Intelligent vehicle lane-changing decision-making planning method, device and system

Abstract

The present invention relates to the field of intelligent vehicle technology, and specifically discloses an intelligent vehicle lane change decision planning method, device and system, including: obtaining data required for lane change decision, the data required for lane change decision at least includes environmental perception data, map positioning data and trajectory prediction data; determining lane change intention according to the data required for lane change decision; when the lane change intention is to change the road, determining the target lane according to the benefit calculation results of the current lane of the intelligent vehicle and the candidate lane; when it is determined that the target lane has a perception blind spot, determining whether the lane change condition is met according to the potential risk level of the perception blind spot; when it is determined that the lane change condition is met, determining the lane change path planning result according to the real-time environmental perception information; outputting the lane change path planning result to the control execution module. The intelligent vehicle lane change decision planning method provided by the present invention can provide lane change decisions in lane change scenarios with large vehicles blocking.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent vehicles, and in particular to an intelligent vehicle lane-changing decision-making planning method, an intelligent vehicle lane-changing decision-making planning device and a lane-changing system. Background Art

[0002] In the lane-changing scenario on highways, intelligent vehicles need to comprehensively consider information such as environmental perception, road maps, and vehicle status to generate safe lane-changing trajectories with accurate decision instructions. Reasonable lane-changing decision-making and planning methods are crucial to improving the driving efficiency of intelligent vehicles and ensuring driving safety.

[0003] Existing intelligent vehicle lane-changing decision-making and planning methods often assume that the vehicle can obtain all the surrounding obstacle information. However, in actual driving scenarios, due to the occlusion of surrounding vehicles or buildings, the assumption that all obstacles are known is not reasonable. When the vehicle's perception field is passive and static obstacles are blocked, potential obstacles in the perception blind spot will affect the safety of the intelligent vehicle; at the same time, the intelligent vehicle generally lacks real-time active adjustment strategies for the perception blind spot during lane changing, which will also reduce the safety of lane-changing decision-making and planning, and easily lead to traffic congestion or accidents.

[0004] For example, existing intelligent vehicle change decisions usually only consider typical lane-changing scenarios, and need to establish a critical safety lane-changing angle model and a safe lane-changing domain. The complete information of the vehicle in front of the target lane must be obtained, which is not suitable for lane-changing decisions and planning of intelligent vehicles when encountering large vehicle occlusion. For lane-changing scenarios on curved roads, lane-changing decisions are usually made only based on the repulsive potential field of the obstacle vehicle in front of the intelligent vehicle. This is too macro and simple, and is based on the assumption that all information about the obstacle in front is known. It is not applicable to lane-changing scenarios with large vehicle occlusion. For lane-changing scenarios with occlusion, only the impact of the lateral sway of the obstructed vehicle on the intelligent vehicle is considered in the path planning process, but the potential risks caused by occlusion are not considered from the lane-changing decision level, and there is a lack of targeted and real-time active adjustment strategies to reduce risks.

[0005] Therefore, how to provide a lane changing decision suitable for the lane changing scenario of being blocked by a large vehicle becomes a technical problem that needs to be solved urgently by those skilled in the art. Summary of the invention

[0006] The present invention provides an intelligent vehicle lane-changing decision-making planning method, an intelligent vehicle lane-changing decision-making planning device and a lane-changing system, which solve the problem in the related art of lack of lane-changing decision-making suitable for lane-changing scenarios blocked by large vehicles.

[0007] As a first aspect of the present invention, a method for intelligent vehicle lane change decision planning is provided, which includes:

[0008] Acquiring data required for lane-changing decision making, wherein the data required for lane-changing decision making includes at least environmental perception data, map positioning data, and trajectory prediction data;

[0009] determining a lane-changing intention according to the lane-changing decision-making required data;

[0010] When the lane change intention is to change the road, the target lane is determined according to the benefit calculation result of the current lane of the intelligent vehicle and the candidate lanes;

[0011] When it is determined that the target lane has a perception blind spot, determining whether a lane change condition is met according to a potential risk level of the perception blind spot;

[0012] When it is determined that the lane changing conditions are met, the lane changing path planning result is determined based on the real-time environment perception information;

[0013] The lane change path planning result is output to the control execution module.

