In-vehicle network channel congestion control method and system based on regional risk level
By calculating vehicle density and hazard level, dividing dangerous vehicle clusters, and setting penalty factors to adjust transmission intervals, the problems of channel congestion and traffic safety hazards in vehicular ad hoc networks are solved, and timely transmission of safety messages and optimization of network load are achieved.
Patent Information
- Application Number
- CN202210639042.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-06-07
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2042-06-07
AI Technical Summary
Existing technologies fail to effectively consider regional hazards in vehicular ad hoc networks, leading to channel congestion and traffic safety risks. In particular, they fail to effectively address the radiating effects of dangerous vehicles on surrounding vehicles in congestion control.
By acquiring vehicle driver behavior and physical performance, vehicle density and hazard level are calculated, dangerous vehicle clusters are divided, penalty factors are set to adjust the safety message transmission interval, and adaptive control is performed using penalty factors HVF and NVF.
It effectively reduces channel congestion, improves traffic safety, ensures timely transmission of safety messages in high-density and dangerous vehicle environments, and avoids network congestion.
Smart Images

Figure CN115038063B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of intelligent transportation systems, and relates to a vehicle network channel congestion control method and system based on regional risk level. Background Art
[0002] With the continuous development of economy and the progress of automobile industry, the concept of Intelligent Traffic System (ITS) has emerged. Intelligent Traffic System uses Vehicular Ad Hoc Network (VANET) to transmit information between vehicles in wireless network, which improves traffic mobility and reduces the number of road accidents. Common communication technologies of VANET include Dedicated Short Range Communications (DSRC) and cellular vehicle-to-everything (C-V2X).
[0003] Since 1999, DSRC technology has dominated the in-vehicle communication solutions until the introduction of C-V2X technology. In terms of congestion management, DSRC technology has a fairly mature research, and the application of C-V2X technology in this regard is a new topic worthy of attention. There are currently several mainstream methods to control problems such as excessive channel load and channel capacity limitation: First, limit the number of sending nodes. When the density is large, the vehicle speed is relatively small, and the information sent by nearby vehicles is not much different. By limiting the sending nodes and selecting typical nodes to send data, the purpose of congestion control can be achieved; second, select an indicator that represents channel congestion. When the indicator exceeds the congestion threshold, it is determined that the channel in the current area is in a congested state, and measures are taken to control the indicator to be below the threshold; third, adaptive control schemes are adopted according to the current communication environment.
[0004] In the V2V (Vehicle to Vehicle) environment, a congestion control solution that is adaptive to the environment is more suitable. The main research objects of congestion control at present are the transmission interval and transmission power of safety messages. The adjustment of transmission power is mainly aimed at the influence of the transmission distance of safety messages. The greater the transmission power, the longer the transmission distance of the safety message, and the stronger the interference to surrounding vehicles. On the contrary, the smaller the communication distance, the weaker the interference. The adjustment of the transmission interval is mainly aimed at the frequency of sending safety messages. Extending the transmission interval can reduce the frequency of sending safety messages and reduce the channel load to alleviate network congestion.
[0005] The existing adaptive congestion control strategy methods based on transmission power adjustment are mainly as follows:
[0006] The real-time distributed control strategy of adjusting the transmission power according to the density around the vehicle is used. The vehicle density within the vehicle communication range is obtained through the Poisson distribution of vehicles in the road scene. The transmission distance required to maintain the maximum PDR is obtained by combining the proposed information delivery rate model. The transmission power of the vehicle is adjusted according to the vehicle, which not only ensures the maximum information delivery rate but also avoids network congestion.
[0007] The channel status is evaluated based on the channel busy rate of the vehicle self-test, and the channel load at the next moment is predicted through Bayesian network learning. The transmission power is adaptively adjusted based on the prediction results.
[0008] A control algorithm that dynamically adjusts transmission power based on road vehicle density ensures connectivity between vehicles.
[0009] The existing adaptive congestion control strategy methods based on transmission interval adjustment mainly include the following: an adaptive beacon generation rate ABGR congestion control mechanism; transplanting a wireless channel congestion control algorithm specified in the Society of Automotive Engineers (SAE) J2945 / 1 standard to C-V2X technology; different congestion management methods under IEEE802.11p and C-V2X.
[0010] In the existing technology, the problem of Beacon adaptation under different road vehicle densities is not taken into account. Sending safety messages at a periodic frequency may cause a large amount of data redundancy, waste of channel bandwidth and broadcast storms, leading to serious network congestion problems, which may eventually cause traffic safety hazards. In addition, the radiation effect of dangerous vehicles on surrounding vehicles in the actual road area is not taken into account, and the regional danger level is not taken into account in the research on safety message congestion control of vehicle-mounted self-organizing networks, resulting in channel congestion. Summary of the invention
[0011] The purpose of the present invention is to provide a vehicle network channel congestion control method and system based on regional risk level to solve the problem of channel congestion.
[0012] In order to achieve the above object, the basic scheme of the present invention is: a vehicle network channel congestion control method based on regional risk, comprising the following steps:
[0013] Obtain the driving behavior of the vehicle driver and the physical performance of the vehicle, calculate the road vehicle density, and adjust the BSM transmission interval;
[0014] Using the driving behavior of the vehicle driver and the physical properties of the vehicle, the dangerousness of the vehicle is calculated, and whether the vehicle in the road network is a dangerous vehicle is determined;
[0015] Cluster the vehicles based on their dangerousness, set the dangerous vehicle as the cluster center, and the area within the preset impact distance factor as the dangerous vehicle cluster;
[0016] According to the classification of dangerous vehicle clusters, the danger index of each dangerous vehicle cluster in the road area is calculated;
[0017] Based on the operation trend and location judgment of each vehicle cluster, the weight factor of the dangerous vehicle cluster and the regional danger level of the entire road are calculated;
[0018] Using the danger level of the vehicle and the regional danger level of the entire road, the penalty factor HVF of the dangerous vehicle and the penalty factor NVF of the vehicle in the dangerous vehicle cluster are calculated respectively;
[0019] According to the penalty factor HVF of the dangerous vehicle and the penalty factor NVF of the vehicles in the dangerous vehicle cluster, the safety message distribution strategy of the dangerous vehicle and the safety message distribution strategy of the vehicles in the dangerous vehicle cluster are adjusted respectively.
