An intelligent auxiliary control method based on in-vehicle wireless communication chips

Abstract

The present invention relates to the technical field of intelligent auxiliary control methods, and particularly relates to an intelligent auxiliary control method based on an in-vehicle wireless communication chip, which includes the following steps: Step 1: Obtain vehicle passing data at different intersections. The vehicle passing data includes vehicle threat types and vehicle routes. The vehicle threat types are divided into low-threat vehicles and high-threat vehicles. Obtain the number of straight, left-turn, and right-turn times of high-threat vehicles within the intersection monitoring period, calculate the risk assessment value, and record the analyzed intersection as the target intersection; Step 2: Establish a wireless communication connection between the image acquisition device and medium and large-sized vehicles; Step 3: Record the medium and large-sized vehicles accessing the wireless communication connection as the accessed vehicles and identify the accessed vehicles; Step 4: Continuously record the passing images of the accessed vehicles, determine the contact risk area based on the blind spot data of the accessed vehicles combined with the vehicle driving data, identify pedestrians and other vehicles within the contact risk area, and generate a warning signal.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent auxiliary control methods, and particularly relates to an intelligent auxiliary control method based on an in-vehicle wireless communication chip. Background Art

[0002] With the improvement of vehicle intelligence level, the application of active or passive safety driving assistance systems in small passenger cars is becoming more and more widespread. For example, the 360 holographic imaging system can effectively reduce the probability of traffic accidents caused by blind spots in small passenger cars. However, for medium and large-sized vehicles, especially large trailers and large trucks, there is often a lack of intelligent imaging systems and large body blind spots. Coupled with the large inner wheel difference caused by the long wheelbase between the front and rear wheels, it is extremely easy to have scratches, collisions and other traffic accidents with surrounding pedestrians and other vehicles when turning through intersections, resulting in losses of life and property.

[0003] In view of the above situation, in the prior art, additional image acquisition devices are usually installed to avoid such problems. However, this method has a high cost and a limited application range, and cannot be applied to all medium and large-sized vehicles passing through intersections, resulting in poor actual application effects. For medium and large-sized vehicles without installed image acquisition devices, there is still a risk of scratching and colliding with surrounding pedestrians and other vehicles. Summary of the Invention

[0004] In view of the above-mentioned drawbacks of the prior art, the present invention provides an intelligent auxiliary control method based on an in-vehicle wireless communication chip, which can effectively solve the problem in the prior art that it is difficult to give accident warnings for medium and large-sized vehicles passing through intersections.

[0005] To achieve the above object, the present invention is realized through the following technical solutions:

[0006] The present invention provides an intelligent auxiliary control method based on an in-vehicle wireless communication chip, which at least includes the following steps:

[0007] Step 1: Obtain vehicle passing data of different intersections. The vehicle passing data includes vehicle threat types and vehicle routes. The vehicle threat types are divided into low-threat vehicles and high-threat vehicles according to the vehicle size and the size of the driving blind spot.

[0008] The vehicle size and the driving blind spot are quantitatively represented as a vehicle type evaluation value and a blind spot evaluation value, which are calculated based on the image analysis when the vehicle passes through the intersection.

[0009] Obtain the number of straight, left-turn and right-turn times of high-threat vehicles within the intersection monitoring period, multiply by the preset weight coefficient and sum to obtain a risk evaluation value. When the risk evaluation value is greater than the risk evaluation threshold, the intersection is recorded as a target intersection.

[0010] Step 2: Build a wireless communication platform at the target intersection to establish a wireless communication connection between the image acquisition device and medium and large-sized vehicles;

[0011] Step 3: Denote the medium and large-sized vehicles connected to the wireless communication platform as connected vehicles, and identify the connected vehicles in the image;

[0012] Step 4: Independently analyze each connected vehicle to obtain the driving direction of the connected vehicle;

[0013] When the connected vehicle goes straight, construct a contact risk area based on the vehicle speed, width, and vehicle blind spot data;

[0014] When the connected vehicle turns, output the corresponding inner wheel difference according to the current steering wheel rotation angle of the connected vehicle, and output the corresponding turning interference area and construct a contact risk area in response to the input inner wheel difference;

[0015] Identify pedestrians and other vehicles in the contact risk area and generate a warning signal.

[0016] Furthermore, the process of obtaining the weight coefficients corresponding to the number of straight, left, and right turns is as follows:

[0017] Obtain the accident liability recognition certificates of multiple vehicle traffic accidents, and extract the accident occurrence locations, accident vehicle types, and accident processes of the accident liability recognition certificates based on natural language processing technology to form an accident data set;

[0018] Screen out the accident data set with the accident occurrence location at the intersection and the accident vehicle type as large-sized vehicles and denote it as the target data set, and obtain the driving direction of the large-sized vehicles in the accident process in the target data set;

[0019] Statistically analyze the proportions of the straight, left, and right turn directions in the driving direction as the weight coefficients corresponding to the number of straight, left, and right turns.

