Vehicle guiding method and system based on driving style recognition

Abstract

The invention provides a vehicle guiding method and system based on driving style recognition, and relates to the technical field of road traffic, and the method comprises the steps: obtaining road traffic data; obtaining the congestion level of the road according to the road traffic data; determining a vehicle guiding scheme according to the congestion level, and guiding a vehicle according to the vehicle guiding scheme; the vehicle guiding scheme comprises a lane guiding scheme, and the lane guiding scheme comprises the steps of identifying the driving style of the vehicle, guiding the vehicle to change the lane according to the driving style, and homogenizing the driving style of the vehicle on each lane. The traffic efficiency of the road can be improved.

Description

Technical Field

[0001] The present invention relates to the field of highway traffic technology, and in particular to a vehicle guiding method and system based on driving style recognition. Background Art

[0002] With the increase in the number of private cars, large-scale congestion often occurs on highways. At this time, traffic induction can be performed on vehicles to provide driving guidance for traffic participants, thereby alleviating congestion, ensuring highway traffic efficiency, and reducing vehicle waiting time and fuel consumption. For example, the invention patent with publication number CN115188194A discloses a highway traffic lane-level precision induction system and method, which predicts the degree of safety by combining road surface data and weather forecast data, and obtains safe vehicle speed and safe vehicle distance through environmental data and road traffic data, and generates multi-level and multi-level vehicle induction schemes to reduce upstream congestion pressure. However, due to the different driving styles of different drivers, there are differences in driving habits and reaction speeds in congested sections. The patent only analyzes from a macro level and does not take into account the driving style characteristics of each specific vehicle. There is room for further improvement in guidance efficiency. Summary of the invention

[0003] In order to address the deficiencies of the prior art, the purpose of the present invention is to provide a vehicle guidance method and system based on driving style recognition, so as to guide the vehicle at lane level by identifying the driving style, thereby improving the road traffic efficiency.

[0004] To achieve the above object, according to some embodiments, a first aspect of the present invention provides a vehicle guidance method based on driving style recognition, comprising: Obtain road traffic data; Obtaining a road congestion level according to the road traffic data; Determine a vehicle guidance plan according to the congestion level, and guide the vehicle according to the vehicle guidance plan; The vehicle guidance scheme includes a lane guidance scheme, and the lane guidance scheme includes: identifying the driving style of the vehicle, guiding the vehicle to change lanes according to the driving style, and making the driving style of the vehicles in each lane homogeneous.

[0005] A second aspect of the present invention provides a vehicle guidance system based on driving style recognition, comprising: A road traffic data acquisition module is configured to acquire road traffic data; A congestion level prediction module is configured to obtain a road congestion level according to the road traffic data; The guidance module is configured to determine a vehicle guidance plan according to the congestion level and guide the vehicle according to the vehicle guidance plan; the vehicle guidance plan includes a lane guidance plan, and the lane guidance plan includes: identifying the driving style of the vehicle, guiding the vehicle to change lanes according to the driving style, and making the driving style of the vehicles in each lane homogenized.

[0006] According to a third aspect of the present invention, an electronic device is provided, comprising a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to complete the steps of the above-mentioned vehicle guidance method based on driving style recognition.

[0007] According to a fourth aspect of the present invention, a computer-readable storage medium is provided for storing computer instructions, which, when executed by a processor, complete the steps of the above-mentioned vehicle guidance method based on driving style recognition.

[0008] According to a fifth aspect of the present invention, a computer program product is provided, comprising a computer program / instruction, which, when executed by a processor, implements the steps of the above-mentioned vehicle guidance method based on driving style recognition.

[0009] Compared with the prior art, the present invention has the following beneficial effects: The present invention provides a vehicle guidance method and system based on driving style recognition, which guides the vehicle to change lanes before entering a congested road section by identifying the vehicle's driving style, so that the driving styles of vehicles on each lane are homogenized, thereby improving traffic efficiency. In addition, the present invention further guides the vehicle's speed and distance based on the identified driving style, and provides the vehicle with a recommended speed and distance in a congested road section, so that the vehicles queue up and pass through the congested road section in an orderly manner, thereby improving the road traffic capacity and traffic flow stability, and ensuring the driving safety of highways.

