Multi-dimensional communication risk assessment method and system based on vehicle-mounted Internet of Things

Abstract

The invention provides a multi-dimensional communication risk assessment method and system based on a vehicle-mounted Internet of Things, and relates to the field of communication risk assessment, and the method comprises the steps: obtaining each driving feature in a current time period through the vehicle-mounted Internet of Things; taking the vehicle state, the environment state and the user state as independent evidence sources, calculating a high-risk joint credibility and a medium-risk joint credibility, and determining a current risk value according to a maximum value in the high-risk joint credibility and the medium-risk joint credibility, the dynamic feature weight of each driving feature and each driving feature; and executing a target communication strategy according to the current risk value and the current driving mode when it is determined that the communication incoming call exists in the current time period. According to the method, the accuracy and comprehensiveness of risk assessment can be improved, semantic risk and environment dynamic coupling is realized, the flexibility and adaptability of interception strategies are improved, different response strategies are executed according to risk levels and driving modes, and the flexibility and applicability of the response strategies are improved.

Description

Technical Field

[0001] The present invention relates to the field of communication risk assessment, and particularly to a multi-dimensional communication risk assessment method and system based on vehicle-mounted Internet of Things. Background Art

[0002] With the development of automotive intelligence and networking, vehicle-mounted communication systems play an increasingly important role in improving driving safety and user experience. However, there are still many technical challenges in risk assessment and response strategies for current vehicle-mounted communication systems. In terms of risk assessment, traditional vehicle-mounted communication systems often rely on a single data source or simple fusion methods, resulting in incomplete and inaccurate risk assessment. At the same time, there is a lack of a dynamic coupling mechanism and it is unable to adapt to complex and changeable driving environments in real time. For example, a large amount of data is generated during vehicle operation, including vehicle location, driving trajectory, driver's behavior habits, etc. If this data is not properly protected and fused, it may be stolen or illegally used by hackers, thus posing a threat to vehicle safety. In terms of response strategies, the response strategies of existing systems are usually based on fixed rules and cannot be dynamically adjusted according to real-time risks. The lack of an adaptive adjustment mechanism results in poor applicability of response strategies in different driving scenarios. For example, in an emergency, the system cannot respond quickly, leading to accidents. In addition, vehicle-mounted communication systems involve multiple heterogeneous data sources, and it is difficult for traditional fusion methods to achieve efficient and real-time data integration. There are problems of semantic inconsistency and spatio-temporal asynchrony in the data fusion process, affecting the accuracy and real-time performance of risk assessment and response strategies. At the same time, traditional dynamic weight adjustment methods are difficult to meet the high real-time and high reliability requirements of vehicle-mounted communication systems, lacking smoothness and traceability control, resulting in an uncontrollable weight adjustment process.

[0003] Chinese invention patent application No. 201710726412.7 discloses a driving behavior data acquisition and analysis system and method based on environmental perception. The single-vehicle data information obtained in chronological order is analyzed through a data fusion algorithm to comprehensively obtain driving behavior information. The data fusion algorithm uses Bayesian inference algorithm, decision tree algorithm or D-S evidence inference algorithm to obtain driving behavior information, thereby providing data support for other related fields such as driverless and traffic accident handling. However, it does not apply it to the communication risk assessment under driving conditions.

[0004] Currently, there is no technical solution that can solve the above technical problems, nor is there a multi-dimensional communication risk assessment method and system based on vehicle-mounted Internet of Things. Summary of the Invention

[0005] The present invention provides a multi-dimensional communication risk assessment method and system based on vehicle-mounted Internet of Things, which can solve the technical problems that traditional vehicle-mounted communication systems often rely on a single data source or simple fusion methods, resulting in incomplete and inaccurate risk assessment.

[0006] In a first aspect, the present invention provides a multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things, including:

[0007] Using the vehicle-mounted Internet of Things to obtain each driving feature in the current period, where the driving features include the average vehicle speed and average yaw rate corresponding to the vehicle state, the real-time proximity and weather threat coefficient corresponding to the environmental state, and the fatigue index and user schedule conflict degree corresponding to the user state, determining the current driving mode according to the average vehicle speed and average yaw rate, and determining the dynamic feature weights of each driving feature in the current driving mode;

[0008] Taking the vehicle state, environmental state, and user state as independent evidence sources, calculating the high-risk joint trust degree and medium-risk joint trust degree, and determining the current risk value according to the maximum value of the high-risk joint trust degree and medium-risk joint trust degree, the dynamic feature weights of each driving feature, and each driving feature;

[0009] In the case of determining that there is a communication call in the current period, executing a target communication strategy according to the current risk value and the current driving mode.

[0010] According to the multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things provided by the present invention, the step of using the vehicle-mounted Internet of Things to obtain each driving feature in the current period includes:

[0011] Using the CAN / LIN bus to obtain the average vehicle speed and average yaw rate in the current period;

[0012] Using the DSRC module to obtain traffic blacklist coordinates and accident section radii, determining the real-time proximity according to the current vehicle position, traffic blacklist coordinates, and accident section radii, and determining the weather threat coefficient from a preset rule library according to the current weather type;

[0013] Using a vehicle-mounted camera to capture the driver's facial image, inputting the driver's facial image into a preset fatigue assessment model to obtain the fatigue index output by the preset fatigue assessment model, using the mobile phone interconnection protocol to obtain user calendar data, and determining the user schedule conflict degree according to the user calendar data and the current period.