[0020] Furthermore, the process of classifying vehicle threat types is as follows:

[0021] S1: Denote the currently identified vehicle as the target vehicle, obtain multiple top-down images of the target vehicle when passing through the analysis intersection as passing images, and obtain the outline of the target vehicle in the passing images as the target outline;

[0022] S2: Construct a contour limiting rectangle, where the contour limiting rectangle is the smallest rectangle containing the target outline, calculate the ratio of the target outline area to the area of the contour limiting rectangle and denote it as the compliance index. There is a preset compliance index threshold, and screen out the passing images in which the compliance index is greater than the compliance index threshold and the target outline covers the zebra stripe pattern as the first analysis images;

[0023] S3: Screen out the second analyzed image based on the position of the contour-defined rectangle in the first analyzed image, where the vehicle body in the second analyzed image is closest to being parallel to the zebra stripes;

[0024] S4: Calculate the area of the contour-defined rectangle in the second analyzed image and divide it by the square of the length of the zebra stripes to obtain the vehicle type evaluation value of the target vehicle, and obtain the maximum value of the compliance index among all the passing images of the target vehicle as the blind area evaluation value;

[0025] Multiply the blind area evaluation value and the vehicle type evaluation value by a preset weight coefficient respectively and sum them to obtain the threat evaluation value. There is a preset danger evaluation threshold, and the target vehicle is classified as a high-threat vehicle or a low-threat vehicle based on the danger evaluation threshold.

[0026] Further, the process of screening the second analyzed image is as follows:

[0027] Obtain the contour-defined rectangle in the first analyzed image, draw a straight line parallel to the zebra stripes and intersecting the short side of the contour-defined rectangle, which is denoted as the zebra straight line, and denote the angle between the zebra straight line and the short side of the contour-defined rectangle as the interaction angle. Each first analyzed image corresponds to an interaction angle. Screen out the first analyzed image with the interaction angle closest to 90° as the second analyzed image.

[0028] Further, the process of generating the warning signal is as follows:

[0029] Obtain the blind area data of the access vehicle. The blind area data includes multiple shadow areas around the access vehicle, which are denoted as invisible areas. The blind area data is determined based on the visible range of the driver's perspective. Denote the contour of the access vehicle as the dangerous contour, and draw the invisible area based on the dangerous contour in each passing image;

[0030] Based on the invisible area of the access vehicle, combined with its driving direction and speed, conduct predictive analysis, delimit the contact risk area around the access vehicle, and identify pedestrians and other vehicles in the contact risk area based on the image recognition algorithm;

[0031] When there are pedestrians and other vehicles in the contact risk area, generate a contact alarm signal.

[0032] Further, the process of obtaining the driving direction of the access vehicle is as follows:

[0033] Obtain the specified driving direction of the lane where the access vehicle is located. When there is only one specified driving direction for the lane where it is located, use this specified driving direction as the driving direction of the access vehicle;

[0034] When there are multiple specified driving directions for the lane where it is located, obtain the steering wheel rotation amplitude angle of the access vehicle in real time, and determine the driving direction of the access vehicle based on the steering wheel rotation amplitude angle.

[0035] Further, when the driving direction of the access vehicle is straight, obtain the invisible area in front of the access vehicle as the first risk area, and obtain the driving speed and vehicle width of the access vehicle as , d, there is a preset reaction time t and a braking influence function f(v), and the braking influence coefficient f(v) is a function of the braking time with respect to the vehicle driving speed v;

[0036] Substitute into the formula for calculation to obtain the length value s. Construct a rectangular second risk area in front of the head of the access vehicle, and use the length value and vehicle width as the two side lengths of the second risk area respectively. Denote the overlapping part of the first risk area and the second risk area as the contact risk area.

[0037] Further, when the driving direction of the access vehicle is turning, calculate the current inner wheel difference corresponding to the access vehicle based on the current steering wheel rotation amplitude angle of the access vehicle;

[0038] Obtain the turning interference area corresponding to the current inner wheel difference of the access vehicle. The turning interference area is in the shape of a crescent formed by the intersection of two arcs. The outer diameter and inner diameter of the turning interference area correspond to the turning arcs of the front inner wheel and the rear inner wheel respectively. Draw the turning interference area in the passing image, and denote the overlapping part of the turning interference area and the invisible area inside the dangerous contour as the contact risk area.

[0039] Further, the process of drawing the turning interference area is as follows:

[0040] Obtain the distance between the rear wheel axle of the access vehicle and the vehicle tail as , obtain the width of the vehicle tail of the access vehicle as , based on , Mark the line segment corresponding to the rear wheel axle in the dangerous contour as the rear wheel axle line segment;

[0041] Denote the inner point of the rear wheel axle line segment as the reference point, and draw the turning interference area in the passing image. The endpoints of the turning interference area coincide with the reference point, and the outer side of the turning interference area is tangent to the inner side of the dangerous contour.

