A lane selection method for autonomous driving vehicles based on driving style

Abstract

The present invention discloses a method for lane selection of an autonomous driving vehicle based on driving style, comprising the following steps: based on traffic signs and markings, obtaining a candidate area for vehicle driving; based on the vehicle speed index and the front driving space index in the candidate area, obtaining the traffic efficiency; based on the risk index, obtaining the risk assessment value in the candidate area; making a comprehensive decision on the traffic efficiency and the risk assessment value to construct a lane selection model; combining the lane selection model with the driving style to obtain the target driving area. The lane selection method of the present invention hierarchically considers the legality, efficiency, and safety of lane selection, and takes into account the influence of different driving styles on the target lane selection, reflecting the process of the driver's intention to change lanes.

Description

Technical Field

[0001] The present invention belongs to the field of automobile automatic control, and in particular relates to a lane selection method for an automatic driving vehicle based on driving style. Background Art

[0002] With the increase in the number of motor vehicles, traffic congestion and traffic conflicts are becoming increasingly serious, which will inevitably lead to an increase in the number of road traffic accidents. Among them, lane change is one of the main factors affecting driving safety, and lane selection is an important part of lane change. Among the existing autonomous driving technologies, there are few detailed analyses of the lane selection module. In the lane selection process, it is necessary to comprehensively analyze various information and make decisions within a certain period of time. The rationality of lane selection is directly related to the efficiency and safety of the driving process. To this end, it is necessary to propose a lane selection method for autonomous driving vehicles that takes driving style into consideration. Summary of the invention

[0003] The purpose of the present invention is to provide a lane selection method for an autonomous driving vehicle based on driving style to solve the problems existing in the above-mentioned prior art.

[0004] To achieve the above object, the present invention provides a lane selection method for an autonomous driving vehicle based on driving style, comprising the following steps:

[0005] Based on traffic signs and markings, obtain candidate areas for vehicle driving;

[0006] Obtaining traffic efficiency based on a vehicle speed index and a front driving space index in the candidate area;

[0007] Acquire a risk assessment value within the candidate area based on the risk indicator;

[0008] Making a comprehensive decision on the traffic efficiency and risk assessment value, and building a lane selection model;

[0009] The lane selection model is combined with the driving style to obtain a target driving area.

[0010] Preferably, the risk indicators include the proportion of large vehicles, the standard deviation of vehicle speed, and the standard deviation of the lateral position of the vehicle relative to the lane line.

[0011] Preferably, the process of obtaining the vehicle speed index of the candidate area includes obtaining the driver's first sensitivity to gain or loss based on the driving speed of the target vehicle, the average speed of vehicles in the candidate area, and the driver's attitude towards risk; presetting a sensitivity threshold, and comparing the driver's first sensitivity to gain or loss with the sensitivity threshold respectively to obtain the vehicle speed index of the candidate area.

[0012] Preferably, the process of obtaining the front driving space index of the candidate area includes obtaining the driver's second sensitivity to gain or loss based on the closest distance between the vehicle and the target vehicle in the candidate area, the maximum range when selecting the lane, and the driver's attitude towards risk; presetting a sensitivity threshold, and comparing the driver's second sensitivity to gain or loss with the sensitivity threshold respectively to obtain the front driving space of the candidate area.

[0013] Preferably, the process of obtaining the traffic efficiency includes obtaining the traffic efficiency in the candidate area based on the vehicle speed index of the candidate area and its weight coefficient, and the front driving space index of the candidate area and its weight coefficient.

[0014] Preferably, the process of obtaining the risk assessment value within the candidate area includes obtaining the risk assessment value within the candidate area based on the total number of vehicles in the candidate area, the number of large vehicles, the standard deviation of vehicle speed, the standard deviation of the lateral position of the vehicle relative to the lane line, and the total number of vehicles within the perception range of the target vehicle.

[0015] Preferably, the process of making a comprehensive decision on the traffic efficiency and risk assessment value includes obtaining and sorting the comprehensive decision values ​​of different candidate areas based on the probability weight of the risk probability in the candidate area and the value function of the vehicle speed index in the candidate area, and selecting the candidate area with the largest comprehensive decision value as the target driving area.