[0014] According to the multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things provided by the present invention, the step of determining the real-time proximity according to the current vehicle position, traffic blacklist coordinates, and accident section radii includes:

[0015] Determine the vehicle proximity distance based on the current vehicle position and the coordinates of the traffic blacklist;

[0016] Determine the real-time proximity based on the vehicle proximity distance and the accident section radius corresponding to the coordinates of the traffic blacklist, where the accident section radius is determined by extending a preset distance outward from the coordinates of the traffic blacklist in a historical accident-prone area.

[0017] According to the multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things provided by the present invention, the weather threat coefficient determined from a preset rule library according to the current weather type includes:

[0018] When the current weather type is hail weather, determine the weather threat coefficient to be 1; when the current weather type is foggy weather, determine the weather threat coefficient to be 0.7;

[0019] When the current weather type is heavy rain weather, determine the weather threat coefficient to be 0.9; when the current weather type is heavy rain weather, determine the weather threat coefficient to be 0.5; when the current weather type is light rain weather, determine the weather threat coefficient to be 0.3.

[0020] According to the multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things provided by the present invention, the determination of the user schedule conflict degree according to the user calendar data and the current time period includes:

[0021] When the current time period conflicts with the current event in the user calendar data, determine the user schedule conflict degree according to the urgency of the current event;

[0022] When the current time period does not conflict with any current event in the user calendar data, determine the user schedule conflict degree to be 0.

[0023] According to the multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things provided by the present invention, the current driving mode includes a sharp turn mode, a high-speed mode, and an urban mode. The determination of the current driving mode according to the average vehicle speed and the average yaw rate includes:

[0024] For any driving mode, determine the numerator value according to the prior probability of the driving mode and the normal distribution probability density function, where the normal distribution probability density function is determined according to the vector composed of the average vehicle speed and the average yaw rate, the mean vector of the vehicle speed and the yaw rate in the driving mode, and the covariance matrix;

[0025] Traverse the temporary index variables of all driving mode categories, determine the sum of the joint probability densities of all driving modes as the denominator value, and determine the probability value of the driving mode according to the numerator value and the denominator value;

[0026] Traverse the sharp turn mode, high-speed mode, and urban mode, determine the probability value of each driving mode, and determine the current driving mode as the driving mode with the maximum probability value.

[0027] According to the multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things provided by the present invention, the calculating the high-risk joint trust degree and the medium-risk joint trust degree by taking the vehicle state, the environmental state, and the user state as independent evidence sources includes:

[0028] Define the first basic probability assignment function of the vehicle state evidence source according to the average vehicle speed, the maximum vehicle speed, the average yaw rate, and the maximum angular velocity, define the second basic probability assignment function of the environmental state evidence source according to the real-time proximity and the weather threat coefficient, and define the third basic probability assignment function of the user state according to the fatigue index and the user schedule conflict degree;

[0029] According to the first basic probability assignment function, the second basic probability assignment function, and the third basic probability assignment function, calculate step by step through the synthesis of two-by-two evidence sources to obtain the high-risk joint trust degree and the medium-risk joint trust degree respectively.

[0030] According to the multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things provided by the present invention, the determining the current risk value according to the maximum value of the high-risk joint trust degree and the medium-risk joint trust degree, the dynamic feature weight of each driving feature, and each driving feature includes:

[0031] For each driving feature, normalize the driving feature to obtain a normalized feature, determine the feature weight value according to the normalized feature and the dynamic feature weight of the driving feature, traverse all driving features, and determine the sum value of all feature weight values;

[0032] Select the maximum value between the high-risk joint trust degree and the medium-risk joint trust degree as the target joint trust degree, and determine the joint risk trust value according to the target joint trust degree and the risk amplification factor;

[0033] Determine the current risk value according to the sum value of all feature weight values and the joint risk trust value.

[0034] According to the multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things provided by the present invention, the executing the target communication strategy according to the current risk value and the current driving mode includes:

[0035] When the current risk value is greater than the preset high - risk threshold and the current driving mode is the sharp - turn mode, if it is an unfamiliar number, the delayed ring - back duration is determined according to the current speed, and delayed ringing is performed according to the delayed ring - back duration; if it is an advertising number, it is intercepted.

[0036] When the current risk value is greater than the preset high - risk threshold and the current driving mode is the high - speed mode or the urban mode, if it is an unfamiliar number, delayed ringing is performed according to a preset fixed duration; if it is an advertising number, it is transferred to the voicemail.

[0037] When the current risk value is less than or equal to the preset high - risk threshold and greater than or equal to the preset low - risk threshold, normal ringing is performed, and the call is answered on behalf of the user by the voice assistant.

[0038] When the current risk value is less than the preset low - risk threshold, normal ringing is performed.

[0039] In a second aspect, a multi - dimensional communication risk assessment system based on vehicle - mounted Internet of Things is provided, including:

[0040] An acquisition unit, which is used to obtain each driving feature in the current period by using the vehicle - mounted Internet of Things. The driving features include the average vehicle speed and average yaw rate corresponding to the vehicle state, the real - time proximity and weather threat coefficient corresponding to the environmental state, and the fatigue index and user schedule conflict degree corresponding to the user state. The current driving mode is determined according to the average vehicle speed and average yaw rate, and the dynamic feature weight of each driving feature in the current driving mode is determined.

[0041] A calculation unit, which is used to use the vehicle state, environmental state and user state as independent evidence sources, calculate the high - risk joint trust degree and medium - risk joint trust degree, and determine the current risk value according to the maximum value of the high - risk joint trust degree and medium - risk joint trust degree, the dynamic feature weight of each driving feature and each driving feature.

[0042] An execution unit, which is used to execute the target communication strategy according to the current risk value and the current driving mode when it is determined that there is a communication call in the current period.