[0020] The working principle and beneficial effects of this basic scheme are as follows: This scheme controls vehicle safety message congestion based on road vehicle density, and then proposes penalty factors for dangerous vehicles and penalty factors for vehicles within dangerous vehicle clusters based on vehicle risk assessment and dangerous vehicle cluster division, and adaptively adjusts the appropriate transmission interval of vehicles based on the penalty factors. Taking regional risk into account in the study of safety message congestion control of vehicle-mounted self-organizing networks enables vehicles with different operating states and vehicle risk to take active congestion control to reduce channel load before channel congestion. This scheme proposes penalty factors NVF and HVF, and adjusts the BSM packet generation cycle accordingly, which not only ensures that basic safety messages are sent at an appropriate transmission frequency when the vehicle density is high, but also avoids channel congestion due to too many vehicles.
[0021] Furthermore, the method of calculating the road vehicle density and adjusting the BSM transmission interval is as follows:
[0022]
[0023] Where Max_ITT(k) is the security message transmission interval in milliseconds, vMax_ITT is the maximum threshold of the transmission interval in the calculation; N s (k) is the regional vehicle density of the current K calculation interval; B is the density coefficient, and α is the calculation factor.
[0024] According to the vehicle density in the actual road area, the basic safety message transmission interval congestion control of the vehicle node is adaptively adjusted to facilitate use.
[0025] Further, the method for calculating the dangerousness of a vehicle is as follows:
[0026] The acquisition and judgment of the real-time behavior RF of the vehicle driver needs to be done through the on-board intelligent device. RF is composed of the real-time vehicle condition Vcondition and the driver's physical condition H driver, Real-time driving behavior RS driving The characteristic weight factor W of the characteristic vector ri The inner product is obtained, and its calculation formula is as follows:
[0027] RF HV =[V condition ,H driver ,RS driving ]·[W Ri ] T
[0028] Where T represents transpose;
[0029] The vehicle's physical performance PF is obtained through the OBD II interface. PF is calculated by the vehicle's mileage K VT 、Vehicle Type V type , Drivers accumulate experience ACE driver The health parameter weight factor W of the eigenvector Li The inner product is obtained, and its calculation formula is as follows:
[0030] PF HV =[K VT ,V type ,ACE driver ]·[W Li ] T ;
[0031] Calculate the danger level H of the vehicle V :
[0032] H v =αPF HV +βRF hv
[0033] Among them, α and β represent the weight factors of physical characteristics and real-time driving characteristics, respectively, and H V Health is a value from 1 to 10.
[0034] Since dangerous vehicles will affect surrounding vehicles, it is necessary to evaluate the dangerousness of the vehicle in order to control the congestion of safety message transmission.
[0035] Further, the method for determining whether a vehicle in the road network is a dangerous vehicle is as follows:
[0036] The degree of danger of vehicles is divided into ten levels, namely I to X. The higher the level, the greater the degree of danger of the vehicle. The health status is I to VII, that is, vehicles with a danger level of I to VII are rated as ordinary vehicles, and vehicles with a danger level of VIII to X are rated as dangerous vehicles.
[0037] The judging operation is simple and easy to use.
[0038] Furthermore, the vehicle clustering process also includes cluster merging and cluster separation methods, and the specific steps are as follows:
[0039] Cluster merging: When the inter-vehicle distance between two dangerous vehicles is lower than the influence distance factor R, the two dangerous vehicle clusters partially overlap, and there are two or more dangerous vehicles within the influence distance;
[0040] When the inter-vehicle distance between two dangerous vehicles is lower than the set influencing distance factor R, the related dangerous vehicle clusters are merged;
[0041] Cluster separation: When there are two or more dangerous vehicles in a dangerous vehicle cluster, and the distance between the dangerous vehicles is greater than the influencing distance factor R, the dangerous vehicle cluster is separated, and a new dangerous vehicle cluster is generated if the dangerous vehicle cluster division conditions are met.
[0042] According to the specific road vehicle information, the cluster structure is adjusted to facilitate use. If the distance between two dangerous vehicles is less than the influencing distance factor R, the two dangerous vehicle clusters will overlap in some areas, which will increase the danger index of the area and may cause serious traffic accidents.
[0043] Furthermore, the method for calculating the danger index of each dangerous vehicle cluster in the road area is as follows:
[0044] If there are vehicles on the road that are in the same driving state as all the vehicles in the area, including driving direction and running speed, the overall energy of the area will decrease, and the risk will also decrease accordingly; on the contrary, if the average speed of the vehicle is significantly different from that of the vehicles in the area, the overall energy will increase and the risk will also be high. The formula for evaluating the danger of a vehicle cluster is as follows:
[0045]
[0046] Among them, cluster_risk{r} represents the total energy under the dangerous vehicle cluster r on the road, that is, the risk of the dangerous vehicle cluster; J(i,j) is the vehicle-to-vehicle relationship between vehicles, H(i) and H(j) are the vehicle risks of vehicle i and vehicle j respectively, and V is the average speed of all vehicles in the dangerous vehicle cluster.
