31 results about "Traffic model" patented technology

The invention relates to the field of intelligent traffic systems, in particular to an intelligent intersection multi-vehicle cooperative control method. The method comprises the following steps: S1, sensing a vehicle state and sharing vehicle node information by adopting a dual-mode communication mechanism; s2, fusing state data at edge computing nodes; establishing a dynamic traffic model according to the fused state data; s3, acquiring node data of the edge calculation node, calculating a vehicle passing priority index according to the node data, and performing vehicle priority decision according to the passing priority index; s4, performing grading processing on the vehicles with the overlapped tracks; s5, the RSU device broadcasts the control instruction, and if the vehicle-mounted unit receives the control instruction within the set time, an execution report is transmitted back to the RSU device; and if the vehicle-mounted unit does not receive the control instruction within the set time, a redundancy execution mechanism is started to detect and control the running state of the vehicle. According to the invention, millisecond-level synchronous transmission and high-precision perception of information are realized, and the real-time performance and accuracy of traffic environment modeling are improved.

The application provides a power deterministic network routing and packet scheduling method and system, the method comprising: determining the feasible path set from the source to the destination node for each deterministic flow according to the network and traffic model; constructing the three-dimensional scheduling variable consisting of the path, time offset and CSQF period specification set for each flow; jointly solving the three-dimensional variable of each flow to obtain the routing and packet scheduling scheme satisfying the resource, sending, delay and period constraint with the minimum objective function, the objective function being used to realize the double load balancing of the link and period window; and calculating the network remaining resource based on the obtained scheme and allocating the transmission resource for the non-deterministic flow according to the network remaining resource. Through the three-dimensional joint scheduling of the routing, time and period and the double load balancing of the link and window, the application can significantly reduce the jitter under the strict delay constraint, effectively improve the network resource utilization and the reliability of the deterministic service.

The invention discloses a network risk identification method and device and a storage medium, and relates to the technical field of network security. The network risk identification method comprises the following steps: extracting historical traffic characteristics based on collected historical traffic, and constructing a normal traffic model; extracting actual network traffic characteristics based on the collected actual network traffic; obtaining the feature similarity of each network device according to the historical traffic features and the actual network traffic features; setting a feature similarity threshold of each network device, and identifying a specific risk device and a risk type according to the feature similarity of each network device; and fusing the actual network traffic characteristics, which are identified as normal, of the equipment with the normal traffic model, and iteratively identifying the network security risk. According to the method, extraction of hidden features and identification and positioning of unknown attacks can be realized, and dynamic adjustment can be carried out according to newly added normal traffic so as to adapt to dynamic changes of network services.

A system receives GPS data and on-board sensor data from several connected vehicles indicating locations of nearby vehicles and objects. The system processes the data to create a shared-world model that includes locations and velocities of the connected vehicles and nearby vehicles and objects, and the system determines whether driving hazards exist, such as potential collisions. The system may transmit an alert to at least one of the connected vehicles, to cause the connected vehicle or a mobile device to present a warning message to a driver, such as a visual, audio, or haptic message, or to cause the connected vehicle to implement an action to avoid the driving hazard, such as activating emergency braking or altering course. The system may create, and transmit to a connected vehicle or mobile device, a lane-level traffic model indicating traffic density, traffic speed, and traffic throughput.

Imitation learning, or artificial intelligence-based learning from demonstration, aims to acquire an agent policy by observing and mimicking the behavior demonstrated in expert demonstrations. Imitation learning can be used to generate reliable and robust learned policies in a variety of tasks involving sequential decision-making, such as autonomous driving and robotics tasks. However, existing methods that use next-token-prediction (NTP) models, where the policy reduces to a classifier over a discrete set of trajectory tokens, suffer from covariate shift due to their open-loop training a closed-loop execution. The present disclosure provides closed-loop fine tuning of autonomous agent policies in a manner that can mitigate covariate shift.