[0043] The present invention proposes a vehicle communication risk assessment method based on cross-layer dynamic coupling, which integrates multi-source heterogeneous data such as vehicle status, environmental status, and user status, improves the accuracy and comprehensiveness of risk assessment, realizes the dynamic coupling of semantic risks and the environment, improves the flexibility and adaptability of interception strategies, uses a Gaussian mixture model for driving scenario classification, improves the classification accuracy and real-time performance, dynamically adjusts feature weights based on driving patterns, realizes the smoothness and traceability of the weight adjustment process. At the same time, the present invention executes different response strategies according to the risk level and driving pattern, improves the flexibility and applicability of the response strategy, and provides strong support for the improvement of driving safety and user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0045] Figure 1 is one of the schematic flowcharts of the multi-dimensional communication risk assessment method based on the vehicle-mounted Internet of Things provided by the present invention;

[0046] Figure 2 is the second schematic flowchart of the multi-dimensional communication risk assessment method based on the vehicle-mounted Internet of Things provided by the present invention;

[0047] Figure 3 is the schematic structural diagram of the multi-dimensional communication risk assessment system based on the vehicle-mounted Internet of Things provided by the present invention;

[0048] Figure 4 is the schematic structural diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions in the present invention in conjunction with the drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments in the present invention belong to the scope of protection of the present invention.

[0050] Figure 1 is one of the schematic flowcharts of the multi-dimensional communication risk assessment method based on the vehicle-mounted Internet of Things provided by the present invention. The multi-dimensional communication risk assessment method based on the vehicle-mounted Internet of Things includes:

[0051] Step 101: Obtain each driving feature in the current period by using the vehicle-mounted Internet of Things. The driving features include the average vehicle speed and average yaw rate corresponding to the vehicle state, the real-time proximity and weather threat coefficient corresponding to the environmental state, and the fatigue index and user schedule conflict degree corresponding to the user state. Determine the current driving mode according to the average vehicle speed and average yaw rate, and determine the dynamic feature weights of each driving feature in the current driving mode.

[0052] Step 102: Take the vehicle state, environmental state, and user state as independent evidence sources, calculate the high-risk joint trust degree and medium-risk joint trust degree, and determine the current risk value according to the maximum value of the high-risk joint trust degree and medium-risk joint trust degree, the dynamic feature weights of each driving feature, and each driving feature.

[0053] Step 103: When it is determined that there is a communication call in the current period, execute the target communication strategy according to the current risk value and the current driving mode.

[0054] In Step 101, the obtaining of each driving feature in the current period by using the vehicle-mounted Internet of Things includes:

[0055] Obtain the average vehicle speed and average yaw rate in the current period by using the CAN / LIN bus.

[0056] Use the DSRC module to obtain the traffic blacklist coordinates and accident section radius, determine the real-time proximity according to the current vehicle position, traffic blacklist coordinates, and accident section radius, and determine the weather threat coefficient from the preset rule library according to the current weather type.

[0057] Use the vehicle-mounted camera to capture the driver's face image, input the driver's face image into the preset fatigue assessment model, obtain the fatigue index output by the preset fatigue assessment model, use the mobile phone interconnection protocol to obtain the user calendar data, and determine the user schedule conflict degree according to the user calendar data and the current period.

[0058] Optionally, the present invention can obtain the vehicle speed and yaw rate at each moment in the current period by using the CAN / LIN bus, and perform mean value processing on the vehicle speed and yaw rate at all moments to obtain the average vehicle speed and average yaw rate in the current period.

[0059] Optionally, the determining of the real-time proximity according to the current vehicle position, traffic blacklist coordinates, and accident section radius includes:

[0060] Determine the vehicle proximity distance according to the current vehicle position and traffic blacklist coordinates.

[0061] Determine the real-time proximity based on the vehicle's proximity distance and the accident section radius corresponding to the traffic blacklist coordinates. The accident section radius is determined by extending a preset distance outward from the traffic blacklist coordinates in a historical accident-prone area.

[0062] Optionally, use the vehicle GPS coordinate P GPS to perform spatial matching with the blacklist area and calculate the real-time proximity:

[0063]

[0064] where B is the real-time proximity, P GPS is the current vehicle position, L black is the traffic blacklist coordinate, r is the accident section radius. The accident section radius refers to a virtual radius range obtained through statistical analysis of accident data based on the fact that the frequency or severity of traffic accidents is relatively high in a certain specific section or area over a past period of time. This radius range is centered on the accident-prone point and extends a certain distance outward to indicate that the road safety within this area is relatively low. When a vehicle travels within this radius range, the system can issue a warning to remind the driver to pay attention to safety. The accident section radius is combined with the roadside unit signal received by the DSRC module to evaluate the environmental risk at the current vehicle position. If the vehicle enters the accident section radius range, the environmental risk confidence level will increase accordingly, thus affecting the final communication risk assessment result. The calculation result of the real-time proximity will affect the degree of support of the environmental evidence source for the risk proposition, and further affect the final communication risk assessment result. For example, when the vehicle enters the accident section radius range, the environmental risk confidence level will increase accordingly, which may trigger a higher-level communication security response strategy.

[0065] Optionally, determining the weather threat coefficient from a preset rule base according to the current weather type includes:

[0066] In the case where the current weather type is hail weather, determine the weather threat coefficient to be 1. In the case where the current weather type is foggy weather, determine the weather threat coefficient to be 0.7;

[0067] In the case where the current weather type is heavy rain weather, determine the weather threat coefficient to be 0.9. In the case where the current weather type is heavy rain weather, determine the weather threat coefficient to be 0.5. In the case where the current weather type is light rain weather, determine the weather threat coefficient to be 0.3.