[0047] The vehicle cluster is a group evaluation. Different from the workshop interaction relationship, it involves the influence between multiple vehicles and is also related to the health status and driving status of a single vehicle. Obtaining the danger level of each cluster in the road area is conducive to a comprehensive understanding of the danger level of the road area and is more conducive to subsequent control.
[0048] Furthermore, the method for calculating the weight factor of the dangerous vehicle cluster and the regional danger of the entire road is as follows:
[0049] Assume that the physical center of the road area is the danger center. When the dangerous vehicle cluster is at the edge of the road and far away from the danger center, the danger of the vehicle cluster has little impact on the evaluation of the regional danger. When the vehicle cluster is at the danger center, the danger of the vehicle cluster has a greater impact on the evaluation of the regional danger. The weight factor of the dangerous vehicle cluster is defined as follows:
[0050]
[0051] Among them, K j (t) represents the weight factor of dangerous vehicle cluster j at time t, (x c ,y c ) represents the physical center of the region, (x j (t),y j (t)) represents the center position of vehicle cluster j at time t; L represents the length of the road area setting;
[0052] When a dangerous vehicle cluster is obtained by merging several vehicle clusters, the center of the cluster is no longer the dangerous vehicle, but the comprehensive average position of the k dangerous vehicles under the cluster, which is calculated as follows:
[0053]
[0054] After obtaining the weight factor of each vehicle cluster, we perform a weighted summation with the danger level of the dangerous vehicle cluster to obtain the regional danger level of the entire road:
[0055]
[0056] Where n represents the number of dangerous vehicle clusters existing in the road area at time t, and cluster_riskj(t) represents the risk level of vehicle cluster j at time t.
[0057] The location factors of vehicle clusters are studied, and the weight factors of dangerous vehicle clusters are comprehensively obtained, and then the danger level of the entire road area is calculated. The operation is simple and easy to use.
[0058] Furthermore, the method for calculating the penalty factor HVF of dangerous vehicles is as follows:
[0059] The relationship between the dangerous vehicle penalty factor and vehicle speed: When the relative speed between the dangerous vehicle and the average speed of the vehicle cluster is zero, it means that the driving state of the cluster is consistent; when the relative speed difference is large, it means that the movement trend of the dangerous vehicle and its neighboring vehicles is quite different; mapped to the dangerous vehicle penalty factor, the greater the relative speed, the greater the penalty factor, and they are linearly correlated:
[0060] HVF~ΔV;
[0061] The relationship between the dangerous vehicle penalty factor and the cluster vehicle density: the greater the cluster vehicle density, the more vehicles are within the dangerous vehicle's influence range, and the priority of sending safety messages to the dangerous vehicle should be increased; mapped to the dangerous vehicle penalty factor, the smaller the cluster vehicle density, the smaller the penalty factor; the larger the cluster density, the larger the penalty factor, and the penalty factor and cluster vehicle density are in a logarithmic growth relationship:
[0062] HVF~log(1+ρ);
[0063] The relationship between the dangerous vehicle penalty factor and the dangerousness of the vehicle: The dangerousness of the vehicle can indicate the dangerousness of the vehicle. The ratio of the health status of the dangerous vehicle to the average health status of the vehicles in the cluster can be used as a cluster dangerousness measurement indicator of the penalty factor. If the ratio between the health status is small, it is considered that the dangerousness of the vehicle cluster is small. When mapped to the penalty factor, the ratio of the health status can be used as a multiple relationship.
[0064] The comprehensive penalty factor is related to the three factors. The penalty factor for dangerous vehicles is:
[0065]
[0066] Where, ΔV i represents the relative speed of dangerous vehicle i and the average speed of the cluster, ρ i represents the density within the cluster of dangerous vehicle i, H(j) represents the danger level of vehicle j within the cluster, and n represents the number of vehicles within the dangerous vehicle cluster.
[0067] The penalty factor of dangerous vehicles can evaluate the driving differences between dangerous vehicles and neighboring vehicle nodes, as well as the driving status of vehicle clusters. This scheme can adjust the safety message transmission interval according to different penalty factors.
[0068] Furthermore, the method for calculating the penalty factor NVF of vehicles in the dangerous vehicle cluster is as follows:
[0069] Vehicles in the dangerous vehicle cluster receive beacon messages from dangerous vehicles within the influencing distance factor, store them in the neighbor dangerous vehicle list in each observation period, and calculate the penalty factor of vehicles in the dangerous cluster:
[0070]
[0071] Among them, the NVF penalty factor is the sum of the vehicle-to-vehicle relationship weights between the normal vehicle i in the cluster and the dangerous vehicle j that affects it, plus the regional risk; J(i,j) is the vehicle-to-vehicle relationship between vehicles, H(i) and H(j) are the vehicle risks of vehicle i and vehicle j respectively; +Risk r Indicates the regional hazard level of the entire road;
[0072] The NVF indicators are standardized, the inverse tan function is used to transform the workshop relationship weight to achieve data normalization, and the log function is used to process the regional hazard:
[0073]
[0074] This solution adjusts the NV security message transmission interval according to the value of NVF, which is simple to operate and easy to use.
[0075] The present invention also provides an in-vehicle network channel congestion control system based on regional hazard level, comprising an in-vehicle intelligent device and a control unit, wherein the in-vehicle intelligent device is used to obtain vehicle information, the output end of the in-vehicle intelligent device is connected to the input end of the control unit, and the control unit executes the method described in the present invention to perform in-vehicle network channel congestion control.
[0076] The system utilizes a control unit to separately adjust the safety message distribution strategy for dangerous vehicles and the safety message distribution strategy for vehicles within a dangerous vehicle cluster, which is easy to operate. BRIEF DESCRIPTION OF THE DRAWINGS
[0077] Figure 1 It is a flow chart of a vehicle network channel congestion control method based on regional risk level of the present invention;
[0078] Figure 2 It is a schematic diagram of the structure of cluster merging of the vehicle network channel congestion control method based on regional risk level of the present invention;
[0079] Figure 3 It is a structural schematic diagram of cluster separation of the vehicle network channel congestion control method based on regional hazard level of the present invention. DETAILED DESCRIPTION
[0080] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.