[0010] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0012] Figure 1 Schematic diagram of the method in Embodiment 1 of the present invention; Figure 2 Schematic diagram of long short-term memory network; Figure 3 Schematic diagram of lane division. DETAILED DESCRIPTION

[0013] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0014] Embodiment 1 like Figure 1-Figure 3 As shown, the first embodiment of the present invention provides a vehicle guidance method based on driving style recognition, comprising: S1. Obtain road traffic data; S2. Obtaining a road congestion level according to the road traffic data; S3. Determine a vehicle guidance scheme according to the congestion level, and guide the vehicle according to the vehicle guidance scheme. The vehicle guidance scheme includes a lane guidance scheme, and the lane guidance scheme includes: identifying the driving style of the vehicle, guiding the vehicle to change lanes according to the driving style, and making the driving style of the vehicles in each lane homogeneous.

[0015] In step S1, first, radar equipment is set up on the road sections where speed guidance is required (such as sections prone to congestion and sections prone to accidents), and vehicle trajectory data on the road is collected by lane. Traffic flow, traffic density, occupancy, average speed and other data are extracted from the trajectory data, and weather, date, time period and other information are obtained from the meteorological department, all of which serve as the data basis.

[0016] (1) Traffic flow Traffic flow refers to the number of vehicles passing through a section in a unit of time. The level of traffic volume directly affects the degree of traffic congestion and traffic efficiency, and is one of the important considerations in traffic planning and management.

[0017] (2) Traffic density Traffic density refers to the number of vehicles passing per unit length or unit area. In the field of transportation, traffic density is an important parameter, usually used to describe the degree of traffic congestion or road use efficiency. Traffic density is closely related to traffic flow and can be calculated based on factors such as traffic flow and road width. The traffic density calculation formula is as follows: ; In the formula, K represents traffic density, unit: vehicle / km; N represents the total number of vehicles passing through a certain section of a single lane, unit: vehicle; L represents the total length of the section, unit: km.

[0018] (3) Market share Occupancy rate includes space occupancy rate and time occupancy rate. Among them, space occupancy rate refers to the ratio of the space occupied by vehicles when driving on the road to the total road space. It reflects the density and congestion of vehicles on the road. It is generally the sum of the lengths of vehicles on the road divided by the length of the observed road. Compared with traffic density, space occupancy rate more accurately reflects the utilization rate of the current traffic section.

[0019] ; In the formula, r s Indicates the spatial occupancy rate; L indicates the total length of the road section, unit: km; l i represents the length of the i-th vehicle, unit: km; n represents the number of vehicles passing within time T, unit: vehicle.

[0020] Time occupancy rate refers to the ratio of the time a vehicle spends traveling on a certain road section to the theoretical shortest travel time of that road section. It is generally the ratio of the sum of the time all vehicles spend on the road to the total observation time. Time occupancy rate can reflect the driving efficiency of vehicles. A higher time occupancy rate may indicate traffic congestion or insufficient road capacity.

[0021] ; In the formula, r s represents the spatial occupancy; T represents the observation time, unit: s; t i It represents the time when the i-th vehicle passes through the observation area, unit: s; n represents the number of vehicles passing within the time T, unit: vehicle.

[0022] (4) Weather factors The weather is divided into five levels, namely red warning, orange warning, yellow warning, blue warning and no warning. In this embodiment, the weather quantification factor w is used to quantify the weather factor: if it is a red warning, w=0.9; if it is an orange warning, w=0.7; if it is a yellow warning, w=0.5; if it is a blue warning, w=0.3; if there is no warning, w=0.1.

[0023] (5) Holiday factors Holidays also have a significant impact on travel. In this embodiment, working days and holidays are divided and quantified using a holiday quantization factor h: if it is a working day, h=0.1; if it is a holiday, h=0.9.

[0024] (6) Time period factors During the morning and evening peak hours of weekdays from 7:00 to 00 and from 17:00 to 20:00, the time period factor is 0.9; during daytime driving from 9:00 to 17:00, the time period factor is 0.7; during other time periods, the time period factor is 0.1. During holidays, during daytime driving from 8:00 to 20:00, the time period factor is 0.7, and during other time periods, the time period factor is 0.1.

[0025] (7) Average vehicle speed The average speed of vehicles on the road refers to the ratio of the sum of the speeds of all vehicles on the road to the number of vehicles.

[0026] ; Among them, V s is the average speed, unit: km / h; N is the total number of vehicles on the road, unit: vehicle; v i is the speed of the ith vehicle, in km / h.