[0068] In other embodiments, the current weather type is not limited to this. For example, when the current weather type is sunny, the weather threat coefficient is determined to be 0; when the current weather type is sandstorm, the weather threat coefficient is determined to be 1; when the current weather type is light snow, the weather threat coefficient is determined to be 0.6, and so on.

[0069] Optionally, the in-vehicle camera of the present invention samples the driver's facial image at 10 fps. The resolution of the driver's facial image can be 128×128, with three RGB channels. The fatigue index is determined using a CNN model. The preset fatigue assessment model can be determined after supervised training using labeled images (labels such as open eyes / closed eyes, yawns, etc.) in a public dataset.

[0070] Optionally, determining the user schedule conflict degree according to the user calendar data and the current time period includes:

[0071] When the current time period conflicts with the current event in the user calendar data, the user schedule conflict degree is determined according to the urgency of the current event;

[0072] When the current time period does not conflict with any current event in the user calendar data, the user schedule conflict degree is determined to be 0.

[0073] Optionally, the user schedule conflict degree is an index that quantifies whether there is an important schedule arrangement for the user during the driving time period, and is used to evaluate whether a call request may interfere with the user's established affairs. For example, if the current time completely conflicts with a key event in the user's schedule, such as an emergency meeting or an emergency flight, and in the case of no conflict, the user schedule conflict degree is determined to be 0.

[0074] For example, if the current events can be a meeting, going to work, and a flight, and their preset urgencies are 0.9, 0.7, and 0.8 respectively, then the user schedule conflict degree can be directly determined. In other embodiments, a conflict degree calculation formula can also be used:

[0075]

[0076] where C s is the user schedule conflict degree, Pr(E i ) is the event priority, Ov(T d , T e ) is the overlapping ratio of the driving time and the event time, M a is the maximum priority preset by the system, and E i is the schedule event.

[0077] Optionally, the current driving mode includes a sharp turn mode, a high-speed mode, and an urban mode. Determining the current driving mode according to the average vehicle speed and the average yaw rate includes:

[0078] For any driving mode, determine the numerator value according to the prior probability of the driving mode and the normal distribution probability density function, where the normal distribution probability density function is determined according to the vector composed of the average vehicle speed and the average yaw rate, the mean vector of the vehicle speed and the yaw rate in the driving mode, and the covariance matrix;

[0079] Traverse the temporary index variables of all driving mode categories, determine the sum of the joint probability densities of all driving modes as the denominator value, and determine the probability value of the driving mode according to the numerator value and the denominator value;

[0080] Traverse the sharp turn mode, the high-speed mode, and the urban mode, determine the probability value of each driving mode, and determine the driving mode with the largest probability value as the current driving mode.

[0081] Optionally, the driving mode classification can refer to the following formula:

[0082]

[0083] Among them, (x|μ k , ∑k) is the normal distribution probability density function, x is [v, w] T , v is the average vehicle speed, w is the average yaw rate, K is 3, that is, the number of driving mode categories, including the high-speed, urban, and sharp turn modes, μ k is the mean vector used to characterize the vehicle speed and the yaw rate in mode k, is the covariance matrix used to characterize the variances and correlations of v and w, π k is the prior probability of mode k, and i is the temporary index variable for traversing all driving mode categories, used to traverse all possible driving mode categories and calculate the sum of the joint probability densities of all modes.

[0084] Further, after traversing the sharp turn mode, the high-speed mode, and the urban mode, the probability value of each driving mode can be calculated, including the probability value of the sharp turn mode, the probability value of the high-speed mode, and the probability value of the urban mode. Determine the driving mode with the largest probability value as the current driving mode.

[0085] Optionally, the present invention determines the dynamic feature weights of each driving feature in the current driving mode. The present invention measures the uncertainty of the feature according to the information entropy of the data. The smaller the entropy, the higher the weight. For each driving feature, the present invention performs normalization processing on it and calculates the information entropy:

[0086]

[0087] Among them, is the sample driving feature, M is the number of historical data samples, k is the driving mode, i is the feature index, j is the summation index, and the calculation of the entropy weight value is as follows:

[0088]

[0089] Furthermore, the dynamic feature weight of each driving feature can be calculated by the following formula:

[0090]

[0091] Among them, exp is the exponential function used to amplify the entropy weight difference, T is the temperature coefficient used to control the smoothness of the weight distribution. Furthermore, the temperature coefficient can be adaptively adjusted by the following formula:

[0092]

[0093] Among them, in the case of the high-speed driving mode, E m is 0.1, in the case of the urban driving mode, E m is 0.5, in the case of the sharp-turn driving mode, E m is 1, γ is the adjustment coefficient, and the default value is 0.5.

[0094] In step 102, the calculating the high-risk joint trust degree and the medium-risk joint trust degree by taking the vehicle state, the environmental state, and the user state as independent evidence sources includes:

[0095] Defining the first basic probability assignment function of the vehicle state evidence source according to the average vehicle speed v, the maximum vehicle speed v max , the average yaw rate w, and the maximum angular velocity w th , defining the second basic probability assignment function of the environmental state evidence source according to the real-time proximity B and the weather threat coefficient E w , and defining the third basic probability assignment function of the user state according to the fatigue index D f and the user schedule conflict degree C s ;

[0096] According to the first basic probability assignment function, the second basic probability assignment function, and the third basic probability assignment function, the high-risk joint trust degree and the medium-risk joint trust degree are respectively obtained through step-by-step calculation by synthesizing two-by-two evidence sources.