[0042] Further, the process of obtaining the turning interference area corresponding to the inner wheel difference is as follows:

[0043] Obtain the two-dimensional plane model of the access vehicle in the state where the steering wheel rotation amplitude angle is the current amplitude angle. The two-dimensional plane model includes the front wheel line segment AB and the rear wheel line segment CD, where points B and D correspond to the front inner wheel and the rear inner wheel respectively;

[0044] Draw the front wheel turning arc. The center of the front wheel turning arc is located on the straight line where the line segment AB is located. The front wheel turning arc passes through point B, and the center of the front wheel turning arc is located inside the access vehicle;

[0045] Draw the rear-wheel turning arc. The center of the rear-wheel turning arc lies on the straight line where the line segment CD is located. The rear-wheel turning arc passes through point D, and the center of the rear-wheel turning arc lies inside the access vehicle. Denote the closed area formed by the intersection of the front-wheel turning arc and the rear-wheel turning arc as the turning interference area.

[0046] The technical solution provided by the present invention has the following beneficial effects compared with the known prior art:

[0047] 1. The present invention can statistically analyze the vehicle passing data in the intersection based on the image data collected by the image acquisition device at the intersection, evaluate the danger of the intersection, and screen out the intersections with a higher degree of danger as target intersections, so as to conduct targeted preventive monitoring when medium and large-sized vehicles pass through the target intersections, reducing the probability of traffic accidents caused by large-sized vehicles.

[0048] 2. The present invention can delimit a contact risk area around the medium and large-sized vehicle when it passes through the target intersection. The contact risk area refers to the area range where the medium and large-sized vehicle poses a threat to the surrounding pedestrians and other vehicles in the current driving state. By delimiting the contact risk area, it is possible to combine the image acquisition device (video monitoring device) at the intersection for collision risk warning, assisting the driver to make a quick response before an accident occurs, reducing the occurrence of traffic accidents, and improving the intelligent control level of the vehicle and the traffic safety at the intersection. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0050] Figure 1 It is a method step diagram of the present invention;

[0051] Figure 2 It is a schematic diagram of the blind area of a medium and large-sized vehicle;

[0052] Figure 3 It is a schematic diagram of the turning interference area in the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0053] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of 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 based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0054] The present invention will be further described below with reference to the embodiments.

[0055] Refer to Figure 1 , an intelligent auxiliary control method based on an in-vehicle wireless communication chip, at least including:

[0056] Step 1: Divide intersections into target intersections and non-target intersections based on vehicle passing data at different intersections. The target intersections are those with a relatively high passing frequency of medium and large-sized vehicles, and the non-target intersections are other intersections except the target intersections, where:

[0057] The intersection where the collected data is analyzed is denoted as the analysis intersection, and the intersection area within 50 meters of the analysis intersection is denoted as the target area. There is a preset monitoring period, and vehicle passing data within the target area is obtained during the monitoring period. The vehicle passing data includes vehicle threat types and vehicle routes. The vehicle threat types are divided according to the vehicle size and the size of the driving blind area, and the vehicle threat types include low-threat vehicles and high-threat vehicles. The vehicle routes include straight, left turn, and right turn. Obtain the number of straight, left turn, and right turn of high-threat vehicles within the analysis intersection, multiply each by a preset weight coefficient and sum them to obtain the risk assessment value of the analysis intersection. Among them, the weight coefficients corresponding to the number of straight, left turn, and right turn are obtained based on the statistical intersection vehicle traffic accident data. When the risk assessment value is greater than the risk assessment threshold, the analysis intersection is denoted as the target intersection.

[0058] Specifically, the process of obtaining the weight coefficients corresponding to the number of straight, left turn, and right turn is as follows:

[0059] Obtain the accident liability recognition certificates of multiple vehicle traffic accidents, extract the accident occurrence locations, accident vehicle types, and accident processes of the accident liability recognition certificates based on natural language processing (NLP) technology to form an accident data set. The accident occurrence locations, accident vehicle types, and accident processes in the same accident data set are bound to each other. Screen out the accident data set with the accident occurrence location at the intersection and the accident vehicle type as large-sized vehicles as the target data set. Obtain the driving directions of large-sized vehicles in the accident process in the target data set and divide them into straight directions, left turn directions, and right turn directions, and statistically calculate the proportions of different driving directions as the weight coefficients corresponding to the number of straight, left turn, and right turn.

[0060] It should be noted that the accident location refers to the road conditions at the accident location, which is used to determine whether the accident occurred at an intersection. The types of accident vehicles refer to the vehicle types of both or multiple parties involved in the accident. The vehicle types are divided into small passenger cars, large trucks, large trailers, etc., which are used to distinguish whether the vehicle involved in the accident is a small vehicle or a large vehicle. The accident process refers to the driving state of the vehicle at the time of the accident, so as to determine whether the large vehicle was going straight, turning left or turning right at the time of the accident.