[0027] In this embodiment, the above parameters are used as input features of the training set and the test set of the subsequent congestion prediction model. The input vector is X=(x1,x2,x3,x4,x5,x6,x7), where x1 is the traffic flow, x2 is the traffic density, x3 is the space occupancy, x4 is the time occupancy, x5 is the weather factor, x6 is the holiday factor, and x7 is the time period factor; the output of the congestion prediction model is the average vehicle speed, and the congestion prediction model is used to predict the road congestion situation, thereby dividing the congestion level.

[0028] Traffic congestion is divided into three levels based on average vehicle speed. Level 1: The average vehicle speed on the road is equal to the free-flowing speed V f , at this time, the road traffic flow is free flow and no guidance is required; Level 2, the average speed on the road is lower than the free-flow speed V f But above the critical speed V m (the speed when the theoretical capacity of the road is the largest), at this time the road traffic flow is a stable flow; level three, the average speed on the road is less than or equal to the critical speed V m At this time, the road traffic flow is mandatory. At level 2 and level 3, vehicle guidance is required.

[0029] Critical speed V m It can be calculated based on the established relationship model between flow and speed. The smooth travel speed V under different weather conditions is obtained from the trajectory data. f and the blocking density K j , establish the relationship model between flow and speed: ; Among them, Q is the actual flow rate, unit: vehicle / h; K j is the blocking density, i.e., the density when the traffic is so dense that vehicles cannot move, in units of vehicles / km; V is the actual vehicle speed, i.e., the average speed of vehicles on the road, in units of km / h; V f The free-flow speed is the average speed when the traffic density approaches zero. The unit is km / h.

[0030] Establish the relationship model between flow and speed under various conditions, and then solve the critical speed V under different conditions m , that is, the speed when the flow reaches its maximum.

[0031] In step S2, the road congestion level (level 1, level 2, level 3) is predicted based on the road traffic data obtained in step S1. In this embodiment, a long short-term memory network (LSTM) is used to predict road congestion. The long short-term memory network predicts traffic congestion by capturing dependent features. It solves the gradient vanishing problem based on a gating mechanism. The long short-term memory network consists of an input layer, an output layer, and a recursive hidden layer with artificially designed memory units. The recursive hidden layer can remove and retain information about the unit state, including three gates: an input gate, an output gate, and a forget gate, such as Figure 2 shown.

[0032] The processing process of the long short-term memory network can be expressed as follows: ; ; ; ; ; ; Among them, i t represents the input gate, f t represents the forget gate, g t represents the updated value of the unit state, o t is the output gate, c t is the cell state at time t, h t is the hidden layer output at time t, x t is the input at time t, h t-1 is the output of LSTM at the last moment, is the deviation vector, and is the weight matrix, is the Hadamard product (i.e., the product of the corresponding position elements of the two matrices), is an activation function (Sigmoid function). The forget gate determines what is related to the previous time step, the input gate determines what information is added from the current time step, and the output gate determines what the next hidden state should be.

[0033] The input of the training set and test set of the long short-term memory network is X=(x1,x2,x3,x4,x5,x6,x7), where x1 is the traffic flow, x2 is the traffic density, x3 is the space occupancy, x4 is the time occupancy, x5 is the weather factor, x6 is the holiday factor, and x7 is the time period factor. According to the actual data adjustment, the batch size is generally set to 4, and the number of iterations can be set to 800. Finally, the congestion prediction model is obtained by training based on historical road traffic data. The road traffic data collected in real time is input into the congestion prediction model to obtain the prediction of the average vehicle speed and congestion level.

[0034] In step S3, if the congestion level is level 1, no guidance is required; if the congestion level is level 2 or level 3, a vehicle guidance plan is determined according to the congestion level.

[0035] In this embodiment, the vehicle guidance scheme includes a lane guidance scheme, a vehicle speed guidance scheme and a vehicle distance guidance scheme. To improve the guidance efficiency, this embodiment first identifies and divides the driving style of the vehicle, and then executes different vehicle guidance schemes according to different driving styles.