[0097] The present invention can set up an identification framework, and the BPA of the three independent evidence sources is defined as follows:

[0098] Vehicle state evidence source:

[0099]

[0100] Environmental evidence source:

[0101]

[0102] User state evidence source:

[0103]

[0104] Furthermore, for the three evidence sources, the combined trust degree is gradually calculated by pairwise combination:

[0105]

[0106] For example, v = 120 km / h (v max = 200), w = 0.8 rad / s (w threshold = 1.0),

[0107]

[0108] Another example, E weather = 0.9 (heavy rain), B = 0.7:

[0109]

[0110] Another example, D fatigue = 0.6, C schedule = 0.8:

[0111]

[0112] In the calculation of pairwise evidence combination, the first combination includes the vehicle and the environment:

[0113]

[0114] Only when the vehicle and environmental propositions are completely in conflict (such as B = {high risk}, C = {medium risk}), but there is no conflict here, so

[0115] In the second combination, that is, when the user state is combined again, m final ({high risk})

[0116]

[0117] In the final trust degree distribution:

[0118]

[0119] In another alternative embodiment, the first step is to define three independent evidence sources. The vehicle state includes vehicle speed and steering wheel angle to determine whether the vehicle is in a high-risk driving state (such as speeding or sharp turning). For example, the faster the vehicle speed and the sharper the turn, the higher the degree of support for high risk. The environmental state includes the severity of the weather and whether it is close to accident-prone sections to evaluate the threat of the external environment to driving safety. For example, when it is rainy and close to accident sections, the degree of support for high risk increases. The user state includes the driver's fatigue level and schedule conflicts to determine whether the driver is easily distracted or needs to prioritize other matters. When driving fatigued and having schedule conflicts, the degree of support for medium risk is higher. Further, a confidence level is assigned to each evidence source. In the vehicle state confidence level, the high-risk confidence level is calculated based on the vehicle speed and turning speed. For example, vehicle speed 120 km / h (maximum 200 km / h), sharp turn intensity 0.8 (threshold 1.0) → high-risk confidence level 48%. In the environmental state confidence level, the high-risk confidence level is the weather threat coefficient × the degree of proximity to the accident area, and the medium-risk confidence level is a partial confidence when the weather is not extreme but close to the blacklist area. For example, heavy rain (coefficient 0.9) and 70% proximity to the accident section → high-risk confidence level 63%. In the user state confidence level, the medium-risk confidence level is the fatigue index × the degree of schedule conflict. For example, fatigue index 60%, schedule conflict 80% → medium-risk confidence level 48%. The third step is to combine the confidence levels of the three evidence sources. If multiple evidences support a certain risk at the same time, the combined confidence level is enhanced. If there is evidence conflict (such as the vehicle supporting high risk and the user supporting medium risk), the contradiction needs to be resolved through mathematical rules. The combination method can be that the total high-risk confidence level = the high-risk confidence levels of the vehicle and the environment × the part not opposed by the user. For example, the high-risk confidence levels of the vehicle and the environment are 48% and 63% respectively, and the user does not explicitly oppose (remaining confidence level 52%) → combined high-risk confidence level 15.7%. The total medium-risk confidence level = the medium-risk confidence level of the user × the part not covered by the environment and the vehicle. For example, the user's medium-risk confidence level is 48%, and the part not covered by the environment and the vehicle is 7.8% → combined medium-risk confidence level 9.2%. Through this step-by-step combination, the system can comprehensively consider the factors of the vehicle, the environment, and the user to quantify the risk level.

[0120] Optionally, determining the current risk value according to the maximum value of the high-risk combined confidence level and the medium-risk combined confidence level, the dynamic feature weight of each driving feature, and each driving feature includes:

[0121] For each driving feature, normalize the driving feature to obtain a normalized feature, determine a feature weight value according to the normalized feature and the dynamic feature weight of the driving feature, traverse all driving features, and determine the sum value of all feature weight values;

[0122] Determine the maximum value between the high - risk combined trust degree and the medium - risk combined trust degree as the target combined trust degree, and determine the combined risk trust value according to the target combined trust degree and the risk amplification factor;

[0123] Determine the current risk value according to the sum value of all the feature weight values and the combined risk trust value.

[0124] Optionally, the current risk value can be calculated by the following formula:

[0125]

[0126] wherein, R total is the current risk value, is the dynamic feature weight of the driving feature, f(F i ) is the normalized feature, β is the preset risk amplification factor, m(high - risk) is the high - risk combined trust degree, and m(medium - risk) is the medium - risk combined trust degree.

[0127] As an alternative embodiment of the present invention, in step 102, the current risk value can also be determined by other means. For example, normalize the average vehicle speed and average yaw rate corresponding to the vehicle state obtained in step 101, the real - time proximity and weather threat coefficient corresponding to the environmental state, and the fatigue index and user schedule conflict degree corresponding to the user state to obtain the normalized average vehicle speed, normalized average yaw rate, normalized real - time proximity, normalized weather threat coefficient, normalized fatigue index, and normalized user schedule conflict degree, and then input the normalized average vehicle speed, normalized average yaw rate, normalized real - time proximity, normalized weather threat coefficient, normalized fatigue index, and normalized user schedule conflict degree into a preset risk assessment model to obtain the current risk value output by the preset risk assessment model, wherein the preset risk assessment model is determined after training based on the sample average vehicle speed, sample average yaw rate, sample real - time proximity, sample weather threat coefficient, sample fatigue index, sample user schedule conflict degree, and sample risk value.