[0061] Through statistical data, the proportion of accidents caused by different routes can be analyzed and calculated, that is, the proportion of traffic accidents caused by large vehicles when going straight, turning left and turning right. The larger the calculated weight coefficient, the higher the probability that the large vehicle causes an accident when passing through the intersection in the corresponding driving direction. Therefore, based on the number of times the large vehicle goes straight, turns left and turns right when passing through the analysis intersection and the corresponding weight coefficients, the danger of the analysis intersection can be evaluated, and the analysis intersections with a higher degree of danger (in terms of traffic accidents) are selected and recorded as target intersections, so as to conduct targeted preventive monitoring when the large vehicle passes through the target intersection and reduce the probability of traffic accidents caused by large vehicles.

[0062] Specifically, the process of classifying vehicle threat types is as follows:

[0063] S1: Independently identify different vehicles, record the currently identified vehicle as the target vehicle, obtain multiple top-down images of the target vehicle when passing through the analysis intersection as passing images (the passing images are obtained by regular shooting, and consecutive passing images show the complete process of the target vehicle passing through the analysis intersection), and obtain the target vehicle contour in the passing images as the target contour;

[0064] It should be noted that in the prior art, image acquisition devices are usually installed at the intersections of urban main roads for safety monitoring. Therefore, obtaining the image data of the vehicle when passing through the analysis intersection is a conventional technical means, that is, the passing images of the vehicle are directly available data sources.

[0065] S2: Construct a contour limiting rectangle. The contour limiting rectangle is the smallest rectangle (in terms of area) containing the target contour. Calculate the ratio of the target contour area to the area of the contour limiting rectangle and record it as the compliance index. There is a preset compliance index threshold. Select the passing images with a compliance index greater than the compliance index threshold and the target contour covering the zebra stripe (indicating that the target vehicle is passing through the zebra stripe) as the first analysis images, and further analyze and screen the first analysis images;

[0066] It should be noted that the conformity index to a certain extent reflects the turning radius of the vehicle. Especially for a towed vehicle, when the conformity index is greater than a preset threshold, it usually means that the vehicle is in the straight - running stage (which does not conflict with the vehicle turning through the analysis intersection). That is to say, at this time, the contour - defining rectangle of the vehicle can usually be used as the basis for judging the length and width of the vehicle.

[0067] S3: Obtain the contour - defining rectangle in the first analysis image. Denote the straight line parallel to the zebra - crossing stripes and intersecting with the short side of the contour - defining rectangle as the zebra straight line, and denote the included angle between the zebra straight line and the short side of the contour - defining rectangle as the interaction angle. Each first analysis image corresponds to an interaction angle. Screen out the first analysis image with the interaction angle closest to 90° and denote it as the second analysis image.

[0068] It should be noted that the interaction angle reflects the included angle between the current vehicle attitude of the target vehicle and the zebra - crossing stripes. In the second analysis image obtained by screening, the forward direction of the target vehicle is basically parallel to the zebra - crossing stripes, and the target vehicle is on the zebra crossing. Therefore, at this time, the size of the target vehicle can be evaluated through the zebra - crossing stripes.

[0069] S4: Calculate the area of the contour - defining rectangle in the second analysis image and divide it by the square of the length of the zebra - crossing stripes to obtain the vehicle type evaluation value of the target vehicle (which is used to reflect the vehicle size, and this result is calculated with the zebra crossing as a specific reference object, which is more in line with the actual vehicle size). The calculation formula is expressed as , where S is the area of the contour - defining rectangle in the second analysis image, and L is the length of the zebra - crossing stripes. Obtain the maximum value of the conformity index in all the passing images of the target vehicle and denote it as the blind - area evaluation value. Multiply the blind - area evaluation value and the vehicle type evaluation value by a preset weight coefficient respectively and sum them to obtain the threat evaluation value. Denote the target vehicle with a threat evaluation value greater than the preset dangerous evaluation threshold as a high - threat vehicle, and vice versa as a low - threat vehicle.

[0070] It should be noted that both the blind - area evaluation value and the vehicle type evaluation value are dimensionless numerical values, which respectively reflect the inner - wheel difference when the vehicle turns (usually, the larger the inner - wheel difference, the greater the degree of deviation of the projection contour of the vehicle when turning from a rectangle) and the vehicle volume size. Through the blind - area evaluation value and the vehicle evaluation value, the threat degree of the target vehicle to surrounding vehicles and pedestrians when turning can be comprehensively reflected. Larger - sized vehicles and vehicles with larger inner - wheel differences when turning are more likely to touch other vehicles or pedestrians during the turning process and cause danger. Therefore, the larger the threat evaluation value, the greater the threat degree of the target vehicle to surrounding vehicles and pedestrians when turning.

