speed is higher, and the real-time performance of the algorithm in processing the dynamic vehicle path optimization problem is improved.

The invention discloses an urban road collapse risk calculation method under an extreme flood condition, and the method comprises the following steps: S1, determining the damage condition of an underground structure, and determining the simulation parameters of a road, a soil body, underground water and a flood scene; s2, establishing a three-dimensional road structure model based on DEM-CFD bidirectional coupling according to the road structure and the particle flow microscopic parameters and hydrodynamic parameters of the soil body, and simulating a cavity development process caused by a damaged underground structure under a long-time extreme flood condition; and S3, according to the simulation result, applying a dynamic vehicle load and an FWD load, and evaluating the road collapse risk and traffic capacity. The method simulates the interaction of the road structure and the underground structure under the whole flood process, can recognize the potential collapse risk of the urban road under the flood condition, evaluates the traffic capacity of the urban road after flood, improves the safety of the urban road, can more accurately give the traffic control opinion of the urban road after flood disaster, and improves the traffic control effect of the urban road after flood disaster. And a certain reference is provided for post-disaster road traffic management and repair.

The present invention relates to a pan-vehicle-road collaboration method and system based on roadside perception and visual guidance, and relates to the field of intelligent transportation technology. The method includes: a roadside holographic perception device obtains the current position data information and current speed information of the target vehicle; the current position data information and current speed information are imported into a constructed vehicle formation driving model, and a dynamic vehicle distance parameter database to obtain the ideal speed information; based on the ideal speed information and current speed information of the target vehicle, visual guidance information is sent to the target vehicle to guide the target vehicle to adjust the speed to the ideal speed. The above-mentioned use of the acquired vehicle current speed, position, distance and other information, with the formation driving model as the core control logic, guides the vehicle to travel in an orderly manner by transmitting optical information, realizes the refinement of key channel scenes and the adaptive controllable characteristics, and improves the stability and safety of road traffic.

The invention belongs to the technical field of communication, and particularly relates to a heterogeneous vehicle edge computing resource allocation method based on deep reinforcement learning, which comprises the following steps: constructing a heterogeneous VEC system model; establishing an optimization problem model by taking minimization of system utility under resource constraint and URLLC constraint as a target; modeling an unloading and resource allocation process as a DRL process, and constructing a deep reinforcement learning framework to obtain a state space, an action space and a return function; based on a deep reinforcement learning framework, optimizing a resource allocation strategy through an improved SAC algorithm to obtain an optimal communication resource and calculation resource allocation proportion; according to the method, the URLLC requirement is met, meanwhile, the system utility is remarkably reduced, the method is suitable for a dynamic vehicle environment, and an effective solution is provided for heterogeneous vehicle edge computing resource allocation.

The invention discloses an Internet of Vehicles low-delay federated learning method based on semi-asynchronous communication, and the method comprises the steps: constructing a vehicle-road cloud cooperative federated learning framework and a system performance model, building a federated learning training time delay optimization model, minimizing the overall training time delay, and meeting the vehicle energy consumption upper limit and global model precision constraint conditions; a multi-dimensional priority dynamic vehicle selection mechanism is realized, and probability sampling is carried out by using a roulette selection algorithm; a vehicle-RSU federated learning strategy based on semi-asynchronous communication and knowledge distillation optimizes the problem of lagging behind and the problem of model obsolessness; a cloud model aggregation opportunity is optimized based on a model difference metric and an adaptive cloud aggregation strategy of a dynamic threshold trigger mechanism. According to the method, collaborative optimization is carried out on multiple levels of equipment selection, intra-layer communication aggregation, old model processing, inter-layer aggregation triggering and the like, so that the overall training time delay of the Internet of Vehicles federated learning system can be effectively reduced on the premise of satisfying model precision and equipment energy consumption constraints.

The invention discloses a method for automatically measuring the driving sight distance of a highway curve. The method comprises the following steps: S1, acquiring three-dimensional space data and vehicle motion parameters of a curve area through a multi-modal sensor; the method comprises the steps of S1, obtaining three-dimensional space data and vehicle motion parameters, S2, carrying out multi-source data fusion on the three-dimensional space data and the vehicle motion parameters, S3, constructing a driving safety sight distance calculation model based on a dynamic vehicle kinematics model, and S4, dynamically correcting a safety sight distance calculation result according to real-time environment parameters. In the frequency domain, double-domain filtering is guided through attention to effectively suppress interference, the purity and accuracy of signals are improved, compared with a traditional single-domain processing method, the double-domain filtering mode can more accurately remove signal interference generated by dynamic interference sources such as construction marks, the reliability of subsequent data processing and decision making is ensured, and the method is suitable for popularization and application. More accurate data support is provided for curve identification and sight distance measurement, and the robustness of the system in a complex environment is improved.