[0036] The driving trajectory and vehicle characteristic data of all vehicles in the shooting area are obtained frame by frame (0.1s / frame) by using roadside communication facilities and camera video shooting. Each data sample includes 18 parameters, including vehicle number, data frame number, total data frame, standard time, X coordinate value of the acquisition area coordinate system, Y coordinate value of the acquisition area coordinate system, X index value of the standard geographic coordinate system, Y coordinate value of the standard geographic coordinate system, vehicle length, vehicle width, vehicle type, vehicle speed, vehicle acceleration, lane number, number of the leading vehicle following, number of the following vehicle following, headway between vehicles and headway time.

[0037] The identification section used in this embodiment is a three-lane straight section, and only an identification device needs to be added to the section without any other modification.

[0038] According to literature survey, almost all normal lane-changing vehicles can complete a single lane change within 16 s, so 32 s is selected as the time length of the experimental data, that is, the time it takes for the vehicle to change lanes twice.

[0039] For the selection of driving style evaluation indicators, the more evaluation indicators there are, the more they can comprehensively characterize the driver's driving style. This embodiment selects 23 evaluation indicators to comprehensively characterize the driver's driving style, and finds evaluation indicators that can better characterize the driving style. The selected evaluation indicators specifically include: maximum longitudinal speed, mean longitudinal speed, standard deviation of longitudinal speed, maximum lateral speed, mean lateral speed, standard deviation of lateral speed, maximum absolute value of acceleration, mean absolute value of acceleration, standard deviation of absolute value of acceleration, minimum headway, mean headway, standard deviation of headway, minimum following distance, mean following distance, standard deviation of following distance, maximum absolute value of acceleration impact, mean absolute value of acceleration impact, standard deviation of absolute value of acceleration impact, free lane change time, number of free lane changes, proportion of lane change time, distance traveled within 32s, and proportion of time above the overall vehicle speed average.

[0040] The vehicle speed, acceleration, following distance and headway distance can be directly obtained from the data set. The calculation of other indicators is as follows: ; ; ; ; ; ; Among them, v yt is the lateral velocity; y t is the lateral displacement at time t; is the time interval between (t-1) and t, which is 0.1s; v m is the overall vehicle speed average, that is, the average vehicle speed within the selected 32s; is the longitudinal displacement of the i-th vehicle in the last frame; is the longitudinal displacement of the i-th vehicle at the first frame; s i is the driving distance within 16 seconds; t is the lane change time; t1 and t2 are the absolute values ​​of the lateral speed that begin to be less than The time point when the distance between the vehicle and the lane change lane line is greater than half of the vehicle width, t1 is the search from the lane change moment to the front, and t2 is the search from the lane change moment to the back; J t is the acceleration impact degree; a t is the acceleration at time t; X 23 The vehicle speed is greater than v m The ratio of the number of frames to 320; It is the maximum distance of all observed vehicles traveling for 32 seconds.

[0041] Since there are many categories of characteristic parameter statistics obtained by preliminary selection and calculation, there is a tendency for information overlap between parameters, which leads to unsatisfactory results in subsequent analysis and increases the amount of calculation, so it is necessary to reduce the dimension of the data. While maximizing the retention of the original variable information, the original complex and multidimensional variables are converted into new low-dimensional variables through linear combination. This process is called data dimensionality reduction. In this embodiment, factor analysis is used to reduce the dimension of the original 23 characteristic parameters to obtain 5 common factors.

[0042] After rotating the factor load matrix, the rotated component matrix can be obtained. The common factors are named according to the matrix. The load on the average speed, speed variance and headway distance average is large, which is named longitudinal speed factor M1; the load on the average lateral speed, lateral speed variance, impact degree average and impact degree variance is large, which is named lateral factor M2; the load on the headway distance average and variance is large, which is named following factor M3; the load on lane change time is large, which is named lane change factor M4. The load coefficient on acceleration variance is large, and acceleration variance can show vehicle speed fluctuation, which is named speed fluctuation factor M5.

[0043] Cluster analysis is performed on the five common factors to obtain classification results of different driving styles. In this embodiment, the driving styles are divided into three categories: aggressive, general and cautious. Accordingly, during cluster analysis, the number of cluster centers k is 3, that is, clustered into three categories.

[0044] Since the initial cluster center of the K-mean clustering algorithm is randomly selected, it needs to be iterated multiple times during clustering to weaken the instability associated with random initialization, which will directly affect the final clustering result. In order to avoid this problem, this embodiment uses the K-means++ algorithm for clustering. The core idea of ​​the K-means++ algorithm is to select data points as far away as possible as the initial cluster center point, which can effectively solve the problem that the clustering result is too dependent on the selection of the initial cluster center.