[0071] It is worth noting that although the vehicle type classification process can quantitatively evaluate the threat level of vehicles, the actual danger warning stage still needs to be specifically analyzed and judged according to the situation of different vehicles, and the vehicle's own structural factors should be fully considered. In other words, vehicle type classification is applicable to the preliminary statistical process of vehicle traffic data in the present invention, helping staff to statistically analyze vehicle traffic data at the intersection based on the image data collected by the image acquisition device, without the need for manual statistics.

[0072] Step 2: Build a wireless communication platform at the target intersection, connect the wireless communication platform to the image acquisition device at the target intersection, establish a wireless communication connection between the wireless communication platform and the on-board wireless communication chip of medium-sized or large vehicles (medium-sized or large vehicles), and transmit data to each other. The wireless communication connection can be compatible with 5G, C-V2X (cellular vehicle-to-everything), DSRC (dedicated short-range communication), Wi-Fi 6, Bluetooth 5.0 and other multi-mode communication methods. In this embodiment, DSRC dedicated short-range communication technology is specifically used to build a wireless communication connection between the wireless communication platform and the on-board wireless communication chip of medium-sized and large vehicles. When the medium-sized and large vehicles enter the target intersection area, they automatically connect to the wireless communication platform and transmit data to each other;

[0073] By building a wireless communication connection between the wireless communication platform and the on-board wireless communication chips of medium and large vehicles, the vehicle control system can be connected to the intelligent monitoring system at the intersection to achieve data interoperability, and then on this basis, the vehicle can gain insight into the surrounding environment and assist the vehicle control system in making intelligent control.

[0074] Step 3: Record the medium and large vehicles connected to the wireless communication platform as connected vehicles. There may be multiple connected vehicles at the same time. Obtain the license plate, appearance and blind spot data of each connected vehicle, obtain the overhead image of the target intersection and record it as the intersection image, identify the connected vehicles in the intersection image based on the image recognition algorithm and mark them, and determine the position of each connected vehicle;

[0075] It should be noted that the access vehicle is identified through its uploaded license plate data and appearance data (when the license plate data and appearance data of any vehicle are the same as those of the access vehicle, the vehicle can be identified as the corresponding access vehicle), and then its specific location is confirmed to facilitate continuous tracking in the subsequent analysis process. Identifying and tracking vehicles through image recognition algorithms is an existing technology and is widely used in the field of traffic violation capture, which will not be elaborated here.

[0076] Step 4: Independently analyze each access vehicle, continuously record the passing images of the access vehicles, determine the travel interference range of the access vehicles based on the passing images in combination with the blind zone data of the access vehicles, identify pedestrians and other vehicles within the driving influence range, and generate a warning signal. By generating the warning signal, it can help the driver react quickly before an accident occurs and avoid traffic accidents caused by the vehicle continuing to drive.

[0077] Among them:

[0078] Obtain the blind zone data of the access vehicle. The blind zone data includes multiple shadow areas surrounding the access vehicle. Denote this shadow area as the invisible area. The blind zone data is determined based on the visible range of the driver's perspective (that is, the area that is invisible around the vehicle through the driver's perspective). Denote the outline of the access vehicle as the dangerous outline, and draw the invisible area based on the dangerous outline in each passing image;

[0079] It should be noted that the blind zone data is calibrated when the vehicle leaves the factory or is tested and calibrated by the staff for each vehicle model. The specific content of the blind zone data is the shadow area around the vehicle contour, as Figure 2 shown, Figure 2 both the dark and light shadow areas around the vehicle contour in

[0080] Obtain the specified driving direction of the lane where the access vehicle is located. When there is only one specified driving direction for the lane where it is located, use this specified driving direction as the driving direction of the access vehicle. When there are multiple specified driving directions for the lane where it is located, obtain the steering wheel rotation amplitude angle of the access vehicle in real time, and determine the driving direction of the access vehicle based on the steering wheel rotation amplitude angle;

[0081] It should be noted that generally, the access vehicle needs to drive along the specified direction of the lane. Therefore, the driving direction of the access vehicle can be quickly determined through the specified driving direction of the lane. On this basis, when the access vehicle is in a two-way or multi-way feasible lane, the driving direction of the access vehicle needs to be further confirmed according to the steering wheel rotation angle of the access vehicle. The two methods are used to detect the driving direction of the access vehicle in a progressive manner, which can improve the detection efficiency and resource occupancy.

[0082] Based on the invisible area of the access vehicle, combined with its driving direction and speed, conduct predictive analysis, delimit the contact risk area around the access vehicle, that is, there is a risk of contact (collision, interference) between the access vehicle and the contact risk area when it is driving. Identify pedestrians and other vehicles in the contact risk area based on the image recognition algorithm. When there are pedestrians and other vehicles in the contact risk area, generate a contact alarm signal and send it to the wireless communication chip of the corresponding access vehicle, triggering the vehicle machine alarm, so as to remind the driver of the access vehicle to decelerate and brake, and avoid traffic accidents between the access vehicle and pedestrians or other vehicles, and avoid causing losses to life and property.