An intrusion detection method for smart substations based on Hidden Markov Models includes the following steps: S1: Segmenting network traffic according to time scale, extracting packet identifiers, packet measurement data, and packet throughput to construct an intrusion detection dataset; S2: Calculating the difference between state variables and measurement values ​​based on the unscented Kalman filter method to establish a state model capable of calculating the attack detection index; S3: Establishing an ARIMA model, selecting the optimal traffic model, determining the detection confidence interval, and establishing a traffic throughput detection model; S4: Establishing multiple protocol compliance detection rules and, based on the Hamming distance calculation method, establishing a standardized detection model; S5: Real-time detection of network traffic in the smart substation, calculating detection vectors based on the models established in S2, S3, and S4 respectively, and using them as input variables for the Hidden Markov Model, comprehensively analyzing system anomaly characteristics, and achieving final discrimination.

The invention discloses a highway charging load simulation method and system based on multi-period traffic balance. The method comprises the following steps: acquiring highway network topological parameters, charging station distribution parameters, user travel origin and destination demand data and electric vehicle battery parameters; the method comprises the following steps of: aiming at the characteristics of long road section distance and large driving time span of an expressway, establishing a multi-period space-time network model of the expressway, and constructing a space-time topological structure comprising a space-time driving arc and a virtual staying arc; based on a traffic balance principle, constructing a highway multi-period traffic distribution model comprising the main problem and the sub-problems; and solving the highway multi-period traffic distribution model in the step 3 by adopting an iterative algorithm, and mapping the converged space-time path flow into a charging demand so as to obtain space-time load distribution of each charging station of the highway. According to the method, the defect that a traditional single-period traffic model cannot accurately describe the long-distance cross-period driving characteristics and the queuing congestion effect of the expressway can be overcome.

The application discloses a traffic generation device. A traffic model is encapsulated into a first data packet by a software processor, and a routing network table is encapsulated into a second data packet, which are respectively stored into corresponding first and second storage modules by a packet distribution circuit in a hardware processor. A traffic generation circuit parses the traffic model in the first data packet from the first storage module, and independently generates an initial message and corresponding index information. A routing processing circuit quickly obtains corresponding routing entries from the second storage module according to the index information and returns in real time, so that the routing access operation and the data packet generation operation are decoupled and executed in parallel at the hardware level. The routing processing circuit and the traffic generation circuit realize accurate cooperation through the index information, so that the real-time query process of the routing network table no longer blocks the message generation pipeline, thereby effectively solving the problems of message loss and traffic distortion caused by external storage access delay.

Imitation learning, or AI-based demonstration-based learning, aims to derive a policy for an agent by observing and imitating the behavior shown in expert demonstrations. Imitation learning can be used to generate reliable and robust learned policies for a variety of tasks requiring sequential decision-making, such as autonomous driving and robotics. However, existing methods that use next-token prediction (NTP) models, where the policy is reduced to a classifier over a discrete set of trajectory tokens, suffer from covariate shifting due to their open-loop training and closed-loop execution.The present disclosure provides a fine-tuning of the policies of autonomous closed-loop agents in a manner that can mitigate covariate shifting.

The invention discloses an automatic OpenStack stability test method and system, and relates to the technical field of cloud platform stability test.The method comprises the steps that test configuration information including test scene configuration parameters, flow model parameters, expected monitoring indexes and test execution strategies is acquired; initializing a tested system environment and test scene deployment based on the test scene configuration parameters; generating and sending test traffic to the tested system based on the traffic model parameters; in the testing process, actual monitoring index data of the tested system are collected, and when a specified abnormal alarm is triggered, testing is stopped, and a field testing environment state is reserved; comparing the actual monitoring index data with the expected monitoring index, and recording an abnormal log entry according to a comparison result; and generating a test report based on the test process data and the exception log entry. According to the method, the stability evaluation capability and the operation and maintenance response level of the cloud platform in a complex scene are enhanced while the test efficiency and the result credibility are improved.