The present invention discloses an active collision avoidance method and vehicle-mounted electronic equipment based on a multi-level risk fusion potential field, comprising the following steps: 1: obtaining information about the vehicle and the surrounding environment; 2: predicting the driving trajectory of the vehicle and other vehicles; 3: judging the positional relationship between the trajectories of the vehicle and other vehicles; 4: if the trajectories intersect, calculating the intersection point in a fixed coordinate system; if the trajectories are parallel and there is a possibility of collision, calculating the rear-end collision point; 5: designing and constructing a risk domain of lane boundaries, static obstacles, dynamic vehicles, estimated trajectory intersection points and collision points in the surrounding environment; 6: performing real-time curvature constraints on the collision avoidance path, and planning a virtual force range within a small time period; 7: determining the optimal virtual force within the current time period based on the principle of low driving risk, and planning a real-time optimal local path; 8: obtaining continuous collision avoidance trajectories based on the optimal virtual forces planned in real time in different time periods; 9: repeating the above steps until the vehicle reaches a safe position.

The invention discloses a logistics vehicle scheduling method and system, electronic equipment and a storage medium, and relates to the technical field of intelligent logistics, and the method comprises the steps: in a cloud center layer, an initial vehicle scheduling scheme is generated based on a dynamic vehicle scheduling model, and the dynamic vehicle scheduling model aims at minimizing the transportation cost; and the edge node layer receives the real-time traffic data and the order information, and dynamically adjusts the initial vehicle scheduling scheme to generate a dynamic scheduling scheme. And the vehicle-mounted terminal layer receives the dynamic scheduling scheme and performs path optimization based on the sensor data to obtain the execution state of the vehicle-mounted terminal. The execution state of the vehicle-mounted terminal is fed back to the edge node layer and the cloud center layer to update the dynamic vehicle scheduling model, a target vehicle scheduling scheme is determined based on the updated dynamic vehicle scheduling model, and the target vehicle scheduling scheme is used for scheduling multiple logistics vehicles. Through cloud side-end cooperative processing, the decision-making quality of logistics vehicle scheduling and the system response speed are improved.

The invention discloses an expressway dynamic vehicle behavior trajectory tracking induction monitoring system, which relates to the technical field of intelligent traffic monitoring, and comprises a multi-source data acquisition module used for responding to a trigger signal of a vehicle entering a monitoring area, acquiring multi-source data, and sending the multi-source data to the monitoring area; the multi-source data comprises three-dimensional point cloud data, a thermal infrared image, a multi-view visible light image, vehicle speed and lane data and a license plate image; and the preprocessing and calibration module is used for registering and fusing the thermal infrared image and the visible light image. According to the expressway dynamic vehicle behavior trajectory tracking induction monitoring system provided by the invention, through multi-source sensor fusion and intelligent algorithm innovation, high-precision tracking and real-time monitoring of vehicle behavior trajectories are realized, road safety and traffic efficiency are effectively improved, and the expressway dynamic vehicle behavior trajectory tracking induction monitoring system is suitable for popularization and application. Multi-dimensional sensors such as a laser radar, an infrared thermal inductance instrument and a visible light camera are adopted for cooperative work, and the capability of capturing vehicle information in a weak light environment is enhanced.

A vehicle includes a chassis, a vehicle body mounted to the chassis, a display device within the vehicle body, and a control system coupled to the chassis and the display device. The control system is configured to detect a change to at least one of the chassis or the vehicle body. The control system is further configured to generate a three-dimensional scene including a representation of a three-dimensional model the vehicle, the three-dimensional scene corresponding to the change. The control system renders the three-dimensional scene to obtain an image and outputs the image on the display device. The image may be rendered using cel shading. The three-dimensional scene may further correspond to an environment of the vehicle such as weather, landscape, and time of day.