[0045] set up is the original data set. Now the data set is divided into , the specific calculation steps for initializing the cluster center point in the K-means++ algorithm are as follows: Step 1: Randomly select k sample points from the data set as the initial cluster center C i ; Step 2: Calculate the distance between each sample point and the initial cluster center C i The shortest Euclidean distance between them is represented by D(x); then the probability of each sample point being selected as the next cluster center is calculated. ;Finally, use the roulette wheel method as the selection rule to select the next cluster center; Step 3: Repeat step 2 until k cluster centers are selected.

[0046] The k-means++ algorithm is used to perform cluster analysis on the five factors, and the number of cluster centers k is defined as 3, so that the driving styles are divided into aggressive, general, and cautious types.

[0047] For the longitudinal speed factor M1, the larger the M1 is, the more likely the driver is to choose a higher speed, and the driving style is more aggressive; conversely, the driving style is more cautious. For the lateral speed factor M2, the larger the M2 is, the more likely the driver is to choose a higher speed to change lanes, intending to overtake or enter the high-speed lane, and the driving style is more aggressive. For the following factor M3, the larger the M3 is, the closer the driver is to the vehicle in front, and the driving style is more aggressive. For the lane change factor M4, the larger the M4 is, the more frequently the driver changes lanes, the more time the lane change takes, and the driving style is more aggressive. For the speed fluctuation factor M5, the larger the M5 is, the more likely the driver is to use the accelerator and brake more frequently, and the driving style is more aggressive. In summary, the larger the five factors are, the more aggressive the driver's driving style is, and the smaller the five factors are, the more cautious the driver's driving style is. This rule can be used to distinguish and verify the classification of driving styles.

[0048] A fully connected neural network model (FNN) is used to identify driving styles. First, the driving style is defined using the above method, and the corresponding training set and validation set are obtained based on historical data. The fully connected neural network model is trained using the BP algorithm to obtain a driving style recognition model.

[0049] Considering that almost all normal lane-changing vehicles can complete a single lane change within 16 seconds, 32 seconds is enough for the driver to change lanes twice in succession. However, in reality, continuous lane changes are rare. Therefore, four longest lane-changing times are reserved, and a length of The road section is used as the identification section. For each vehicle entering the identification section, the driving style recognition model is used to identify its driving style before it enters the lane change section, and then the corresponding lane guidance scheme is executed after the vehicle enters the lane change section.

[0050] Before the vehicle enters the target section, there are two sections, namely the identification section and the lane-changing section; the identification section is the section where historical vehicle trajectory data is collected, and the lane-changing section is the section where decision-making is executed. ) performs lane guidance and determines whether to execute the lane guidance plan based on the congestion level. When the congestion level is level one, no guidance is performed. When the congestion level is level two or level three, lane change guidance is performed.

[0051] The recognition section has three lanes, and the FNN model is used for learning to identify the driving style of the vehicle; the change section has three lanes, such as Figure 3As shown in the figure, they are the aggressive lane L1, the general lane L2 and the cautious lane L3. Before entering the lane change section, the FNN model can already identify the driving styles of all vehicles. And notify each vehicle through the roadside communication equipment, prompting the driver to prepare for the lane change.

[0052] Considering that the lane-changing section will frequently prompt drivers to change lanes, if continuous lane changes occur in the lane-changing section, the risk of traffic conflict will increase and the possibility of congestion will increase. Therefore, when vehicles enter the lane-changing section, all vehicles are considered to change lanes only once, that is: Lane change guidance suggestions for vehicles with three driving styles in the aggressive lane L1: vehicles with aggressive and cautious driving styles do not change lanes, and vehicles with normal driving styles change lanes to L2; Lane change guidance suggestions for vehicles with three driving styles in the general lane L2: general style vehicles do not change lanes, aggressive style vehicles change lanes to L1, and cautious style vehicles change lanes to L3; There are three driving styles in the cautious lane L3: aggressive driving style and cautious driving style vehicles do not change lanes, and general style vehicles change lanes to L2.

[0053] Through lane change guidance, the driving style in each lane is homogenized, and speed and vehicle distance guidance can be easier to implement after entering congested sections, thereby improving traffic efficiency in congested sections.