[0081] In the description of the present invention, it is necessary to understand that the terms "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.
[0082] In the description of the present invention, unless otherwise specified and limited, it should be noted that the terms "installed", "connected" and "connected" should be understood in a broad sense. For example, it can be a mechanical connection or an electrical connection, or it can be the internal connection between two components. It can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to the specific circumstances.
[0083] The present invention discloses a vehicle network channel congestion control method based on regional risk level, such as Figure 1 As shown, the following steps are included:
[0084] Obtain the driving behavior of vehicle drivers and the physical performance of vehicles, calculate the road vehicle density (vehicle density refers to the number of vehicles per lane per kilometer, which can be obtained using roadside units RSU), and adjust the BSM (Business Service Management, which is software that establishes a connection between business-focused IT services and basic IT infrastructure) transmission interval;
[0085] Using the driving behavior of the vehicle driver and the physical properties of the vehicle, the dangerousness of the vehicle (i.e. the health status of the vehicle) is calculated, and whether the vehicle in the road network is a dangerous vehicle is determined;
[0086] There are multiple types of vehicles, vehicle driving states and vehicle traffic flow densities in the road area. In view of the multiple different vehicle group states in the road area, the road needs to be clustered. The road area is equipped with a road side unit RSU (Road Side Unit). The RSU unit obtains road vehicle messages within the management range, beacon messages of driving states, and obtains the danger level and movement trend of vehicles on the road. Based on the danger level of the vehicle, the RSU unit clusters the vehicle, sets the dangerous vehicle as the cluster center, and presets the area within the influence distance factor R (that is, the area within the influence range with R as the radius) as the dangerous vehicle cluster.
[0087] According to the dangerous vehicle cluster division, the danger index of each dangerous vehicle cluster in the road area is calculated; the existing roadside unit RSU obtains the road vehicle information within the management range, and according to the dangerous vehicles in the road area, the dangerous vehicle cluster is set as the center and the impact factor R is set as the radius;
[0088] Based on the running trend and position judgment of each vehicle cluster, the weight factor of the dangerous vehicle cluster and the regional danger level of the entire road are calculated; based on the existing roadside units RSU in the road area, mobile vehicles broadcast beacon information containing driving status, RSU obtains road vehicle information within the management range, obtains the danger level and movement trend of vehicles on the road, and obtains location information based on the beacon information broadcast by RSU;
[0089] Using the danger level of the vehicle and the regional danger level of the entire road, the penalty factor HVF of the dangerous vehicle and the penalty factor NVF of the vehicle in the dangerous vehicle cluster are calculated respectively;
[0090] According to the penalty factor HVF of dangerous vehicles and the penalty factor NVF of vehicles in dangerous vehicle clusters, the safety message distribution strategy of dangerous vehicles and the safety message distribution strategy of vehicles in dangerous vehicle clusters are adjusted respectively. According to the specific values obtained, when the penalty factor is high, the message sending frequency of dangerous vehicles is increased, that is, the BSM transmission interval of vehicles is shortened. When HVF is less than 15, the transmission interval is 100ms; when HVF is between 15 and 18, the transmission interval is adjusted to 50; when HVF is greater than 18, the transmission interval is adjusted to 20. For NVF, the relationship between HVF and message transmission interval is that when NVF is less than 11, the transmission interval is 100ms; when NVF is between 11 and 14, the transmission interval is adjusted to 50; when NVF is greater than 14, the transmission interval is adjusted to 20.
[0091] In a preferred embodiment of the present invention, the smaller the BSM transmission interval, the higher the frequency of surrounding vehicles receiving update messages, and the more accurate the node's operating status and perception of the surrounding environment. However, the faster the update frequency, the better. The smaller the message transmission interval, the more data in the wireless communication channel, resulting in insufficient channel resources and channel congestion. On the contrary, the larger the BSM transmission interval, the greater the timeliness of the vehicle receiving information. While the node is waiting to receive the next basic safety message, the driving environment around the vehicle node may change dramatically, and it is impossible to provide reliable information for safety applications, which is not conducive to the vehicle making more accurate driving decisions.
[0092] This scheme proposes a method to calculate the road vehicle density and adaptively adjust the BSM transmission interval according to the vehicle density of the actual road area. The algorithm is:
[0093]
[0094] Where Max_ITT(k) is the security message transmission interval in milliseconds, vMax_ITT is the maximum threshold of the transmission interval in the calculation; N s (k) is the regional vehicle density of the current K calculation interval; B is the density coefficient, and α is the calculation factor. The density coefficient B is the key to the congestion control algorithm for the safety message transmission interval, and its selection is crucial. Traffic density is related to the service level of the road. According to the Highway Capacity Manual (HCM, 2010), when the road vehicle density is the number of vehicles per lane per kilometer, the road service level is divided into six levels. The AF service levels represent different traffic densities and traffic conditions, as shown in Table 1 below:
[0095] Table 1 AF service level
[0096]
[0097]
[0098] When the traffic flow is in a non-blocking state, the road vehicle density is considered to be a stable flow, and the transmission interval of the vehicle node can be set according to the default transmission interval (100ms). When the road vehicle density is greater than 25pcu / (km.ln), it is an unstable flow, and the transmission interval of the vehicle safety message needs to be adjusted through an adaptive channel congestion control strategy; when the road vehicle density is greater than 45pcu / (km.ln), the traffic flow density is a forced flow, and the vehicle safety message transmission interval is set to the maximum threshold value vMax_ITT. Therefore, the density coefficient B is set to 25pcu / (km.ln) and the calculation factor is set to 0.36.