[0128] In step 103, when it is determined that there is a communication call in the current time period, first, obtain the phone number of the communication call, match the phone number with each fraud phone number in the preset fraud phone number library. If there is a matching relationship, determine that the communication call is a fraud call, directly hang up and mark it as a fraud. If there is no matching relationship, then execute the target communication strategy according to the current risk value and the current driving mode.

[0129] Optionally, the executing the target communication strategy according to the current risk value and the current driving mode includes:

[0130] When the current risk value is greater than the preset high-risk threshold and the current driving mode is the sharp-turn mode, if it is an unfamiliar number, the delayed ringing duration is determined according to the current speed, and delayed ringing is performed according to the delayed ringing duration; if it is an advertising number, it is blocked.

[0131] When the current risk value is greater than the preset high-risk threshold and the current driving mode is the high-speed mode or the urban mode, if it is an unfamiliar number, delayed ringing is performed according to a preset fixed duration; if it is an advertising number, it is transferred to the voicemail.

[0132] When the current risk value is less than or equal to the preset high-risk threshold and greater than or equal to the preset low-risk threshold, normal ringing is performed, and the call is answered by the voice assistant.

[0133] When the current risk value is less than the preset low-risk threshold, normal ringing is performed.

[0134] Optionally, when the current risk value is greater than the preset high-risk threshold and the current driving mode is the sharp-turn mode, it is determined that the risk of communication in the current driving state is relatively high. At this time, it is necessary to determine whether the incoming call is an unfamiliar number to avoid missing a key call. If it is an unfamiliar number, the delayed ringing duration is determined according to the current speed, and delayed ringing is performed according to the delayed ringing duration; if it is an advertising number, it is blocked.

[0135] Specifically, determining the delayed ringing duration according to the current speed includes:

[0136]

[0137] where t delay is the delayed ringing duration, v is the current speed. To ensure that the driver will not be distracted by long-term ringing in the high-speed driving state, when v = 0 km / h (stopped): t delay = 0 seconds (no delay required), when v = 50 km / h: t delay = 15 seconds; when v = 100 km / h: t delay = 20 seconds. In the low-speed scenario (v < 30 km / h), short delay or direct ringing is implemented to reduce interference to driving. In the high-speed scenario (v > 80 km / h), the delay time exceeds 18 seconds. The present invention can also trigger automatic blocking in combination with fatigue detection. However, if it is an advertising number, it is blocked to avoid unnecessary disturbance.

[0138] Optionally, when the current risk value is less than or equal to the preset high-risk threshold and greater than or equal to the preset low-risk threshold, it is confirmed that the communication risk of the current driving state is relatively high. Although the call can ring normally in this case, after it is answered, the voice assistant can be used to answer the call on behalf of the driver. At this time, according to the NLP keyword matching technology, if it is determined that the matter is urgent, the driver can be switched to answer the call, or their emergency contact can be notified. When the current risk value is less than the preset low-risk threshold, the call rings normally.

[0139] The present invention proposes a vehicle-mounted communication risk assessment method based on cross-layer dynamic coupling, which integrates multi-source heterogeneous data such as vehicle state, environmental state, and user state, improves the accuracy and comprehensiveness of risk assessment, realizes the dynamic coupling of semantic risk and environment, improves the flexibility and adaptability of the interception strategy, uses the Gaussian mixture model for driving scenario classification, improves the classification accuracy and real-time performance, dynamically adjusts the feature weights based on the driving mode, realizes the smoothness and traceability of the weight adjustment process. At the same time, the present invention executes different response strategies according to the risk level and driving mode, improves the flexibility and applicability of the response strategy, and provides strong support for enhancing driving safety and user experience.

[0140] Figure 2 is the second schematic flowchart of the multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things provided by the present invention. In Figure 2 the present invention combines the features of three dimensions, namely the vehicle's own state, the external environmental state, and the driver's state, and combines the evidence reasoning algorithm provided by the present invention to calculate the current risk value. During the calculation process, the influence brought by dynamic weight adjustment and driving mode is considered, so as to further improve the accuracy of the current risk value. Finally, relevant risk defense strategies are executed according to the incoming call situation.

[0141] Figure 3 is the structural schematic diagram of the multi-dimensional communication risk assessment system based on vehicle-mounted Internet of Things provided by the present invention. The multi-dimensional communication risk assessment system based on vehicle-mounted Internet of Things includes an acquisition unit 1. The acquisition unit 1 is used to obtain each driving feature in the current period by using the vehicle-mounted Internet of Things. The driving features include the average vehicle speed and average yaw rate corresponding to the vehicle state, the real-time proximity and weather threat coefficient corresponding to the environmental state, and the fatigue index and user schedule conflict degree corresponding to the user state. Determine the current driving mode according to the average vehicle speed and average yaw rate, and determine the dynamic feature weights of each driving feature in the current driving mode. The working principle of the acquisition unit 1 can refer to the foregoing step 101 and will not be elaborated here.

[0142] The multi-dimensional communication risk assessment system based on vehicle-mounted Internet of Things further includes a calculation unit 2, which is configured to use the vehicle state, the environmental state, and the user state as independent evidence sources, calculate the high-risk joint trust degree and the medium-risk joint trust degree, and determine the current risk value according to the maximum value of the high-risk joint trust degree and the medium-risk joint trust degree, the dynamic feature weight of each driving feature, and each driving feature. The working principle of the calculation unit 2 can refer to the foregoing step 102 and will not be elaborated herein.

[0143] The multi-dimensional communication risk assessment system based on vehicle-mounted Internet of Things further includes an execution unit 3, which is configured to execute a target communication strategy according to the current risk value and the current driving mode when it is determined that there is a communication incoming call in the current time period. The working principle of the execution unit 3 can refer to the foregoing step 103 and will not be elaborated herein.