[0014] Further, determining the lane changing intention according to the lane changing decision required data includes:

[0015] determining a lane-changing intention influencing parameter according to the lane-changing decision-making required data;

[0016] Performing fuzzy processing on the lane-changing intention influencing parameter to obtain a fuzzy processing result;

[0017] Reasoning the fuzzification processing result according to the fuzzy rule base and fuzzy reasoning to obtain a fuzzy reasoning result;

[0018] Defuzzifying the fuzzy reasoning result to obtain a lane-changing intention degree value;

[0019] comparing the lane-changing intention degree value with a lane-changing threshold;

[0020] If the lane change intention degree value is greater than the lane change threshold, determining that the lane change intention requires changing roads;

[0021] If the lane-changing intention degree value is not greater than the lane-changing threshold value, it is determined that the lane-changing intention does not require changing roads.

[0022] Furthermore, the target lane is determined according to the benefit calculation results of the lane where the intelligent vehicle is currently located and the candidate lanes, including:

[0023] Calculate the driving benefit of each candidate lane and the driving benefit of the lane where the intelligent vehicle is currently located;

[0024] Compare the driving benefit of the current lane of the intelligent vehicle with the driving benefit of each candidate lane;

[0025] If the driving benefit of the lane where the intelligent vehicle is currently located is less than the driving benefits of each candidate lane, the candidate lane with the maximum driving benefit among all the candidate lanes is taken as the target lane;

[0026] If the driving benefit of the lane where the intelligent vehicle is currently located is not less than the driving benefit of each candidate lane, the lane where the intelligent vehicle is currently located is taken as the target lane.

[0027] Furthermore, the driving benefit of each candidate lane and the driving benefit of the lane where the intelligent vehicle is currently located are calculated respectively, including:

[0028] Calculate the driving efficiency benefit and driving space benefit of each candidate lane respectively, and calculate the driving efficiency benefit and driving space benefit of the lane where the intelligent vehicle is currently located;

[0029] The driving benefit of each candidate lane is calculated according to the driving efficiency benefit and the driving space benefit of each candidate lane, and the driving benefit of the intelligent vehicle is calculated according to the driving efficiency benefit and the driving space benefit of the intelligent vehicle.

[0030] Further, when it is determined that the target lane has a perception blind spot, determining whether a lane change condition is met according to a potential risk level of the perception blind spot includes:

[0031] When it is determined that the target lane has a perception blind spot, determining the potential risk level of the perception blind spot according to speed information of the vehicle closest to the intelligent vehicle in the target lane;

[0032] If the latest position of the vehicle closest to the intelligent vehicle in the target lane is within the perception blind area, determining that the potential risk level of the perception blind area is high risk;

[0033] If the latest position of the vehicle closest to the intelligent vehicle in the target lane is outside the perception blind zone, the potential risk level of the perception blind zone is determined to be low risk;

[0034] When the potential risk level of the perception blind spot is high risk, an active adjustment instruction of the intelligent vehicle is generated to adjust the lateral position or longitudinal position of the intelligent vehicle, and obtain the latest real-time environmental status of the target lane;

[0035] When the potential risk level of the perception blind spot is low risk, obtaining the real-time environmental status of the target lane;

[0036] Determine whether the lane changing conditions are met based on the latest real-time environmental status of the target lane.

[0037] Further, when the potential risk level of the perception blind spot is high risk, an active modulation instruction of the intelligent vehicle is generated to adjust the lateral position or longitudinal position of the intelligent vehicle, including:

[0038] When the potential risk level of the perception blind area is high risk, the lateral action benefit and the longitudinal action benefit are calculated respectively;

[0039] If the lateral motion gain is greater than or equal to the longitudinal motion gain, generating a lateral motion adjustment instruction to adjust the lateral position of the intelligent vehicle;

[0040] If the lateral motion gain is less than the longitudinal motion gain, a longitudinal motion adjustment instruction is generated to adjust the longitudinal position of the intelligent vehicle.

[0041] Furthermore, judging whether the lane change condition is met according to the real-time environment state of the target lane includes:

[0042] Determine the safety distance constraint condition, the expression of the safety distance constraint condition is:

[0043] ,

[0044] in, Indicates the position of the vehicle in front of the target lane, Indicates the position of the vehicle behind the target lane, Indicates the safe distance between the smart vehicle and the vehicle ahead in the target lane without collision. Indicates the safe distance between the smart vehicle and the vehicle behind the target lane to avoid collision;

[0045] Obtain the real-time position of the vehicle in front of the target lane and the real-time position of the vehicle behind the target lane;

[0046] Determine whether the distance between the real-time position of the front vehicle in the target lane and the real-time position of the rear vehicle in the target lane satisfies the safety distance constraint condition;

[0047] If so, it is determined that the intelligent vehicle meets the lane-changing condition;

[0048] If not, it is determined that the intelligent vehicle does not meet the lane changing conditions.