[0099] For example, the transmission interval of the safety message can be set in the range of 20 milliseconds to 1000 milliseconds. Considering the timeliness and effectiveness of the vehicle obtaining the safety message, 500 milliseconds is selected as the maximum threshold of the transmission interval. Substituting the selected parameters into the adaptive control strategy of the safety message transmission interval, the formula is as follows:
[0100]
[0101] When the vehicle density in the area of the current observation interval is less than 30, the BSM transmission interval ITT of the vehicle nodes in the area is set to 100ms; when the vehicle density is greater than 30 and less than 45, the BSM transmission interval is determined by the road vehicle density; when the vehicle density is greater than 45, the BSM transmission interval is the set maximum threshold of 500ms.
[0102] In a preferred embodiment of the present invention, since dangerous vehicles will affect surrounding vehicles, it is necessary to evaluate the danger level of the vehicle. The danger level of the vehicle is determined by two elements: the driver's cognitive driving behavior and the physical performance of the vehicle. The physical performance of the vehicle decreases with the increase of the vehicle's mileage and is affected by the type of vehicle. The driver's accumulated experience can be determined by the driver's driving age. In a road traffic system, the greater the danger level of the assessed vehicle, the greater the probability of a traffic accident on the road. The vehicle becomes a source of danger, threatening the driving status of surrounding vehicles, so the poor health status of the vehicle has a certain impact on the driving decisions of surrounding vehicles. The method for calculating the danger level of a vehicle is as follows:
[0103] The acquisition and judgment of the real-time behavior RF (Real Time Feature) of the vehicle driver needs to be done through the on-board intelligent device. RF is composed of the real-time vehicle condition Vcondition, the driver's physical condition H driver , Real-time driving behavior RS driving The characteristic weight factor W of the characteristic vector ri The inner product is obtained, Vconditon is obtained through the vehicle-mounted sensor, and Hdriver is obtained through the device worn by the driver. driving Obtained through real-time driving recorder; W ri According to different parameter value definitions, Vcondition, Hdriver, RSdriving are normal when W ri Take 0, which is bad W ri Take 1, very bad W ri Take 2. The calculation formula is as follows:
[0104] RF HV =[V condition ,H driver ,RS driving ]·[W Ri ] T
[0105] Where T represents transpose;
[0106] The vehicle's physical performance PF (Physics Feature) is obtained through the OBD II (On Board Diagnostics II) interface. PF is calculated based on the vehicle's mileage K VT 、Vehicle Type V type , Drivers accumulate experience ACE driver The health parameter weight factor W of the eigenvector Li The inner product is obtained; ACE driver The driver's driving age weight takes different values according to the previous parameter values. ACE driver If the driver's driving age is greater than or equal to 15 years, the corresponding health parameter weight factor is 1; if the driver's driving age is between 11 and 15 years, the corresponding health parameter weight factor is 2; if the driver's driving age is between 6 and 10 years, the corresponding health parameter weight factor is 3; if the driver's driving age is between 1 and 5 years, the corresponding health parameter weight factor is 4; K VT:When the vehicle mileage is 0-30000km, the corresponding health parameter weight factor is 0; when the vehicle mileage is 30000-70000km, the corresponding health parameter weight factor is 1; when the vehicle mileage is 70000-140000km, the corresponding health parameter weight factor is 2; when the vehicle mileage is 140000-200000km, the corresponding health parameter weight factor is 3; when the vehicle mileage is 200000-300000km, the corresponding health parameter weight factor is 4; When the mileage is greater than 300,000 km, the corresponding health parameter weight factor is 5; Vtype: the vehicle type is a low-speed truck, the corresponding health parameter weight factor is 1, the vehicle type is a 8-19-seat bus, the corresponding health parameter weight factor is 2, the vehicle type is a medium-sized truck or a bus with more than 20 seats, the corresponding health parameter weight factor is 3, the vehicle type is a heavy truck, trailer, semi-trailer, the corresponding health parameter weight factor is 4, and the vehicle type is a car with less than 7 seats, the corresponding health parameter weight factor is 5. The calculation formula is as follows:
[0107] PF HV =[K VT ,V type ,ACE driver ]·[W Li ] T ;
[0108] Calculate the danger level H of the vehicle V :
[0109] H v =αPF HV +βRF hv
[0110] Among them, α and β represent the weight factors of physical characteristics and real-time driving characteristics, respectively, and H V Health is a value from 1 to 10.
[0111] In a preferred embodiment of the present invention, the method for determining whether a vehicle in a road network is a dangerous vehicle is as follows:
[0112] The vehicle danger level is divided into ten levels, namely I to X. The higher the level, the greater the danger level of the vehicle. The health status is I to VII, that is, vehicles with a danger level of I to VII are rated as ordinary vehicles (also known as normal vehicles, Normal Vehicle, NV), and the danger level of VIII to X is represented as a risky vehicle (Risk Vehicle, RV).
[0113] In a preferred solution of the present invention, due to the dynamic nature of the road area, attention needs to be paid to cluster merging and separation. The vehicle clustering process also includes cluster merging and cluster separation methods. The specific steps are as follows:
[0114] like Figure 2 As shown, cluster merging: when the inter-vehicle distance between two dangerous vehicles is lower than the influencing distance factor R, the two dangerous vehicle clusters partially overlap, and there are two or more dangerous vehicles within the influencing distance;
[0115] When the inter-vehicle distance between two dangerous vehicles is lower than the set influencing distance factor R, the related dangerous vehicle clusters are merged;
[0116] like Figure 3 As shown, cluster separation: when there are two or more dangerous vehicles in a dangerous vehicle cluster, and the distance between the dangerous vehicles is greater than the influencing distance factor R, the dangerous vehicle cluster is separated, and a new dangerous vehicle cluster is generated under the conditions for satisfying the dangerous vehicle cluster division.