The invention relates to the technical field of stacking, in particular to a stacking scene unknown unicast flooding suppression method and device and a program product, and the method comprises the steps: determining the exit type of a locally forwarded target message and the type of an effective far-end port; the type of the outlet and the type of the effective far-end port comprise an aggregation port and a non-aggregation port; when it is determined that an outlet of a locally forwarded target message is an aggregation port, whether the flow of the target message exceeds a threshold value is judged, and if yes, a flow model is established based on the flow exceeding the threshold value; if not, deleting the existing traffic model; copying the target message according to the type of the effective far-end port; adding identification information to the copied message based on the traffic model; and sending the message containing the identification information to an effective far-end device, so that the effective far-end device forwards the message according to the identification information. According to the technical scheme provided by the invention, the forwarding performance of the equipment can be improved.

The invention discloses a flow generation device. A flow model is packaged into a first data packet through a software processor, a routing netlist is packaged into a second data packet, and the first data packet and the second data packet are respectively stored to a corresponding first storage module and a corresponding second storage module through a packet distribution circuit in a hardware processor. The traffic generation circuit analyzes the traffic model in the first data packet from the first storage module, and independently generates an initial message and corresponding index information. And the route processing circuit quickly acquires the corresponding route entry from the second storage module according to the index information and returns the route entry in real time, so that decoupling and parallel execution of the route access operation and the data packet generation operation on the hardware level are realized, and the route processing circuit and the flow generation circuit realize accurate collaboration through the index information. The message generation assembly line is not blocked in the real-time query process of the routing netlist, so that the problems of message loss and traffic distortion caused by external storage access delay are effectively solved.

The present application relates to a kind of 5G / B5G power communication network traffic analysis method based on multi-time series data mining, specific method steps include: 1) 5G / B5G power communication network traffic characteristic analysis 2) establish 5G / B5G power communication network traffic model, select multi-time series data mining method to establish traffic model, describe the real situation of 5G / B5G communication network service based on power service.Different from the traditional sense of power communication traffic analysis, new division is carried out to 5G / B5G power communication traffic level, using multi-time series data mining algorithm, 5G / B5G power communication traffic is reasonably statistically described, multiple traffic characteristic time series are analyzed as a whole, produce effective abnormal network traffic characteristic association rules, accurately describe the security situation of entire 5G / B5G power communication network, power communication fault prediction, network design, traffic control, resource management and network design have very important application value.

Disclosed are systems and methods that provide a computerized framework for accurately and efficiently estimating the backhaul bandwidth and core capacity, which can be utilized to automatically and dynamically configure and / or curate network configurations and topologies to handle network loads and traffic. The framework can utilize and / or operate RF system simulators to perform the estimation of each sector's delivered UL / DL throughputs, as well as user inputs related to the traffic model, the dedicated network's topology and product specifications related to the processing limits of BBUs and private cores in order to determine network bottlenecks. Such determinations and / or predictions can be leveraged to ensure the allocation of backhaul bandwidth and core capacity, among other network features and / or characteristics, are optimized to achieve overall high throughput and remove traffic bottlenecks.

The application discloses a kind of road network topology updating method based on GPS data, comprising: S1. the GPS trajectory data of vehicle is collected;S2. all road sections of GPS trajectory data are traversed, and each road section in all road sections is processed as follows: the road section in GPS trajectory data is regarded as target road section ID_re, from GPS trajectory data, the road section sequence (ID_1, ID_re, ID_2) including target road section ID_re is extracted;The topology relationship of the road section sequence including target road section ID_re is judged, and the update of road network topology relationship is carried out according to the judgment result;S3. repeat step S2 until road network topology relationship is no longer updated.The application can accurately update road network topology information in real time, and provides technical support for road traffic statistics, road network simulation, the construction of comprehensive traffic model.