[0054] The vehicle guidance scheme also includes a vehicle speed guidance scheme and a vehicle distance guidance scheme. After completing the lane change guidance, the vehicle speed and vehicle distance are further guided.

[0055] If the congestion level is level 3, the road is very saturated and there are a lot of vehicles. When the vehicle density on the road exceeds the road capacity, the traffic congestion is serious and the speed drops sharply, it usually means that the road service level is poor, the vehicle traffic is difficult, the traffic efficiency is low, and it belongs to the category of forced flow. At this time, the critical speed V is used. m It is the target value for vehicle speed guidance.

[0056] If the congestion level is level 2, the traffic flow is stable and the average vehicle speed V is used. s It is the target value for vehicle speed guidance.

[0057] The vehicle distance guidance uses the following model to determine the guidance value of the vehicle distance. The general formula of the following model is: ; in, is sensitivity; is the acceleration of the car n+1 after the delay T, in m / s 2 ; is the proportional constant; m and l are constants; T is the reaction time, ranging from 1.0 to 2.2 seconds. It needs to be adjusted in different weather conditions. It is best to conduct field tests with multiple people in different weather conditions and take the median of the data; is the speed of the n+1th car after the delay T time, in m / s; X n (t) is the position of the nth vehicle (lead vehicle) at time t; X n+1 (t) is the position of the n+1th vehicle (the following vehicle) at time t; is the speed of car No. n at time t, in m / s; is the speed of car n+1 at time t, in m / s.

[0058] When m=0 and l=0, it is a linear following model. This embodiment uses the linear following model and introduces a reaction sensitivity coefficient to enhance the guidance effect based on the classification of vehicles with different driving styles by lane guidance. The linear following model in this embodiment is expressed as: ; ; Where, L is the headway distance in the blocked condition; The reaction sensitivity coefficient usually refers to the sensitivity of the driver of the following vehicle to adjust the state of his own vehicle (such as acceleration or deceleration) after sensing the change of the state of the front vehicle (such as speed change). This coefficient reflects the driving behavior characteristics of the driver, including factors such as the driver's attention, judgment, and reaction speed.

[0059] Based on the collected historical trajectory data, a large number of following vehicle pairs in the three lanes of the interchange section are analyzed respectively in the aggressive lane L1, the general lane L2, and the cautious lane L3. The aggressive lane reaction sensitivity coefficients of the vehicles in the corresponding lanes under different weather conditions are calibrated through genetic algorithms. , General lane reaction sensitivity coefficient and cautious lane response sensitivity coefficient .

[0060] The stability of the linear car-following model is calculated using the following formula:

[0061] Where C represents the value of the swing characteristic of the distance between the two vehicles. The larger C is, the greater the swing of the distance value is; the smaller C is, the distance swing tends to zero. , the spacing swing is: no swing, basically stable; if , the spacing swing is: attenuation swing; if , the spacing swing is: non-attenuated swing; if , the spacing swing situation is: the swing amplitude increases.

[0062] If the stability C of the linear model is less than , it means that the traffic flow is stable and the linear following model is applicable to this lane under this weather condition on this road section; if the stability C value of the linear model is greater than , it means that the traffic flow is unstable and the linear following model is not applicable in this lane under this weather condition on this road section.

[0063] According to the linear following model, the target value of vehicle distance guidance is obtained: ; If the predicted congestion level of a road during a certain period of time is level one, the traffic flow on the road at this time is free flow, and there is no need for speed and distance guidance.

[0064] If it is level 2, the traffic flow is in the range of stable flow. The speed and distance between vehicles are guided to improve the stability of the traffic flow on the road and avoid the harm caused by the decrease of stability. According to the average speed V of the traffic flow on the road, s , as the target of vehicle speed guidance, the target value of vehicle distance guidance is The roadside radar monitors the position of vehicles on the road in real time, and the roadside communication equipment sends a signal to the guide vehicle to follow the critical speed V m The radar will then continue to monitor the following vehicle. If the distance between the following vehicle and the leading vehicle is not , the following vehicle is notified to adjust the distance, and the expected distance and current distance are provided to the driver in real time. After the adjustment is completed, V m That's it.

[0065] If it is level 3, the traffic flow is in a forced flow state, and the critical speed V m is the target of vehicle speed guidance, and the target value of vehicle distance guidance is , guiding vehicles to maintain the same speed and distance to improve traffic flow stability.