[0117] In a preferred embodiment of the present invention, the method for calculating the danger index of each dangerous vehicle cluster in the road area is as follows:
[0118] The vehicle cluster is a group evaluation, which is different from the workshop interaction relationship. It involves the impact between multiple vehicles and is also related to the health status and driving status of a single vehicle. Therefore, this solution refers to the total energy concept of the magnet in the Ising model and maps it to the hazard assessment of the road area.
[0119] If there are vehicles on the road that are in the same driving state as all the vehicles in the area, including driving direction and running speed, the overall energy of the area will decrease, and the risk will also decrease accordingly; on the contrary, if the average speed of the vehicle is significantly different from that of the vehicles in the area, the overall energy will increase and the risk will also be high. The formula for evaluating the danger of a vehicle cluster is as follows:
[0120]
[0121] Among them, cluster_risk{r} represents the total energy under the dangerous vehicle cluster r on the road, that is, the risk of the dangerous vehicle cluster; J(i,j) is the vehicle-to-vehicle relationship between vehicles, H(i) and H(j) are the vehicle risks of vehicle i and vehicle j respectively, and V is the average speed of all vehicles in the dangerous vehicle cluster.
[0122] In a preferred embodiment of the present invention, the method for calculating the weight factor of the dangerous vehicle cluster and the regional danger degree of the entire road is as follows:
[0123] There may be one or more dangerous vehicle clusters in the road area. This scheme equates the regional hazard assessment to the assessment of the vehicle cluster. Assume that the physical center of the road area is the hazard center. When the dangerous vehicle cluster is at the edge of the road and far away from the hazard center, the hazard of the vehicle cluster has little impact on the assessment of the regional hazard. When the vehicle cluster is at the hazard center, the hazard of the vehicle cluster has a greater impact on the assessment of the regional hazard. The weight factor of the dangerous vehicle cluster is defined as follows:
[0124]
[0125] Among them, K j (t) represents the weight factor of dangerous vehicle cluster j at time t, (x c ,y c ) represents the physical center of the region, (x j (t),y j (t)) represents the center position of vehicle cluster j at time t; L represents the length of the road area setting;
[0126] When a dangerous vehicle cluster is obtained by merging several vehicle clusters, the center of the cluster is no longer the dangerous vehicle, but the comprehensive average position of the k dangerous vehicles under the cluster, which is calculated as follows:
[0127]
[0128] After obtaining the weight factor of each vehicle cluster, we perform a weighted summation with the danger level of the dangerous vehicle cluster to obtain the regional danger level of the entire road:
[0129]
[0130] Where n represents the number of dangerous vehicle clusters in the road area at time t, cluster_risk j (t) represents the danger level of vehicle cluster j at time t.
[0131] In a preferred embodiment of the present invention, in order to evaluate the driving difference between the dangerous vehicle and the neighboring vehicle nodes, as well as the driving status of the vehicle cluster, the penalty factor HVF (Hazard Vehicle Factor, i.e. Figure 1 RVF in the following way:
[0132] The relationship between the dangerous vehicle penalty factor and vehicle speed: When the relative speed between the dangerous vehicle and the average speed of the vehicle cluster is zero, it means that the driving state of the cluster is consistent; when the relative speed difference is large, it means that the movement trend of the dangerous vehicle and its neighboring vehicles is quite different; mapped to the dangerous vehicle penalty factor, the greater the relative speed, the greater the penalty factor, and they are linearly correlated:
[0133] HVF~ΔV;
[0134] The relationship between the dangerous vehicle penalty factor and the cluster vehicle density: the greater the cluster vehicle density, the more vehicles are within the dangerous vehicle's influence range, and the priority of sending safety messages to the dangerous vehicle should be increased; mapped to the dangerous vehicle penalty factor, the smaller the cluster vehicle density, the smaller the penalty factor; the larger the cluster density, the larger the penalty factor, and the penalty factor and cluster vehicle density are in a logarithmic growth relationship:
[0135] HVF~log(1+ρ);
[0136] The relationship between the dangerous vehicle penalty factor and the dangerousness of the vehicle: The dangerousness of the vehicle can indicate the dangerousness of the vehicle. The ratio of the health status of the dangerous vehicle to the average health status of the vehicles in the cluster can be used as a cluster dangerousness measurement indicator of the penalty factor. If the ratio between the health status is small, it is considered that the dangerousness of the vehicle cluster is small. When mapped to the penalty factor, the ratio of the health status can be used as a multiple relationship.
[0137] The comprehensive penalty factor is related to the three factors. The penalty factor for dangerous vehicles is:
[0138]
[0139] Where, ΔV i represents the relative speed of dangerous vehicle i and the average speed of the cluster, ρ i represents the density within the cluster of dangerous vehicle i, H(j) represents the danger level of vehicle j within the cluster, and n represents the number of vehicles within the dangerous vehicle cluster.
[0140] The safety message transmission interval is adjusted according to different penalty factors. According to the mode 4 of the C-V2X protocol, if it is detected that the current moment is the time for the vehicle node to reserve resources for sending, the safety message is sent; each time data is sent, the random number is reduced by one. When the random number meets the termination condition (reduced to zero), the reset probability is set to determine whether to retain the reserved resources and reset the SPS (security policy) algorithm random number, or reselect the resources.
[0141] In a preferred embodiment of the present invention, the method for calculating the penalty factor NVF of vehicles in a dangerous vehicle cluster is as follows:
[0142] When the distance between an ordinary vehicle and a dangerous vehicle is less than the impact distance threshold, the ordinary vehicle is marked as a dangerous vehicle in the cluster. Due to the impact of dangerous vehicles, the safety message transmission interval of vehicles in the cluster cannot be determined only according to the adaptive safety message control model of road vehicle density, but also in combination with the regional danger degree and the dangerous vehicle's relationship with the vehicle.