[0016] Preferably, the process of obtaining the probability weight of the risk probability in the candidate area includes, based on the risk probability of the candidate area, the vehicle speed index of the candidate area, presetting the first parameter and the second parameter, and obtaining the probability weight of the risk probability in the candidate area.

[0017] Preferably, the process of obtaining the value function of the vehicle speed index in the candidate area includes obtaining the driver's third sensitivity to gain or loss based on the vehicle speed index of the candidate area and the driver's attitude towards risk; presetting a sensitive threshold, and comparing the driver's third sensitivity to gain or loss with the sensitive threshold respectively to obtain the value function of the vehicle speed index in the candidate area.

[0018] The technical effects of the present invention are:

[0019] The lane selection method of the present invention obtains candidate driving areas according to traffic signs and markings, obtains traffic efficiency and risk assessment values ​​within the candidate areas, decouples the corresponding legality, high efficiency, and safety evaluation indicators, and selects weight coefficients for comprehensive evaluation based on different driving style types, so that the present invention can optimize the ideal area for future driving according to different driving styles, while having the advantages of legal, efficient, and safe driving. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0021] Figure 1 is a flow chart of a lane selection method in an embodiment of the present invention;

[0022] Figure 2 A schematic diagram of the division of traffic signs and road markings in an embodiment of the present invention;

[0023] Figure 3 It is a division diagram of the candidate area in the embodiment of the present invention. DETAILED DESCRIPTION

[0024] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0025] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0026] Embodiment 1

[0027] like Figure 1-3 As shown, this embodiment provides a method for lane selection of an autonomous driving vehicle based on driving style. The input of the lane selection model in this embodiment is the dynamic driving environment around the vehicle and the driving style of the driver of the vehicle, and the output is the target area. The specific process includes:

[0028] 1.Legality Assessment

[0029] Legality assessment is to deal with traffic regulations and ensure the legality of decision-making results.

[0030] First, the driving environment of the car is divided into regions according to traffic regulations to create candidate regions. Traffic signals include traffic lights, traffic signs, traffic markings and traffic police's instructions. Traffic signs and traffic markings do not conflict with traffic lights. Without considering traffic police, the traffic regulations here mainly refer to traffic signs and markings. The division of traffic signs and markings is as follows: Figure 2 shown.

[0031] When creating candidate areas, it is necessary to ensure that the traffic markings and signs in each candidate area are the same, that is, the line type and traffic signs in the candidate area are consistent, to avoid the situation where one lane is solid and the other is dotted, one part is a motor vehicle lane and the other is a bus lane. Figure 3 In the case of a vehicle, the horizontal direction (X) is first divided into three areas with three lanes. On this basis, the longitudinal direction (Y) is divided according to traffic regulations to make the traffic elements in the same area consistent. In other words, there are five candidate areas. When creating candidate areas, a special case should be considered - the area completely behind the vehicle. Because the vehicle cannot change lanes to the rear of the vehicle, this area cannot be classified as a candidate area.

[0032] 2. Traffic efficiency evaluation

[0033] The traffic flow attributes of each drivable area are counted, and areas with sparse traffic flow are given priority to ensure higher traffic efficiency.

[0034] When driving, drivers usually hope to reach their destination as quickly as possible and reduce driving time while ensuring safety and legality. To this end, they usually choose areas with higher traffic efficiency. The traffic efficiency of the candidate area takes into account the combined influence of the vehicle speed in the candidate area and the driving space ahead.

[0035] The traffic efficiency function established by this embodiment taking into account the vehicle speed in the candidate area and the front driving space is as follows:

[0036]

[0037] Among them, E area (i) represents the traffic efficiency of candidate area i, w v represents the vehicle speed index and its weight coefficient of candidate area i, w d Represents the front space index and its weight coefficient of the vehicle in candidate area i. If the candidate area is a special area such as a crosswalk, a dedicated lane for buses, HOVs, or a non-motorized vehicle lane, the traffic efficiency is 0.