[0083] It should be noted that the contact risk area is delimited instantaneously based on the driving speed and current position of the connected vehicle, and it reflects the blind spot range of the connected vehicle's vision that poses a threat to surrounding pedestrians and other vehicles under the current driving state. If pedestrians and other vehicles are within this area, they cannot be detected by the driver. In this case, if the connected vehicle continues to drive, it will cause a traffic accident and the driver cannot avoid this traffic accident. Therefore, by delimiting the contact risk area, collision risk warnings can be combined with the image acquisition devices (video surveillance devices) at intersections.

[0084] Specifically, the process of delimiting the contact risk area is as follows:

[0085] When the driving direction of the connected vehicle is straight, the invisible area in front of the connected vehicle is obtained and denoted as the first risk area. The driving speed and vehicle width of the connected vehicle are obtained and denoted as and d respectively. There is a preset reaction duration t and a braking influence function f(v). The braking influence coefficient f(v) is a function of the braking time with respect to the vehicle driving speed. f(v0) represents the duration required for the vehicle to brake from the driving speed v0 until the speed drops to 0. The braking influence coefficient is obtained through vehicle braking ability measurement. Substitute it into the formula for calculation to obtain the length value s. Use the length value and the vehicle width as the two side lengths of a rectangle to construct a second risk area in front of the connected vehicle's head. Denote the overlapping part of the first risk area and the second risk area as the contact risk area. One side of the second risk area coincides with the head edge line;

[0086] When the driving direction of the connected vehicle is turning (left turn and right turn), obtain the current steering wheel rotation amplitude angle of the connected vehicle and denote it as the current amplitude angle . There is a preset front inner wheel turning radius function and a rear inner wheel turning radius function . Among them, θ represents the steering wheel rotation amplitude angle. The front inner wheel turning radius function and the rear inner wheel turning radius function are obtained based on the measurement of the turning radius of the inner wheels when the vehicle turns. Substitute the current amplitude angle into the formula to obtain the current inner wheel difference;

[0087] It should be noted that the front inner wheel turning radius function and the rear inner wheel turning radius function respectively represent the functions of the turning radius of the front inner wheels (when turning, the vehicle closer to the inner side is denoted as the inner wheel, the front inner wheel refers to the inner wheel at the front of the vehicle, and the rear inner wheel refers to the inner wheel at the rear of the vehicle) and the rear inner wheels of the connected vehicle with respect to the steering wheel rotation amplitude angle. Taking the front inner wheel turning radius function as an example, represents that when the steering wheel rotation amplitude angle is The turning radius of the front inner wheel before a certain time. For each steering wheel rotation angle, there is a corresponding fixed turning radius for the front inner wheel and the rear inner wheel. The functions of the turning radii of the front inner wheels and rear inner wheels of access vehicles of different sizes and types are different.

[0088] Obtain the turning interference area corresponding to the current inner wheel difference of the access vehicle. The turning interference area is a crescent shape formed by the intersection of two arcs. The outer diameter of the crescent shape corresponds to the turning arc of the front inner wheel, and the inner diameter of the crescent shape corresponds to the turning arc of the rear inner wheel. Obtain the distance between the rear axle connection of the access vehicle and the vehicle tail and record it as and obtain the tail width of the access vehicle and record it as , based on 、 Mark the line segment corresponding to the rear axle connection in the danger profile as the rear axle line segment (the positional relationship of the rear axle line segment in the danger profile is the same as the position of the rear axle connection in the access vehicle, that is, at a certain proportion of the distance from the vehicle tail). Mark the inner point (the inner side of the turn) of the rear axle line segment as the reference point. Draw the turning interference area in the image. The endpoints of the turning interference area coincide with the reference point, and the outer side of the turning interference area is tangent to the inner side of the danger profile. Mark the overlapping part of the turning interference area and the invisible area inside the danger profile as the contact risk area.

[0089] The advantage of using the position of the rear axle connection and the vehicle tail to determine the turning interference area is that no matter how the vehicle is driving, the relative position of its rear axle connection and the vehicle tail is fixed, which makes the determination of the position of the turning interference area unambiguous. That is to say, the inner point of the rear axle line segment is suitable as the reference point for determining the turning interference area.

[0090] As the overlapping part of the turning interference area and the invisible area, the contact risk area further narrows the warning range. That is to say, only pedestrians who are simultaneously in the driver's vision blind spot and the vehicle turning interference area will trigger the warning signal. By specifically limiting the range, the robustness of the signal output can be improved and false alarms can be reduced. Compared with the conventional blind spot radar detection and 360 holographic image detection technologies in the prior art, it has higher accuracy.