[0143] Vehicles in the dangerous vehicle cluster receive beacon messages from dangerous vehicles within the influencing distance factor, store them in the neighbor dangerous vehicle list in each observation period, and calculate the penalty factor of vehicles in the dangerous cluster:
[0144]
[0145] Among them, the NVF penalty factor is the sum of the vehicle-to-vehicle relationship weights between the normal vehicle i in the cluster and the dangerous vehicle j that affects it, plus the regional risk; J(i,j) is the vehicle-to-vehicle relationship between vehicles, H(i) and H(j) are the vehicle risks of vehicle i and vehicle j respectively; +Risk r Indicates the regional hazard level of the entire road;
[0146] Since different evaluation indicators may have different dimensions and dimensional units, the analysis of the results will be affected. Therefore, in order to eliminate the dimensional impact between indicators, the NVF indicators are standardized to solve the comparability between indicators. The inverse tan function is used to transform the workshop relationship weight to achieve data normalization, and the log function is used to process the regional hazard:
[0147]
[0148] The present invention also provides an on-vehicle network channel congestion control system based on regional hazard level, including an on-vehicle intelligent device and a control unit. The on-vehicle intelligent device is used to obtain vehicle information. The output end of the on-vehicle intelligent device is electrically connected to the input end of the control unit. The control unit executes the method described in the present invention to perform on-vehicle network channel congestion control.
[0149] This scheme mainly studies the congestion control strategy of the vehicle network. First, the vehicle safety message congestion control is carried out based on the road vehicle density; then, according to the vehicle risk assessment and dangerous vehicle cluster division, the penalty factor of the dangerous vehicle and the penalty factor of the vehicle in the dangerous vehicle cluster are proposed, and the appropriate transmission interval of the vehicle is adaptively adjusted according to the penalty factor. The present invention takes the regional risk into consideration in the research on the safety message congestion control of the vehicle self-organizing network, so that vehicles with different operating states and health levels can take active congestion control to reduce the channel load before the channel is congested. The penalty factors NVF and HVF are proposed, and the BSM packet generation cycle is adjusted accordingly. It not only ensures that the basic safety messages are sent at an appropriate transmission frequency when the vehicle density is high, but also avoids channel congestion due to too many vehicles.
[0150] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.
[0151] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.
Claims
1. A vehicle network channel congestion control method based on regional risk, characterized in that: The steps include: Obtain the driving behavior of the vehicle driver and the physical performance of the vehicle, calculate the road vehicle density, and adjust the BSM transmission interval; Using the driving behavior of the vehicle driver and the physical properties of the vehicle, the dangerousness of the vehicle is calculated, and whether the vehicle in the road network is a dangerous vehicle is determined; Cluster vehicles based on their dangerousness, set dangerous vehicles as cluster centers, and preset areas within the influencing distance factor as dangerous vehicle clusters; According to the classification of dangerous vehicle clusters, the danger index of each dangerous vehicle cluster in the road area is calculated; Based on the operation trend and location judgment of each vehicle cluster, the weight factor of the dangerous vehicle cluster and the regional danger level of the entire road are calculated; Using the danger level of the vehicle and the regional danger level of the entire road, the penalty factor HVF of the dangerous vehicle and the penalty factor NVF of the vehicle in the dangerous vehicle cluster are calculated respectively; According to the penalty factor HVF of the dangerous vehicle and the penalty factor NVF of the vehicle in the dangerous vehicle cluster, the safety message distribution strategy of the dangerous vehicle and the safety message distribution strategy of the vehicle in the dangerous vehicle cluster are adjusted respectively; The method to calculate the road vehicle density and adjust the BSM transmission interval is: Where Max_ITT(k) is the security message transmission interval in milliseconds, vMax_ITT is the maximum threshold of the transmission interval in the calculation; N s (k) is the regional vehicle density of the current K calculation interval; B is the density coefficient, and α is the calculation factor; The method for calculating the dangerousness of a vehicle is as follows: The acquisition and judgment of the real-time behavior RF of the vehicle driver needs to be done through the on-board intelligent device. RF is composed of the real-time vehicle condition Vcondition and the driver's physical condition H driver , Real-time driving behavior RS driving The characteristic weight factor W of the characteristic vector ri The inner product is obtained, and its calculation formula is as follows: RF HV =[V condition ,H driver ,RS driving ]·[W Ri ] T Where T represents transpose; The vehicle's physical performance PF is obtained through the OBD II interface. PF is calculated by the vehicle's mileage K VT 、Vehicle Type V type , Drivers accumulate experience ACE dr i ver The health parameter weight factor W of the eigenvector Li The inner product is obtained, and its calculation formula is as follows: PF HV =[K VT ,V type ,ACE driver ]·[W Li ] T ; Calculate the danger level H of the vehicle V : H v =αPF HV +βRF hv Among them, α and β represent the weight factors of physical characteristics and real-time driving characteristics, respectively, and H V Health status is a value from 1 to 10; The method for calculating the danger index of each dangerous vehicle cluster in the road area is as follows: If there are vehicles on the road that are in the same driving state as all the vehicles in the area, including driving direction and running speed, the overall energy of the area will decrease, and the risk will also decrease accordingly; on the contrary, if the average speed of the vehicle is significantly different from that of the vehicles in the area, the overall energy will increase and the risk will also be high. The formula for evaluating the danger of a vehicle cluster is as follows: Among them, cluster_risk{r} represents the total energy under the dangerous vehicle cluster r on the road, that is, the dangerousness of the dangerous vehicle cluster; J(i,j) is the vehicle-to-vehicle relationship between vehicles, H(i) and