[0038] In this function, the values ​​of parameters can be used to reflect the attitudes of different decision makers towards losses, gains and risks, that is, the attitudes of drivers with different driving styles towards losses, gains and risks.

[0039] The vehicle speed index function is as follows:

[0040]

[0041] in, is the average speed of vehicles in candidate area i, v egois the speed of the vehicle. ξ and θ reflect the driver's sensitivity to gains or losses. If the driver is more sensitive to gains than losses, then θ=1, ξ>1; if the driver is more sensitive to losses than gains, then ξ=1, θ>1. α and β reflect the driver's attitude towards risk. If the driver is a conservative driver, then α, β>1; if the driver is a general driver, then α, β=1; if the driver is an aggressive driver, then 0<α, β<1.

[0042] The calculation formula for the average vehicle speed in candidate area i is as follows:

[0043]

[0044] Among them, v ij is the speed of the jth vehicle in candidate area i, and N is the total number of vehicles in candidate area i.

[0045] The front space indicator function is as follows:

[0046]

[0047] Among them, d i is the distance between the nearest vehicle in front of candidate area i and the vehicle. If there is no vehicle in front, then d i is the maximum distance from the vehicle to the front of candidate area i after it moves to candidate area i, d limit is the maximum range considered when making decisions. ξ and θ reflect whether the driver is more sensitive to gains or losses. If the driver is more sensitive to gains than losses, then θ=1, ξ>1; if the driver is more sensitive to losses than gains, then ξ=1, θ>1. α and β reflect the driver's attitude towards risk. If the driver is a conservative driver, then α, β>1; if the driver is a general driver, then α, β=1; if the driver is an aggressive driver, then 0<α, β<1.

[0048] 3. Risk Assessment

[0049] Risk assessment is to ensure the safety of decision-making results.

[0050] The risk assessment module objectively assesses the potential risk of entering each candidate area. The main sources are two: the proportion of large vehicles and the proportion of vehicles with strange behaviors. The risk assessment function is as follows:

[0051]

[0052] Among them, R area (i) is the risk in candidate region i, k i is the number of large vehicles in candidate area i, S is the total number of vehicles within the sensing range, N is the total number of vehicles in candidate area i, is the standard deviation of vehicle j’s speed in candidate area i, is the standard deviation of the lateral position of vehicle j in candidate area i relative to the lane line.

[0053] The calculation formula is as follows:

[0054]

[0055] Among them, v m is the speed of the mth vehicle in candidate area i, is the average speed of vehicles in region i, and n is the total number of vehicles in candidate region i.

[0056] The calculation formula is as follows:

[0057]

[0058] Among them, l m is the lateral position of the mth vehicle relative to the lane line in candidate area i, is the mean lateral position of the vehicles in candidate area i relative to the lane line, and n is the total number of vehicles in candidate area i.

[0059] For each drivable area, the risk of entering is measured by the proportion of large vehicles and vehicles with strange behaviors, and priority is given to areas with lower risks to ensure higher safety.

[0060] 4. Comprehensive decision-making

[0061] Convert the results of risk assessment into risk probability within the candidate area.

[0062] The risk probability of candidate region i is:

[0063]

[0064] Among them, R area (i) is the risk in candidate region i, and N is the number of candidate regions

[0065] According to the probability weight function and the value function, the final comprehensive decision formula is obtained, and the order is sorted according to the size of the V value. The larger the V value, the greater the prospect, and the area with the larger prospect is the output target area.

[0066] The comprehensive decision-making formula is as follows:

[0067] V i =w(p i )v(E area (i))

[0068] Among them, ω is the probability weight function and v is the value function.

[0069] The calculation formula of the value function is as follows:

[0070]

[0071] in, is the vehicle speed index of candidate area i, θ, ξ reflects the driver's sensitivity to gains or losses. If the driver is more sensitive to gains than losses, then θ=1, ξ>1; if the driver is more sensitive to losses than gains, then ξ=1, θ>1. α, β reflect the driver's attitude towards risk. If the driver is a conservative driver, then α, β>1; if the driver is a general driver, then α, β=1; if the driver is an aggressive driver, then 0<α, β<1.