The present application provides a traffic prediction system, a traffic prediction method and a computer readable medium, which efficiently predict the change in required time for a route in a case where the traffic volume of the route changes along with a change in action. A traffic volume data acquisition unit acquires traffic volume data. A required time acquisition unit acquires an actual value of required time. A traffic volume calculation unit calculates the inflow traffic volume of each route using the traffic volume data and the actual value of required time. A traffic state prediction unit calculates, for each route, a predicted value of required time for the calculated inflow traffic volume and a predicted value of required time in a case where the inflow traffic volume changes, using a traffic model generated in advance. A change amount calculation unit calculates, for each route, a change amount of the predicted value of required time in a case where the inflow traffic volume changes.

The application relates to the technical field of network security, and discloses a training method and a detection method of an encrypted traffic detection model and an electronic device. The training method of the encrypted traffic detection model comprises the following steps: training a first plaintext traffic model by using plaintext traffic feature data; training a backup plaintext traffic model by using at least part of encrypted traffic feature data to obtain a first encrypted traffic model; cooperatively training the first plaintext traffic model and the first encrypted traffic model by using the encrypted traffic feature data to obtain a second encrypted traffic model; training an adversarial generative network by using the encrypted traffic feature data; distilling training a teacher model and a student model by using the encrypted traffic feature data, and taking the trained student model as the encrypted traffic detection model. The encrypted traffic detection model trained by the application is lightweight and easy to deploy, and can realize high-precision detection of encrypted traffic.

This invention discloses a method for generating time-dependent business data, relating to the field of communication network simulation. The method includes: synthesizing first variable data following a normal distribution based on sine waves of different frequencies, and synthesizing second variable data following a normal distribution that is orthogonal to the first variable data based on cosine waves; establishing data following a target distribution combination based on the first and second variable data; and selecting target business data from the data following the target distribution combination, the first variable data, and / or the second variable data according to the business scenario to establish a business traffic model. Compared to existing technologies, this invention can generate time-dependent business data for different business types even in the absence of real traffic datasets. It can more accurately and effectively simulate actual user traffic data, adapt to multiple business scenarios, and can be used for tasks such as traffic analysis, business classification, and business prediction, reducing research costs.

Traffic modeling network application services is described. An example system includes a network traffic model that defines traffic relationships for data traffic in a network between a network topology of an application and a plurality of different access points serving network connections through different physical or logical tiers of the network topology with at least one instance of the application based in part on domain-subdomain relationships obtained from a domain name system record model for communicating on the network. A first access point in the system, in response to receiving a first communication between a first endpoint and a first instance of the application, serves the first communication on a first network connection through a first tier of the network topology with the first instance of the application based on a first traffic relationship defined by the network traffic model for the first network connection.

This invention discloses a virtual traffic generator device for network switching units, comprising: a clock-driven scheduler, an XML configuration parsing module, a traffic generation module, a queue management module, and a statistics module. This invention possesses the ability to parse real network data packets and generate random data packets conforming to various network protocol specifications, exhibiting high flexibility. This enables efficient and accurate simulation of various complex network traffic scenarios in ESL simulation environments, improving the realism and effectiveness of simulation testing. Simultaneously, through timestamp management and rate control mechanisms, it can simulate burst traffic and precise bandwidth requirements under specific network scenarios. This invention supports two modes: in realistic stimulus mode, it can reproduce data packets captured in actual networks, suitable for regression testing and specific scenario reproduction; while in generative stimulus mode, users can customize traffic models according to their needs for stress testing, boundary condition testing, etc.

The application discloses a vehicle state perception method and system in a vehicle-road cooperative environment, and relates to the technical field of intelligent transportation, comprising: collecting real-time traffic data of a target intersection; establishing a temporary V2X communication link; receiving real-time traffic data and K real-time vehicle state information of K driving vehicles; if the single-dimensional risk detection results of the real-time traffic data and the K real-time vehicle state information by a roadside edge computing device are all 0, compensating the K real-time vehicle state information with the real-time traffic data to obtain K compensated vehicle state information, constructing a global dynamic traffic model, and predicting trajectory conflicts to locate potential collision risks; according to a risk avoidance priority, broadcasting the potential collision risks to the K driving vehicles through the V2X communication link. The application solves the technical problems of inaccurate traffic modeling and untimely collision risk identification caused by incomplete vehicle-mounted perception information in the prior art, and achieves the technical effects of improving the accuracy of vehicle state perception and the timeliness of collision risk warning.