[0049] Furthermore, the lane change path planning result is determined according to the real-time environment perception information, including:

[0050] Determine the obstacle ellipse scope and obstacle hazard factor of the target lane;

[0051] Determine the obstacle repulsion potential field of the target lane according to the obstacle ellipse scope and obstacle hazard coefficient of the target lane;

[0052] Generate the resultant potential field of the target lane according to the obstacle gravitational potential field, blind spot repulsive potential field, obstacle velocity repulsive potential field and obstacle repulsive duration of the target lane;

[0053] Determine the curvature speed constraint condition according to the curvature of the target lane;

[0054] A lane changing path is determined according to the curvature speed constraint condition and the resultant potential field of the target lane.

[0055] As another aspect of the present invention, a smart vehicle lane change decision planning device is provided, which is used to implement the smart vehicle lane change decision planning method described above, and includes:

[0056] An acquisition module, used to acquire data required for lane-changing decision-making, wherein the data required for lane-changing decision-making at least includes environmental perception data, map positioning data, and trajectory prediction data;

[0057] A lane-changing intention determination module, used to determine the lane-changing intention according to the data required for the lane-changing decision;

[0058] A target lane determination module, used for determining a target lane according to a benefit calculation result between a current lane of the intelligent vehicle and a candidate lane when the lane change intention requires changing the road;

[0059] A lane-changing condition determination module, for determining whether a lane-changing condition is met according to a potential risk level of the perception blind spot when it is determined that the target lane has a perception blind spot;

[0060] A path planning module is used to determine a lane change path planning result according to real-time environment perception information when it is determined that the lane change conditions are met;

[0061] The output module is used to output the lane change path planning result to the control execution module.

[0062] As another aspect of the present invention, a lane changing system is provided, which includes: an environment perception module, a trajectory prediction module, a map positioning module, a control execution module and the above-mentioned intelligent vehicle lane changing decision planning device, wherein the environment perception module, the trajectory prediction module, the map positioning module and the control execution module are all in communication connection with the intelligent vehicle lane changing decision planning device;

[0063] The environment perception module is used to perceive the road environment in real time to obtain environment perception data;

[0064] The trajectory prediction module is used to predict the driving trajectory of the intelligent vehicle to obtain trajectory prediction data;

[0065] The map positioning module is used to map the driving route of the intelligent vehicle to obtain map positioning data;

[0066] The intelligent vehicle lane change decision planning device is used to determine the lane change intention according to the data required for the lane change decision, and determine the target lane when the lane change intention is to change the road, and determine whether the lane change condition is met according to the perception blind spot of the target lane, and determine the lane change path planning result according to the real-time environment perception information when it is determined that the lane change condition is met;

[0067] The control execution module is used to control the intelligent vehicle to change lanes according to the lane change path planning result.

[0068] The intelligent vehicle lane change decision planning method provided by the present invention determines the lane change intention through the data required for the lane change decision, and determines the target lane when the lane change intention requires changing roads, and determines whether the lane change conditions are met based on the perception blind spot of the target lane, and determines the lane change path planning results based on the real-time environmental perception information when it is determined that the lane change conditions are met. This intelligent vehicle lane change decision planning method comprehensively considers the occlusion factors in the lane change decision, which is closer to the actual driving conditions. Therefore, the lane change path planning results obtained based on the intelligent vehicle lane change decision planning method can be more suitable for lane changes in the presence of large vehicle occlusion. In summary, the intelligent vehicle lane change decision planning method provided by the present invention is suitable for lane change scenarios under large vehicle occlusion, and can achieve safe lane changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the present invention, but do not constitute a limitation of the present invention.

[0070] Figure 1 For existing intelligent vehicles, there are scenarios where large vehicles are blocking the view.

[0071] Figure 2 This is a flow chart of the intelligent vehicle lane changing decision planning method provided by the present invention.

[0072] Figure 3 A flow chart of a method for determining lane-changing intention provided by the present invention.

[0073] Figure 4 A flow chart of a method for determining a target lane provided by the present invention.

[0074] Figure 5 A flow chart of the method for calculating driving revenue provided by the present invention.

[0075] Figure 6 A flow chart of a method for determining whether a lane-changing condition is met provided by the present invention.

[0076] Figure 7 A flow chart of a method for adjusting the lateral position or longitudinal position of an intelligent vehicle provided by the present invention.

[0077] Figure 8 A flow chart of a method provided by the present invention for determining whether lane-changing conditions are met based on the real-time environmental status of a target lane.

[0078] Fig. 9 A flow chart of a method for determining lane change path planning results provided by the present invention.