[0072] The calculation formula of the probability weight function is as follows:

[0073]

[0074] Among them, p i is the risk probability of candidate region i, is the vehicle speed index of candidate area i, γ and δ are parameters less than 1, that is, the weight is greater than the probability when the probability is small, and the weight is less than the probability when the probability is medium or large.

[0075] A comprehensive assessment of the legality, efficiency, and safety value functions is set for conservative, normal, and aggressive driving styles. The selectivity of each drivable area is calculated, and the area corresponding to the maximum selectivity is selected as the target area or lane to achieve the lane selection function during driving. Among them, the driving style in the comprehensive decision-making can be set by the driver in the system, and can also be used in combination with the online driving style recognition technology.

[0076] The above is only a preferred specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A lane selection method for an autonomous driving vehicle based on driving style, characterized in that: The following steps are involved: Based on traffic signs and road markings, obtain candidate areas for vehicle travel; Obtaining traffic efficiency based on a vehicle speed index and a front driving space index in the candidate area; Acquire a risk assessment value within the candidate area based on the risk indicator; Making a comprehensive decision on the traffic efficiency and risk assessment value, and constructing a lane selection model; Combining the lane selection model with the driving style to obtain a target driving area; The process of obtaining the vehicle speed index of the candidate area includes obtaining the first sensitivity of the driver to gain or loss based on the driving speed of the target vehicle, the average speed of the vehicles in the candidate area, and the driver's attitude towards risk; Preset a sensitivity threshold, compare the driver's first sensitivity to gain or loss with the sensitivity threshold respectively, and obtain a vehicle speed index of the candidate area; The process of obtaining the front driving space index of the candidate area includes obtaining the second sensitivity of the driver to gain or loss based on the shortest distance between the vehicle and the target vehicle in the candidate area, the maximum range when the lane is selected, and the driver's attitude towards risk; Preset a sensitivity threshold, compare the second sensitivity of the driver to gain or loss with the sensitivity threshold respectively, and obtain the front driving space of the candidate area; The process of making a comprehensive decision on the traffic efficiency and risk assessment value includes obtaining and sorting comprehensive decision values ​​of different candidate areas based on the probability weight of the risk probability in the candidate area and the value function of the vehicle speed index in the candidate area, and selecting the candidate area with the largest comprehensive decision value as the target driving area; The process of obtaining the value function of the vehicle speed index in the candidate area includes obtaining the driver's third sensitivity to gain or loss based on the vehicle speed index in the candidate area and the driver's attitude towards risk; A sensitivity threshold is preset, and the third sensitivity of the driver to gain or loss is compared with the sensitivity threshold respectively to obtain a value function of the vehicle speed index of the candidate area.

2. The method for lane selection of an automatic driving vehicle based on driving style according to claim 1, characterized in that: The risk indicators include the proportion of large vehicles, the standard deviation of vehicle speed, and the standard deviation of the lateral position of the vehicle relative to the lane line.

3. The method for lane selection of an automatic driving vehicle based on driving style according to claim 1, characterized in that: The process of acquiring the traffic efficiency includes acquiring the traffic efficiency in the candidate area based on the vehicle speed index of the candidate area and its weight coefficient, and the front driving space index of the candidate area and its weight coefficient.

4. The method for lane selection of an automatic driving vehicle based on driving style according to claim 1, characterized in that: The process of obtaining the risk assessment value within the candidate area includes obtaining the risk assessment value within the candidate area based on the total number of vehicles in the candidate area, the number of large vehicles, the standard deviation of vehicle speed, the standard deviation of the lateral position of the vehicle relative to the lane line, and the total number of vehicles within the perception range of the target vehicle.

5. The method for lane selection of an automatic driving vehicle based on driving style according to claim 1, characterized in that: The process of obtaining the probability weight of the risk probability in the candidate area includes obtaining the probability weight of the risk probability in the candidate area based on the risk probability of the candidate area, the vehicle speed index of the candidate area, and presetting the first parameter and the second parameter.

Images