[0144] The present invention provides a vehicle-mounted communication risk assessment method based on cross-layer dynamic coupling, which integrates multi-source heterogeneous data such as vehicle state, environmental state, and user state, improves the accuracy and comprehensiveness of risk assessment, realizes the dynamic coupling of semantic risk and environment, improves the flexibility and adaptability of the interception strategy, uses the Gaussian mixture model for driving scenario classification, improves the classification accuracy and real-time performance, dynamically adjusts the feature weights based on the driving mode, realizes the smoothness and traceability of the weight adjustment process. At the same time, the present invention executes different response strategies according to the risk level and the driving mode, improves the flexibility and applicability of the response strategy, and provides strong support for the improvement of driving safety and user experience.

[0145] Figure 4 is a schematic structural diagram of the electronic device provided by the present invention, as Figure 4As shown in the figure, the electronic device may include: a processor 110, a communications interface 120, a memory 130, and a communication bus 140. Among them, the processor 110, the communications interface 120, and the memory 130 complete communication with each other through the communication bus 140. The processor 110 may call logical instructions in the memory 130 to execute a multi-dimensional communication risk assessment method based on the vehicle-mounted Internet of Things. The method includes: obtaining each driving feature in the current period by using the vehicle-mounted Internet of Things. The driving features include the average vehicle speed and average yaw rate corresponding to the vehicle state, the real-time proximity and weather threat coefficient corresponding to the environmental state, and the fatigue index and user schedule conflict degree corresponding to the user state. Determine the current driving mode according to the average vehicle speed and average yaw rate, and determine the dynamic feature weights of each driving feature in the current driving mode; use the vehicle state, environmental state, and user state as independent evidence sources to calculate the high-risk joint trust degree and medium-risk joint trust degree, and determine the current risk value according to the maximum value of the high-risk joint trust degree and medium-risk joint trust degree, the dynamic feature weights of each driving feature, and each driving feature; in the case of determining that there is a communication call in the current period, execute the target communication strategy according to the current risk value and the current driving mode.

[0146] In addition, the logical instructions in the above-mentioned memory 130 can be implemented in the form of software function units and, when sold or used as an independent product, can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store program codes.

[0147] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute a multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things provided by each of the above methods. The method includes: using the vehicle-mounted Internet of Things to obtain each driving feature in the current period, where the driving features include the average vehicle speed and average yaw rate corresponding to the vehicle state, the real-time proximity and weather threat coefficient corresponding to the environmental state, and the fatigue index and user schedule conflict degree corresponding to the user state; determining the current driving mode according to the average vehicle speed and average yaw rate, and determining the dynamic feature weights of each driving feature in the current driving mode; taking the vehicle state, environmental state, and user state as independent evidence sources, calculating the high-risk joint trust degree and medium-risk joint trust degree, and determining the current risk value according to the maximum value of the high-risk joint trust degree and medium-risk joint trust degree, the dynamic feature weights of each driving feature, and each driving feature; in the case of determining that there is a communication call in the current period, executing a target communication strategy according to the current risk value and the current driving mode.

[0148] On another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it realizes the execution of a multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things provided by each of the above methods. The method includes: using the vehicle-mounted Internet of Things to obtain each driving feature in the current period, where the driving features include the average vehicle speed and average yaw rate corresponding to the vehicle state, the real-time proximity and weather threat coefficient corresponding to the environmental state, and the fatigue index and user schedule conflict degree corresponding to the user state; determining the current driving mode according to the average vehicle speed and average yaw rate, and determining the dynamic feature weights of each driving feature in the current driving mode; taking the vehicle state, environmental state, and user state as independent evidence sources, calculating the high-risk joint trust degree and medium-risk joint trust degree, and determining the current risk value according to the maximum value of the high-risk joint trust degree and medium-risk joint trust degree, the dynamic feature weights of each driving feature, and each driving feature; in the case of determining that there is a communication call in the current period, executing a target communication strategy according to the current risk value and the current driving mode.

[0149] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without creative work.

[0150] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0151] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things, characterized in that, Including: Using the vehicle-mounted Internet of Things to obtain each driving feature in the current period, where the driving features include the average vehicle speed and average yaw rate corresponding to the vehicle state, the real-time proximity and weather threat coefficient corresponding to the environmental state, and the fatigue index and user schedule conflict degree corresponding to the user state. Determine the current driving mode according to the average vehicle speed and average yaw rate, and determine the dynamic feature weights of each driving feature in the current driving mode. Taking the vehicle state, environmental state, and user state as independent evidence sources, calculating the high-risk joint trust degree and medium-risk joint trust degree, and determining the current risk value according to the maximum value of the high-risk joint trust degree and medium-risk joint trust degree, the dynamic feature weights of each driving feature, and each driving feature. When it is determined that there is a communication call in the current period, execute the target communication strategy according to the current risk value and the current driving mode.

2. The multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things according to claim 1, wherein The using the vehicle-mounted Internet of Things to obtain each driving feature in the current period includes: Using the CAN / LIN bus to obtain the average vehicle speed and average yaw rate in the current period. Using the DSRC module to obtain traffic blacklist coordinates and the accident section radius. Determine the real-time proximity according to the current vehicle position, traffic blacklist coordinates, and accident section radius, and determine the weather threat coefficient from the preset rule library according to the current weather type. Using the vehicle-mounted camera to capture the driver's face image, input the driver's face image into the preset fatigue assessment model to obtain the fatigue index output by the preset fatigue assessment model, use the mobile phone interconnection protocol to obtain the user calendar data, and determine the user schedule conflict degree according to the user calendar data and the current period.