The invention provides an intelligent optimization method and system for a parking lot system. Wherein parking space pressure change data is collected through a parking space pressure sensor, vehicle size data of a vehicle identification system is synchronized, and a three-dimensional thermodynamic diagram containing parking demands is constructed based on spatial distribution; the stand column displacement sensing array monitors the height of a vehicle chassis, a dynamic vehicle size database is constructed by combining size data, and parking space physical constraints are matched; in cooperation with the parking demand and the constraint condition, screening and matching an idle parking space set according to the size of the entering vehicle and the thermodynamic diagram prediction trend; and a vehicle guiding path is generated by combining the incidence relation between the historical parking path and the thermodynamic diagram and integrating the parking space position and the path occupation state, so that intelligent optimization of the parking lot is realized. According to the technical scheme provided by the invention, precise parking space matching and dynamic path guiding of vehicle size self-adaption are realized, and the turnover efficiency and the space utilization rate of the parking lot are improved.

The invention discloses a traffic signal optimization method and system based on artificial intelligence, and the method comprises the steps: collecting vehicle microscopic behavior data, and obtaining a driving intention prediction result; according to the driving intention prediction result, constructing a dynamic vehicle motion track; and completing the optimization of the intersection traffic signal by using the constructed dynamic vehicle motion trail. The traffic signal response speed and accuracy can be effectively improved, the traffic jam is reduced, and the traffic efficiency is improved. By adopting the convolutional neural network and the long and short term memory network, the system can automatically extract and analyze the driving behavior characteristics and mine the driving intention information, thereby optimizing the duration and phase sequence of the signal lamp and ensuring the smoothness of the traffic flow. Besides, the dynamic planning method based on the Markov process model enables the signal optimization decision to be more scientific and reasonable, can adapt to the change of traffic flow in real time, and further improves the intelligent level of traffic management.

The invention discloses a vehicle speed prediction method and system in a dynamic vehicle following scene considering a traffic environment, and the method comprises the steps: extracting vehicle following case data through setting a standard and a threshold value of a vehicle following process in a natural driving data set; using a target detection algorithm to obtain traffic environment characteristics related to the vehicle following case data, extracting vehicle dynamics parameters in the vehicle following case data, fusing the vehicle dynamics parameters and the traffic environment characteristics, and performing normalization processing to obtain fusion characteristics; removing redundant variables in the fused features by adopting a combined feature selection method; and dividing the fusion features from which the redundant variables are removed into a data set, training a pre-established Informer-FDR vehicle speed prediction model, and obtaining a vehicle speed prediction result by using the trained vehicle speed prediction model. The method effectively reflects the dynamic interaction characteristics of the driver and the traffic environment and potential traffic conflicts, enhances the understanding of the model for the complex driving environment, and improves the real-time performance and accuracy of vehicle speed prediction.

The invention discloses an Internet of Vehicles multi-hop unloading optimization method and system based on a graph neural network, and the method comprises the steps: system initialization and environment modeling, task unloading strategy decision making, multi-hop path construction, link state detection and RIS configuration, and task transmission and execution, the system comprises a path construction module, a link enhancement module and a task unloading scheduling module, multi-target and multi-granularity unloading cooperative scheduling of tasks among local, adjacent vehicles, roadside units and a cloud is supported by sensing node positions and link state changes in the dynamic Internet of Vehicles, and compared with a traditional fixed strategy or single agent method, the task unloading cooperative scheduling efficiency is greatly improved. The method provided by the invention has higher path robustness, stronger link adaptability and better scheduling performance, and is suitable for scenes with high requirements on calculation timeliness and communication stability, such as automatic driving and smart traffic.

The invention relates to the technical field of carbon emission evaluation, and relates to a carbon emission evaluation method and system based on multi-source heterogeneous data, and the method comprises the steps: obtaining the space matching data of a target region boundary and parking spaces, and generating the number and distribution information of effective parking spaces; the time sequence dynamic remote sensing data, the traffic density fluctuation data and the demographic data are fused, a parking space activeness factor model is constructed, and the parking space use intensity is output; predicting the vehicle type distribution proportion of the target area by using a neural network, and calculating the annual average driving distance of the vehicle; calculating a dynamic vehicle carbon emission factor and a static vehicle carbon emission factor according to the time dynamic nature and the regional difference of the power grid energy; and calculating the total carbon emission of the target area by integrating the vehicle type distribution proportion, the annual average driving distance and the carbon emission factor based on a collaborative correction mechanism of the parking space resource tensity and the parking space use intensity. The objective of the invention is to realize community-level high-precision carbon emission accounting and improve the spatial resolution and adaptability of carbon emission accounting.