H(j) are the vehicle dangerousness of vehicle i and vehicle j respectively, and V is the average speed of all vehicles in the dangerous vehicle cluster; The method for calculating the weight factor of the dangerous vehicle cluster and the regional danger of the entire road is as follows: Assume that the physical center of the road area is the danger center. When the dangerous vehicle cluster is at the edge of the road and far away from the danger center, the danger of the vehicle cluster has little impact on the evaluation of the regional danger. When the vehicle cluster is at the danger center, the danger of the vehicle cluster has a greater impact on the evaluation of the regional danger. The weight factor of the dangerous vehicle cluster is defined as follows: Among them, K j (t) represents the weight factor of dangerous vehicle cluster j at time t, (x c ,y c ) represents the physical center of the region, (x j (t),y j (t)) represents the center position of vehicle cluster j at time t; L represents the length of the road area setting; When a dangerous vehicle cluster is obtained by merging several vehicle clusters, the center of the cluster is no longer the dangerous vehicle, but the comprehensive average position of the k dangerous vehicles under the cluster, which is calculated as follows: After obtaining the weight factor of each vehicle cluster, we perform a weighted summation with the danger level of the dangerous vehicle cluster to obtain the regional danger level of the entire road: Where n represents the number of dangerous vehicle clusters in the road area at time t, cluster_risk j (t) represents the danger level of vehicle cluster j at time t; The method for calculating the penalty factor HVF of dangerous vehicles is as follows: The relationship between the dangerous vehicle penalty factor and vehicle speed: When the relative speed between the dangerous vehicle and the average speed of the vehicle cluster is zero, it means that the driving state of the cluster is consistent; when the relative speed difference is large, it means that the movement trend of the dangerous vehicle and its neighboring vehicles is quite different; mapped to the dangerous vehicle penalty factor, the greater the relative speed, the greater the penalty factor, and they are linearly correlated: HVF~ΔV; The relationship between the dangerous vehicle penalty factor and the cluster vehicle density: the greater the cluster vehicle density, the more vehicles are within the dangerous vehicle's influence range, and the priority of sending safety messages to the dangerous vehicle should be increased; mapped to the dangerous vehicle penalty factor, the smaller the cluster vehicle density, the smaller the penalty factor; the larger the cluster density, the larger the penalty factor, and the penalty factor and cluster vehicle density are in a logarithmic growth relationship: HVF~log(1+ρ); The relationship between the dangerous vehicle penalty factor and the dangerousness of the vehicle: The dangerousness of the vehicle can indicate the dangerousness of the vehicle. The ratio of the health status of the dangerous vehicle to the average health status of the vehicles in the cluster can be used as a cluster dangerousness measurement indicator of the penalty factor. If the ratio between the health status is small, it is considered that the dangerousness of the vehicle cluster is small. When mapped to the penalty factor, the ratio of the health status can be used as a multiple relationship. The comprehensive penalty factor is related to the three factors. The penalty factor for dangerous vehicles is: Where, ΔV i represents the relative speed of dangerous vehicle i and the average speed of the cluster, ρ i represents the density of dangerous vehicle i in the cluster, H(j) represents the danger level of vehicle j in the cluster, and n represents the number of vehicles in the dangerous vehicle cluster; The method for calculating the penalty factor NVF of vehicles in the dangerous vehicle cluster is as follows: Vehicles in the dangerous vehicle cluster receive beacon messages from dangerous vehicles within the influencing distance factor, store them in the neighbor dangerous vehicle list in each observation period, and calculate the penalty factor of vehicles in the dangerous cluster: Among them, the NVF penalty factor is the sum of the vehicle-to-vehicle relationship weights between the normal vehicle i in the cluster and the dangerous vehicle j that affects it, plus the regional risk factor. J(i,j) is the vehicle-to-vehicle relationship between vehicles, H(i) and H(j) are the vehicle risks of vehicle i and vehicle j respectively. Risk r Indicates the regional hazard level of the entire road; The NVF indicators are standardized, the inverse tan function is used to transform the workshop relationship weight to achieve data normalization, and the log function is used to process the regional hazard:
2. The vehicle network channel congestion control method based on regional risk as claimed in claim 1, characterized in that: The method for determining whether a vehicle in the road network is a dangerous vehicle is as follows: The degree of danger of vehicles is divided into ten levels, namely I to X. The higher the level, the greater the degree of danger of the vehicle. The health status is I to VII, that is, vehicles with a danger level of I to VII are rated as ordinary vehicles, and vehicles with a danger level of VIII to X are rated as dangerous vehicles.
3. The vehicle network channel congestion control method based on regional risk as claimed in claim 1, characterized in that: The process of clustering vehicles also includes cluster merging and cluster separation methods. The specific steps are as follows: Cluster merging: When the inter-vehicle distance between two dangerous vehicles is lower than the influence distance factor R, the two dangerous vehicle clusters partially overlap, and there are two or more dangerous vehicles within the influence distance; When the inter-vehicle distance between two dangerous vehicles is lower than the set influencing distance factor R, the related dangerous vehicle clusters are merged; Cluster separation: When there are two or more dangerous vehicles in a dangerous vehicle cluster, and the distance between the dangerous vehicles is greater than the influencing distance factor R, the dangerous vehicle cluster is separated, and a new dangerous vehicle cluster is generated if the dangerous vehicle cluster division conditions are met.
4. A vehicle network channel congestion control system based on regional risk, characterized in that: It includes an on-board intelligent device and a control unit, wherein the on-board intelligent device is used to obtain vehicle information, the output end of the on-board intelligent device is connected to the input end of the control unit, and the control unit executes the method described in one of claims 1-3 to perform on-board network channel congestion control.