3. The multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things according to claim 2, wherein, The determining the real-time proximity according to the current vehicle position, traffic blacklist coordinates, and accident section radius includes: Determining the vehicle proximity distance according to the current vehicle position and traffic blacklist coordinates. Determining the real-time proximity according to the vehicle proximity distance and the accident section radius corresponding to the traffic blacklist coordinates, where the accident section radius is determined by extending a preset distance outward from the traffic blacklist coordinates in the historical accident-prone area.

4. The multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things according to claim 2, wherein, The determining the weather threat coefficient from the preset rule library according to the current weather type includes: When the current weather type is hail weather, determining the weather threat coefficient as 1; when the current weather type is fog weather, determining the weather threat coefficient as 0.

7. When the current weather type is heavy rain weather, determining the weather threat coefficient as 0.9; when the current weather type is heavy rain weather, determining the weather threat coefficient as 0.5; when the current weather type is light rain weather, determining the weather threat coefficient as 0.

3.

5. The multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things according to claim 2, wherein The determining the user schedule conflict degree according to the user calendar data and the current period includes: When the current period conflicts with the current event in the user calendar data, determining the user schedule conflict degree according to the urgency of the current event. When there is no conflict between the current time period and any current event in the user calendar data, determine that the user schedule conflict degree is 0.

6. The multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things according to claim 1, wherein The current driving mode includes an emergency turning mode, a high-speed mode, and an urban mode. The determination of the current driving mode according to the average vehicle speed and the average yaw rate includes: For any driving mode, determine the numerator value according to the prior probability of the driving mode and the normal distribution probability density function, where the normal distribution probability density function is determined according to the vector composed of the average vehicle speed and the average yaw rate, the mean vector of the vehicle speed and the yaw rate in the driving mode, and the covariance matrix; Traverse the temporary index variables of all driving mode categories, determine the sum of the joint probability densities of all driving modes as the denominator value, and determine the probability value of the driving mode according to the numerator value and the denominator value; Traverse the emergency turning mode, the high-speed mode, and the urban mode, determine the probability value of each driving mode, and determine the driving mode with the largest probability value as the current driving mode.

7. The multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things according to claim 1, characterized in that The calculation of the high-risk joint trust degree and the medium-risk joint trust degree by taking the vehicle state, the environmental state, and the user state as independent evidence sources includes: Define the first basic probability assignment function of the vehicle state evidence source according to the average vehicle speed, the maximum vehicle speed, the average yaw rate, and the maximum angular velocity, define the second basic probability assignment function of the environmental state evidence source according to the real-time proximity and the weather threat coefficient, and define the third basic probability assignment function of the user state according to the fatigue index and the user schedule conflict degree; According to the first basic probability assignment function, the second basic probability assignment function, and the third basic probability assignment function, calculate step by step through the synthesis of two evidence sources respectively to obtain the high-risk joint trust degree and the medium-risk joint trust degree.

8. The multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things according to claim 1, characterized in that The determination of the current risk value according to the maximum value of the high-risk joint trust degree and the medium-risk joint trust degree, the dynamic feature weight of each driving feature, and each driving feature includes: For each driving feature, normalize the driving feature to obtain a normalized feature, determine the feature weight value according to the normalized feature and the dynamic feature weight of the driving feature, traverse all driving features, and determine the sum value of all feature weight values; Select the maximum value between the high-risk joint trust degree and the medium-risk joint trust degree as the target joint trust degree, and determine the joint risk trust value according to the target joint trust degree and the risk amplification coefficient; Determine the current risk value according to the sum value of all feature weight values and the joint risk trust value.

9. The multi-dimensional communication risk assessment method based on vehicle-mounted Internet of Things according to claim 1, wherein The execution of the target communication strategy according to the current risk value and the current driving mode includes: When the current risk value is greater than the preset high-risk threshold and the current driving mode is the emergency turning mode, if it is an unknown number, determine the delayed ringing duration according to the current speed, and perform delayed ringing according to the delayed ringing duration. If it is an advertising number, intercept it; When the current risk value is greater than the preset high-risk threshold and the current driving mode is the highway mode or the urban mode, if it is an unfamiliar number, ring with a delay according to a preset fixed duration, and if it is an advertising number, transfer it to the voicemail; When the current risk value is less than or equal to the preset high-risk threshold and greater than or equal to the preset low-risk threshold, ring normally and answer the call on behalf of the user by the voice assistant; When the current risk value is less than the preset low-risk threshold, ring normally.

10. A multi-dimensional communication risk assessment system based on vehicle-mounted Internet of Things, characterized in that, It includes: An acquisition unit, which is used to use the vehicle-mounted Internet of Things to acquire each driving feature in the current period. The driving features include the average vehicle speed and the average yaw rate corresponding to the vehicle state, the real-time proximity and the weather threat coefficient corresponding to the environmental state, and the fatigue index and the user schedule conflict degree corresponding to the user state. Determine the current driving mode according to the average vehicle speed and the average yaw rate, and determine the dynamic feature weight of each driving feature in the current driving mode; A calculation unit, which is used to use the vehicle state, the environmental state and the user state as independent evidence sources, calculate the high-risk joint trust degree and the medium-risk joint trust degree, and determine the current risk value according to the maximum value of the high-risk joint trust degree and the medium-risk joint trust degree, the dynamic feature weight of each driving feature and each driving feature; An execution unit, which is used to execute the target communication strategy according to the current risk value and the current driving mode when it is determined that there is a communication incoming call in the current period.

Images