[0091] More specifically, the process of obtaining the turning interference area corresponding to the inner wheel difference is as follows:

[0092] Obtain the two-dimensional plane model of the access vehicle when the steering wheel rotation angle is the current angle state, as Figure 3 shown, Figure 3 in which represents the current angle The corresponding offset angle of the vehicle head (compared to the vehicle body). In the two-dimensional plane model, it includes the front wheel line segment AB and the rear wheel line segment CD, where points B and D correspond to the front inner wheel and the rear inner wheel respectively. The front wheel turning arc Y is drawn. The center of the front wheel turning arc is located on the straight line where the line segment AB is located. The front wheel turning arc passes through point B, and the center of the front wheel turning arc is located inside the accessing vehicle. Similarly, the rear wheel turning arc X is drawn. The center of the rear wheel turning arc is located on the straight line where the line segment CD is located. The rear wheel turning arc passes through point D, and the center of the rear wheel turning arc is located inside the accessing vehicle. The closed area formed by the intersection of the front wheel turning arc and the rear wheel turning arc is denoted as the turning interference area (the shaded part in the figure).

[0093] When a large truck turns, there is an "inner wheel difference". The difference in the turning radii of the front inner wheel and the rear inner wheel is commonly known as the "death crescent". If other people or vehicles stay in the inner wheel difference area, the consequences are very serious. By determining the turning interference area, the "death crescent" area corresponding to the current driving state of the accessing vehicle can be determined, so as to conduct early warning analysis for the risks in this area.

[0094] It should be noted that the inner side in the embodiments of the present invention refers to the inner side when the vehicle turns. The relevant data of the vehicle itself, such as the vehicle license plate, appearance, and blind spot data, are uploaded to the wireless communication platform through the in-vehicle wireless communication chip and processed. Constructing a two-dimensional plane model at different rotation angles based on the vehicle structure is a prior art.

[0095] A computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps in the above method are implemented.

[0096] A computer-readable storage medium stores a computer program thereon, and when the computer program is executed by a processor, the steps in the above method are implemented.

[0097] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; 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 recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements will not make the essence of the corresponding technical solutions deviate from the protection scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent auxiliary control method based on a vehicle-mounted wireless communication chip, characterized in that, It includes the following steps: Step 1: Obtain the vehicle passing data of different intersections. The vehicle passing data includes vehicle threat types and vehicle routes. The vehicle threat types are divided into low-threat vehicles and high-threat vehicles according to the vehicle size and the size of the driving blind area. The vehicle size and the driving blind area are quantitatively represented as a vehicle type evaluation value and a blind area evaluation value, which are calculated based on the image analysis when the vehicle passes through the intersection. Obtain the number of straight, left-turn, and right-turn times of high-threat vehicles during the intersection monitoring period, multiply them by the preset weight coefficient and sum to obtain a danger evaluation value. When the danger evaluation value is greater than the danger evaluation threshold, the intersection is recorded as a target intersection. Step 2: Build a wireless communication platform at the target intersection and establish a wireless communication connection between the image acquisition device and medium and large-sized vehicles. Step 3: Record the medium and large-sized vehicles accessing the wireless communication platform as accessed vehicles and identify the accessed vehicles in the image. Step 4: Independently analyze each accessed vehicle to obtain the driving direction of the accessed vehicle. Obtain the blind area data of the accessed vehicle. The blind area data contains multiple shadow areas around the accessed vehicle, and this shadow area is recorded as an invisible area. The blind area data is determined based on the visible range of the driver's perspective. Record the contour of the accessed vehicle as a danger contour, and draw the invisible area based on the danger contour in each passing image. When the accessed vehicle goes straight, construct a contact risk area based on the vehicle speed and width combined with the vehicle blind area data: The invisible area in front of the access vehicle is obtained and recorded as the first risk area, and the driving speed and vehicle width of the access vehicle are obtained and recorded respectively as , d. There is a preset reaction duration t and a braking influence function f(v). The braking influence coefficient f(v) is a function of the braking time with respect to the vehicle driving speed v; Substitute into the formula to calculate the length value s. A second risk area in the shape of a rectangle is constructed in front of the front of the access vehicle. The length value and the vehicle width are used as the two side lengths of the second risk area respectively, and the overlapping part of the first risk area and the second risk area is denoted as the contact risk area; When the accessed vehicle turns, output the corresponding inner wheel difference according to the current steering wheel rotation angle of the accessed vehicle, and output the corresponding turning interference area in response to the input inner wheel difference and construct a contact risk area: Calculate the current inner wheel difference corresponding to the accessed vehicle based on the current steering wheel rotation angle of the accessed vehicle. Obtain the turning interference area corresponding to the current inner wheel difference of the accessed vehicle. The turning interference area is in the shape of a crescent formed by the intersection of two arcs. The outer diameter and inner diameter of the turning interference area correspond to the turning arcs of the front inner wheel and the rear inner wheel respectively. Draw the turning interference area in the passing image, and record the overlapping part of the turning interference area and the invisible area inside the danger contour as the contact risk area. Identify pedestrians and other vehicles in the contact risk area and generate a warning signal.