The invention relates to the technical field of ecological driving, in particular to a self-adaptive ecological cruise method and device based on dynamic vehicle behavior prediction.The method comprises the steps that on the basis of at least one driving parameter of a surrounding vehicle within a vehicle target range, the transverse motion state of the surrounding vehicle is recognized, predicting a dynamic driving track of the surrounding vehicle in the future target time according to the transverse motion state; in combination with the at least one driving parameter and the dynamic driving track, generating a vehicle distance variation between the surrounding vehicles in the future target time; and generating a self-adaptive ecological cruise strategy of the vehicle at the current moment based on the dynamic driving track, the vehicle distance variation and a preset driving constraint condition. Therefore, the problems that the prediction dimension of the ecological cruise control strategy of the new energy automobile is still limited to the longitudinal motion state change, the decision complexity of the new energy automobile as an intelligent driving main body is ignored, and obvious insufficient environmental adaptation is achieved in a dynamic traffic scene in the related technology are solved.

The invention discloses a driving sunscreen cooling control method based on sun illumination tracking, and the main design concept of the method is that the recognition precision of the relation between dynamic light and a dynamic vehicle is improved through sun vector space modeling. Specifically, accurate measurement and calculation of the sun angle is matched with dynamic response of a vehicle system to realize a sunlight tracking type sunscreen mode, and the method mainly comprises the following steps: acquiring data such as longitude and latitude, date time, advancing direction and the like of a current vehicle, calculating an irradiation angle of sunlight irradiated to the vehicle, and combining a real-time measurement value of a light sensor to obtain a sunlight tracking type sunscreen mode. And deciding whether the vehicle enters the sunscreen mode at the current moment. On the basis of real-time tracking of the sun position and by means of intelligent cooperative control of devices such as an electric sun shield curtain, electrochromic glass and air conditioner directional blowing, dynamic and accurate automatic non-inductive sun protection and cooling are achieved, and comfortable driving and riding experience is brought to drivers and passengers.

The invention discloses a configurable dynamic vehicle-mounted dimming method and system for tunnel lamp illumination, and the method comprises the following steps: dividing a target tunnel into a plurality of dimming sections, further dividing a dimming loop in each dimming section, and carrying out the deployment of an illumination unit and an incoming vehicle detector based on a division result; a dimming counting interval of a logic level is constructed, a dynamic mapping relation is established between the dimming counting interval and a dimming loop of a physical level, and the dimming counting interval is a minimum logic unit used for vehicle tracking and counting; an independent vehicle counter is configured for each dimming counting interval, and a counting updating rule is set; configuring a security policy for counting function activation and dimming execution; and responding to the triggering of the coming vehicle detector, updating the vehicle count in the dimming counting interval in real time, and adjusting the brightness of the dimming loop based on the security policy. Compared with the prior art, the configurable dynamic vehicle-mounted dimming method and the configurable dynamic vehicle-mounted dimming system have remarkable progress in the aspects of energy conservation and safety.

The invention relates to the technical field of intelligent mine and vehicle safety, in particular to a vehicle collision prediction method and prediction system for a mine laneway environment. The vehicle collision prediction method for the mine laneway environment comprises the following steps: constructing a static laneway environment map based on a mine digital map; at each time step t, constructing a dynamic vehicle interaction diagram according to the vehicle positioning information; extracting the spatial-temporal characteristics of the dynamic vehicle interaction diagram and the state characteristics of the driver, fusing the spatial-temporal characteristics and the state characteristics to form a comprehensive characteristic vector, and outputting a predicted value of the running speed of the current target vehicle at the next moment through a regression layer; and judging whether a collision risk exists or not according to the deviation between the predicted value of the vehicle speed and the real speed. According to the method, the space-time diagram neural network is constructed, vehicle dynamic interaction, static roadway environment topology and driver state multi-source information are fused, and high-precision and advanced prediction of the collision risk is achieved.

Provided are system, method, and device for automatically manage communications between vehicles. According to embodiments, the system may include: a memory storage storing computer-executable instructions; and at least one processor communicatively coupled to the memory storage, wherein the at least one processor may be configured to execute the instructions to: receive an identification of a target vehicle from a server; obtain an identification of a nearby vehicle from the nearby vehicle; determine whether the identification of the target vehicle matches the identification of the nearby vehicle; and in response to determining that the identification of the target vehicle matches the identification of the nearby vehicle, transmit a current time and location to the server.