[0066] Embodiment 2 This embodiment provides a vehicle guidance system based on driving style recognition, including: A road traffic data acquisition module is configured to acquire road traffic data; A congestion level prediction module is configured to obtain a road congestion level according to the road traffic data; The guidance module is configured to determine a vehicle guidance plan according to the congestion level and guide the vehicle according to the vehicle guidance plan; the vehicle guidance plan includes a lane guidance plan, and the lane guidance plan includes: identifying the driving style of the vehicle, guiding the vehicle to change lanes according to the driving style, and making the driving style of the vehicles in each lane homogenized.

[0067] It should be noted here that the various modules in this embodiment correspond one-to-one to the steps of the method in Example 1, and the specific implementation process is the same, which will not be repeated here.

[0068] Embodiment 3 This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to complete the steps of the method in the first embodiment.

[0069] Embodiment 4 This embodiment provides a computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, the steps of the method in the first embodiment are completed.

[0070] Embodiment 5 This embodiment provides a computer program product, including a computer program / instruction, which implements the steps of the method in Embodiment 1 when executed by a processor.

[0071] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A vehicle guidance method based on driving style recognition, characterized in that: include: Obtain road traffic data; Obtaining a road congestion level according to the road traffic data; Determine a vehicle guidance plan according to the congestion level, and guide the vehicle according to the vehicle guidance plan; The vehicle guidance scheme includes a lane guidance scheme, and the lane guidance scheme includes: identifying the driving style of the vehicle, guiding the vehicle to change lanes according to the driving style, and making the driving style of the vehicles in each lane homogeneous.

2. A vehicle guidance method based on driving style recognition as claimed in claim 1, characterized in that: The driving styles include aggressive, general and cautious; the driving styles are obtained by clustering driving style characterization indicators collected when the vehicle is driving on the identified road section within a set time length.

3. A vehicle guidance method based on driving style recognition as claimed in claim 2, characterized in that: The method of guiding vehicles to change lanes according to driving styles to homogenize the driving styles of vehicles in each lane includes: dividing the lanes into an aggressive lane, a general lane, and a cautious lane that are adjacent to each other in sequence; guiding vehicles with a general driving style in the aggressive lane to the general lane, guiding vehicles with an aggressive driving style in the general lane to the aggressive lane, guiding vehicles with a cautious driving style in the general lane to the cautious lane, and guiding vehicles with a general driving style in the cautious lane to the general lane.

4. The vehicle guidance method based on driving style recognition as claimed in claim 1, characterized in that: The congestion level of the road is divided according to the average speed of vehicles on the road.

5. The vehicle guidance method based on driving style recognition as claimed in claim 1, characterized in that: The vehicle guidance scheme also includes a vehicle speed guidance scheme, which includes: if the road traffic flow is a stable flow, the average speed of vehicles on the road is used as the target value for vehicle speed guidance to guide the vehicles; if the road traffic flow is a forced flow, the critical speed of the road is used as the target value for vehicle speed guidance to guide the vehicles.

6. The vehicle guidance method based on driving style recognition as claimed in claim 1, characterized in that: The vehicle guidance scheme also includes a vehicle distance guidance scheme, which includes: if the road traffic flow is a stable flow, the target value of the vehicle distance guidance is calculated based on the average speed of vehicles on the road, the reaction sensitivity coefficient of the lane where the vehicle is located, and the distance between the heads of the vehicles; if the road traffic flow is a forced flow, the target value of the vehicle distance guidance is calculated based on the critical speed of the road, the reaction sensitivity coefficient of the lane where the vehicle is located, and the distance between the heads of the vehicles.

7. A vehicle guidance system based on driving style recognition, characterized in that: include: A road traffic data acquisition module is configured to acquire road traffic data; A congestion level prediction module is configured to obtain a road congestion level according to the road traffic data; a guidance module, configured to determine a vehicle guidance plan according to the congestion level, and guide the vehicle according to the vehicle guidance plan; The vehicle guidance scheme includes a lane guidance scheme, and the lane guidance scheme includes: identifying the driving style of the vehicle, guiding the vehicle to change lanes according to the driving style, and making the driving style of the vehicles in each lane homogeneous.

8. An electronic device, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to complete the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: Used to store computer instructions, which, when executed by a processor, complete the steps of the method according to any one of claims 1 to 6.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images