According to the pressure test method and device, the storage medium and the computer equipment provided by the invention, the pre-created standardized traffic model is introduced, and then traffic recording is performed on the target interface based on the traffic model, so that the recorded data is obtained. And then the performance of each press machine cluster is evaluated, and the optimal press machine cluster is determined in each press machine cluster in combination with the performance index and the recorded data of each press machine cluster. And finally, performing a pressure test on the target interface by using the optimal press cluster according to the recorded data. Therefore, the pressure test can be carried out by using the press cluster which is most matched with the current recorded data, on one hand, the resource utilization can be optimized, and the test efficiency is improved. And on the other hand, the originally dispersed resource evaluation and selection decision process is integrated into a coherent adaptive process, so that the problem of'operation fault 'caused by frequent switching of tools and environments is fundamentally solved, and the efficiency of the pressure test is further improved.

The invention discloses an overall automatic driving method, device and system for a vehicle, and relates to the technical field of intelligent traffic, and the method comprises the steps: constructing a traffic model according to the driving related information of the vehicle, and taking the starting point, destination, driving purpose, passenger / article and time demand of the vehicle as the input, the traffic model is utilized to generate an overall automatic driving scheme of the traffic tool, the overall automatic driving scheme of the traffic tool is sent to all the automatic driving traffic tools in the traffic model, and each automatic driving traffic tool performs automatic driving according to the overall automatic driving scheme of the traffic tool. Therefore, the problems that information acquisition is not comprehensive and decisions of the vehicles conflict with one another can be effectively avoided, and the driving safety and the driving efficiency of the vehicles are greatly improved.

This invention provides a method and system for reconstructing network traffic behavior patterns based on flow matching, belonging to the field of traffic analysis models. It addresses the problems of low generation quality, loss of temporal structure, and slow convergence speed in existing network traffic data processing. This invention, without accessing private training data and model internal parameters, relies solely on limited black-box queries to reconstruct traffic patterns at the burst level that are highly consistent with the private behavior of the victim model. The generation process is a deterministic one-step or few-step update, balancing attack quality and computational efficiency. This provides a practical implementation path for subsequent assessment and protection against model reverse engineering risks in consumer network environments. Compared to existing methods, this invention demonstrates significant advantages in reverse engineering accuracy, structural consistency, and computational efficiency, and can more comprehensively reveal the potential privacy leakage risks faced by deep traffic models deployed in consumer network environments.

The invention discloses an intelligent traffic model credibility verification method based on zero knowledge proof and a reputation mechanism, and the method comprises the steps: recognizing malicious nodes through a dual-reference verification criterion based on cosine similarity and model precision; a lazy trainer is identified through a verification scheme based on zero knowledge proof. The combination of the two mechanisms significantly improves the robustness of the model to the Byzantine attack. Besides, a reputation mechanism is introduced into the block chain system, malicious and lazy trainers can be detected at the same time by using the strategy, and a basis is provided for reward and punishment of the trainers through the reputation mechanism.

The invention relates to the technical field of intelligent transportation, and discloses a method for acquiring a public transportation network, which comprises the following steps: acquiring the number of people getting on and getting off a bus in a plurality of grid areas in a first preset time period; obtaining a first candidate bus stop corresponding to the first preset time period according to the number of passengers getting on the bus and the number of passengers getting off the bus in each grid region in the first preset time period; acquiring first driving time and first travel demand quantity between the first candidate bus stops; according to each first candidate bus stop, each first driving time and each first travel demand, constructing a first traffic model diagram; and inputting the first traffic model diagram into a preset public transportation network acquisition model to obtain a target public transportation network of the preset area. According to the method, the efficiency and practicability of obtaining the public transportation network can be improved. The invention further discloses a device for acquiring the public transportation network, electronic equipment and a storage medium.