2. An intelligent auxiliary control method based on an in-vehicle wireless communication chip according to claim 1, characterized in that, The process of obtaining the weight coefficients corresponding to the number of straight, left-turn, and right-turn times is as follows: Obtain the accident liability recognition certificates of multiple vehicle traffic accidents, and extract the accident occurrence locations, accident vehicle types, and accident processes of the accident liability recognition certificates based on natural language processing technology to form an accident data set. Screen out the accident data set with the accident occurrence location at the intersection and the accident vehicle type as large-sized vehicles as the target data set, and obtain the driving directions of large-sized vehicles in the accident process in the target data set. Statistically calculate the proportions of the straight, left-turn, and right-turn directions in the driving directions as the weight coefficients corresponding to the number of straight, left-turn, and right-turn times.

3. An intelligent auxiliary control method based on an in-vehicle wireless communication chip according to claim 1, characterized in that, The process of dividing vehicle threat types is as follows: S1: Record the currently identified vehicle as the target vehicle, obtain multiple top-down images of the target vehicle when passing through the analyzed intersection as passing images, and obtain the contour of the target vehicle in the passing images as the target contour. S2: Construct a contour-defining rectangle, which is the smallest rectangle containing the target contour. Calculate the ratio of the area of the target contour to the area of the contour-defining rectangle and denote it as the compliance index. There is a preset compliance index threshold. Select the processed images in which the compliance index is greater than the compliance index threshold and the target contour covers the zebra stripes as the first analysis images; S3: Based on the position of the contour-defining rectangle in the first analysis images, select the second analysis images in which the vehicle body is closest to being parallel to the zebra stripes; S4: Calculate the area of the contour-defining rectangle in the second analysis images and divide it by the square of the length of the zebra stripes to obtain the vehicle type evaluation value of the target vehicle; Obtain the maximum compliance index among all the processed images of the target vehicle and denote it as the blind area evaluation value; Multiply the blind area evaluation value and the vehicle type evaluation value by a preset weight coefficient respectively and sum them to obtain the threat evaluation value. There is a preset danger evaluation threshold. Based on the danger evaluation threshold, classify the target vehicle as a high-threat vehicle or a low-threat vehicle.

4. An intelligent auxiliary control method based on an in-vehicle wireless communication chip according to claim 3, characterized in that, The process of screening the second analysis images is as follows: Obtain the contour-defining rectangle in the first analysis images. Draw a straight line parallel to the zebra stripes and intersecting the short side of the contour-defining rectangle, which is denoted as the zebra straight line. Denote the included angle between the zebra straight line and the short side of the contour-defining rectangle as the interaction angle. Each first analysis image corresponds to an interaction angle. Select the first analysis image with the interaction angle closest to 90° as the second analysis image.

5. The intelligent auxiliary control method based on an in-vehicle wireless communication chip according to claim 1, wherein The process of obtaining the driving direction of the accessed vehicle is as follows: Obtain the specified driving direction of the lane where the accessed vehicle is located. When there is only one specified driving direction for the lane where it is located, use this specified driving direction as the driving direction of the accessed vehicle; When there are multiple specified driving directions for the lane where it is located, obtain the steering wheel rotation amplitude angle of the accessed vehicle in real time, and determine the driving direction of the accessed vehicle based on the steering wheel rotation amplitude angle.

6. The intelligent auxiliary control method based on an in-vehicle wireless communication chip according to claim 5, characterized in that, The process of drawing the turning interference area is as follows: Obtain the distance between the rear axle of the access vehicle and the vehicle's tail and record it as , obtain the tail width of the access vehicle and record it as , based on , Mark the line segment corresponding to the rear axle in the danger profile as the rear axle line segment; Denote the inner point of the rear wheel axis segment as the reference point, and draw the turning interference area in the processed image. The endpoints of the turning interference area coincide with the reference point, and the outer side of the turning interference area is tangent to the inner side of the dangerous contour.

7. An intelligent auxiliary control method based on an in-vehicle wireless communication chip according to claim 6, characterized in that, The process of obtaining the turning interference area corresponding to the inner wheel difference is as follows: Obtain the two-dimensional plane model of the accessed vehicle in the state where the steering wheel rotation amplitude angle is the current amplitude angle. The two-dimensional plane model includes the front wheel line segment AB and the rear wheel line segment CD, where points B and D correspond to the front inner wheel and the rear inner wheel respectively; Draw the front wheel turning arc. The center of the front wheel turning arc is located on the straight line where the line segment AB is located. The front wheel turning arc passes through point B, and the center of the front wheel turning arc is located inside the accessed vehicle; Draw the rear wheel turning arc. The center of the rear wheel turning arc is located on the straight line where the line segment CD is located. The rear wheel turning arc passes through point D, and the center of the rear wheel turning arc is located inside the accessed vehicle. Denote the closed area formed by the intersection of the front wheel turning arc and the rear wheel turning arc as the turning interference area.

Images