[0079] Fig.10 A schematic diagram of the improved artificial potential field method model architecture provided by the present invention.

[0080] Fig.11 This is a structural block diagram of the intelligent vehicle lane change decision-making and planning device provided by the present invention.

[0081] Fig.12 This is a structural block diagram of the lane changing system provided by the present invention.

[0082] Fig.13 A flowchart of lane change decision provided by the present invention. DETAILED DESCRIPTION

[0083] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0084] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of the present invention.

[0085] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so as to describe the embodiments of the present invention described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0086] In lane-changing scenarios, intelligent vehicles need to perceive the surrounding environment. However, due to factors such as occlusion areas and abnormal behaviors, the perception ability of intelligent vehicles is limited, which affects the safety and effectiveness of intelligent vehicle lane-changing decision-making planning. Existing intelligent vehicle lane-changing decision-making planning methods often ignore occlusion factors and lack targeted and real-time adjustment strategies. Specific scenarios such as Figure 1 As shown in the figure, the smart vehicle (black) and other vehicles are traveling on a multi-lane road in the same direction. There is a large vehicle in front of the lane where the smart vehicle is located. Due to the obstruction of the large vehicle, the smart vehicle will have a certain perception blind spot, such as Figure 1 If there is an obstacle in the perception blind spot, it will cause potential safety risks to the vehicle's lane change.

[0087] Based on this, an intelligent vehicle lane change decision planning method is provided in this embodiment. Figure 2 is a flow chart of an intelligent vehicle lane change decision planning method provided according to an embodiment of the present invention. Figure 2 As shown, including:

[0088] S100, acquiring data required for lane-changing decision-making, where the data required for lane-changing decision-making at least includes environment perception data, map positioning data, and trajectory prediction data;

[0089] In an embodiment of the present invention, data required for lane changing decision is acquired according to the current driving environment and vehicle status. Specifically, the data required for lane changing decision includes at least environmental perception data, map positioning data and trajectory prediction data.

[0090] S200, determining a lane-changing intention according to the data required for the lane-changing decision;

[0091] In the embodiment of the present invention, the lane change intention determines whether to execute the lane change, so the lane change intention needs to be determined first according to the data required for the lane change decision. It should be understood that the lane change intention can be determined specifically according to fuzzy logic reasoning. The lane change intention in the embodiment of the present invention can specifically include the need to change the road and the need not to change the road.

[0092] S300, when the lane change intention is to change the road, determine the target lane according to the benefit calculation result of the current lane of the intelligent vehicle and the candidate lanes;

[0093] In an embodiment of the present invention, when the lane change intention requires changing roads, it is necessary to determine the target lane for changing roads. Specifically, the target lane can be determined by comparing the benefit calculation results of the lane where the intelligent vehicle is currently located with the candidate lanes.

[0094] S400: When it is determined that the target lane has a perception blind spot, determining whether a lane change condition is met according to a potential risk level of the perception blind spot;

[0095] After determining the target lane, it is necessary to further determine whether the target lane has a perception blind spot. When it is determined that the target lane has a perception blind spot, determine whether the lane change condition is currently met based on the potential risk level of the perception blind spot.

[0096] Specifically, when determining whether a target lane has a perception blind spot, first determine the driving interest space of the target lane; then determine whether the driving interest space of the target lane has a perception blind spot.

[0097] S500: When it is determined that the lane change condition is met, determining a lane change path planning result according to the real-time environment perception information;

[0098] In an embodiment of the present invention, a lane change path planning result may be determined specifically based on real-time environmental perception information, so that the intelligent vehicle can achieve safe lane change in the presence of large vehicle obstruction.

[0099] S600: Output the lane change path planning result to a control execution module.

[0100] Therefore, the intelligent vehicle lane change decision planning method provided by the present invention determines the lane change intention through the data required for the lane change decision, and determines the target lane when the lane change intention requires changing roads, and determines whether the lane change conditions are met based on the perception blind spot of the target lane, and determines the lane change path planning results based on the real-time environmental perception information when it is determined that the lane change conditions are met. This intelligent vehicle lane change decision planning method comprehensively considers the occlusion factors in the lane change decision, which is closer to the actual driving conditions. Therefore, the lane change path planning results obtained based on the intelligent vehicle lane change decision planning method can be more suitable for lane changes in the presence of large vehicle occlusion. In summary, the intelligent vehicle lane change decision planning method provided by the present invention is suitable for lane change scenarios under large vehicle occlusion, and can achieve safe lane changes.