The traffic mode recognition method based on fusion of global and local spatiotemporal features comprises the following steps: (1) segmenting the original GPS trajectory data; (2) calculating seven kinematic features of each GPS trajectory point; (3) pre-processing the segmented GPS trajectory data; (4) constructing a traffic mode recognition model; the model comprises a spatiotemporal feature extraction block, a time domain convolution network and an attention mechanism, wherein the spatiotemporal feature extraction block can extract local spatiotemporal features, the time domain convolution network can extract global spatiotemporal features, and the attention mechanism can realize fusion and feature reconstruction of the global and local spatiotemporal features; (5) splitting the data obtained in step (3) to generate a training data set and a test data set, and training the traffic mode recognition model; (6) calculating various evaluation indexes according to the recognition result of step (5); (7) visually displaying the recognition result and the evaluation indexes of the model. The application also comprises a system for implementing the traffic mode recognition method based on fusion of global and local spatiotemporal features. The application improves the traffic model recognition precision.

The invention discloses a method, device, equipment and product for constructing a pressure measurement scene based on artificial intelligence, and the method comprises the steps: predicting the reference pressure measurement flow of a target interface through a flow prediction model based on a target pressure measurement constraint; based on the system design architecture information and / or the interface upstream and downstream dependency information, constructing a calling relation topological graph taking the target interface as a starting point; inputting the calling relation topological graph, the reference pressure measurement flow, the historical flow data of each node and the historical calling frequency data of each edge into a trained graph neural network model; and generating a linkage pressure measurement flow model of the target interface and upstream and downstream interfaces of the target interface based on the reference pressure measurement flow, the weight of each node in the calling relation topological graph output by the graph neural network model and the key calling path. According to the embodiment of the invention, the pressure measurement scene which truly reflects the service calling link can be automatically constructed, and the linkage pressure measurement model which accords with the actual service flow distribution is generated.

The invention discloses a data leakage risk detection method, system and device for abnormal traffic analysis, and the method comprises the steps: firstly obtaining the network traffic data of historical operation data of a data management platform in different time periods, obtaining the traffic parameters of the data management platform, and building a behavior model of normal traffic; monitoring traffic parameters of the data management platform, comparing the traffic parameters with a normal traffic model, and identifying abnormal behaviors; performing risk detection first performance and second performance evaluation analysis on the data of abnormal traffic analysis of the data management platform to obtain a user behavior risk evaluation index of abnormal traffic data leakage and an abnormal traffic real-time analysis evaluation index; according to the abnormal traffic data leakage user behavior risk assessment index and the abnormal traffic real-time analysis assessment index, obtaining an abnormal traffic data leakage comprehensive risk assessment index; and dividing risk levels based on threshold comparison and carrying out early warning. The reliability of the data leakage risk detection method for abnormal traffic analysis is improved.

The invention discloses a pressure measurement method and device based on artificial intelligence, equipment, a storage medium and a product. The method comprises the following steps: automatically triggering pressure measurement decision analysis based on activity planning information or an interface performance state; according to a system design architecture and / or interface upstream and downstream dependency information, a target interface and upstream and downstream interfaces thereof are included in a pressure measurement interface range; constructing a multi-dimensional feature vector of each interface in the pressure measurement interface range, and inputting the multi-dimensional feature vector into a pressure measurement priority ranking model to obtain a pressure measurement interface list ranked according to priorities; constructing a pressure measurement flow model, pressure measurement request data and pressure measurement response data based on the flow prediction model, the graph neural network model and the pressure measurement parameter prediction model according to pressure measurement requirements; on the basis of the pressure measurement flow model, the pressure measurement request data and the pressure measurement response data, pressure measurement is carried out on multiple interfaces in the pressure measurement interface list; and acquiring pressure measurement data of the plurality of interfaces for evaluation to obtain a performance evaluation report. According to the invention, the pressure measurement efficiency and accuracy can be improved.