[0101] In an embodiment of the present invention, the lane changing intention is determined according to the lane changing decision required data, such as Figure 3 As shown, including:

[0102] S210, determining a lane-changing intention influencing parameter according to the lane-changing decision-making required data;

[0103] In an embodiment of the present invention, the lane changing intention influencing parameters can be determined specifically based on the environmental perception data, map positioning data and trajectory prediction data. The lane changing intention influencing parameters can specifically include the speed difference between the current smart vehicle and the preceding vehicle, the relative distance between the current smart vehicle and the preceding vehicle and the degree of threat to the current smart vehicle from other vehicles.

[0104] S220, performing fuzzy processing on the lane-changing intention influencing parameter to obtain a fuzzy processing result;

[0105] Specifically, the speed difference between the current intelligent vehicle and the preceding vehicle, the relative distance between the current intelligent vehicle and the preceding vehicle, and the degree of threat posed to the current intelligent vehicle by other vehicles are fuzzy processed to obtain a fuzzy processing result.

[0106] S230, reasoning the fuzzification processing result according to the fuzzy rule base and fuzzy reasoning to obtain a fuzzy reasoning result;

[0107] In the embodiment of the present invention, a fuzzy rule base may be formulated, and the fuzzification processing result may be inferred based on the fuzzy rule base and fuzzy reasoning to obtain a fuzzy reasoning result.

[0108] S240, performing defuzzification processing on the fuzzy inference result to obtain a lane-changing intention degree value;

[0109] The above fuzzy reasoning result is subjected to defuzzification processing, that is, quantization processing, to obtain a lane-changing intention degree value. It should be understood that the lane-changing intention degree value can represent the magnitude of the lane-changing intention.

[0110] S250, comparing the lane-changing intention degree value with a lane-changing threshold;

[0111] S260: If the lane change intention degree value is greater than the lane change threshold, determining that the lane change intention requires changing roads;

[0112] S270: If the lane-changing intention degree value is not greater than the lane-changing threshold, determine that the lane-changing intention does not require changing roads.

[0113] It should be understood that when the lane change intention degree value is greater than the lane change threshold value, the lane change intention is determined to require a road change, otherwise the lane change intention is determined to not require a road change.

[0114] It should be noted that, in the embodiment of the present invention, the use of fuzzy logic to infer the lane changing intention of an intelligent vehicle may specifically include the following reasons: first, fuzzy logic can effectively simulate the way the human brain processes uncertain reasoning, and is suitable for processing complex and imprecise information; second, when establishing an intelligent vehicle lane changing decision model, it is often difficult to accurately obtain accurate driving environment information, and there is a complex nonlinear relationship between the lane changing decision and the environmental information; third, through fuzzy logic design, multiple state parameters related to lane changing are comprehensively considered to realize multi-input single-output lane changing intention judgment, which can effectively quantify the inference process and improve the accuracy of the judgment result.

[0115] In an embodiment of the present invention, the step of determining whether there is an intention to change lanes based on fuzzy logic reasoning may specifically include: defining input and output and fuzzifying, formulating fuzzy rules, fuzzy reasoning, defuzzifying and outputting results.

[0116] First, define the input and output and fuzzify them. It should be understood that there are many factors that affect the lane change intention. The embodiment of the present invention mainly considers the speed difference with the preceding vehicle. , relative distance to the vehicle in front , the degree of threat from other vehicles (If you encounter a large vehicle, the threat level is high.) As input, the lane change intention As the output, a multi-input-single-output fuzzy reasoning model is established to associate these four state variables, and the vehicle's lane-changing intention is judged based on the output results of the fuzzy reasoning model.

[0117] i. Speed ​​difference The fuzzification is divided into 2 levels, {large, small} is selected as the fuzzy subset, the domain value is [-30, 30], and the triangular membership function is used to fuzzify it.

[0118] ii. Relative distance The fuzzification is divided into two levels, {near, far} is selected as the fuzzy subset, the domain value is [0, 50], and the triangular membership function is used to fuzzify it.

[0119] iii. Level of threat The fuzzification is divided into two levels, {high, low} is selected as the fuzzy subset, the domain value is [0, 1], and the trapezoidal membership function is used to fuzzify it.

[0120] iv. Lane change intention The fuzzification is divided into 2 levels, {strong, weak} is selected as the fuzzy subset, and the membership function is the trapezoidal function.

[0121] Each fuzzy function in the above fuzzy subsets of each state quantity can be determined based on empirical values. At the same time, the fuzzy level and subset can be further subdivided according to actual conditions, such as speed difference The fuzzy level can be divided into 5 levels, and the fuzzy subsets are {small, smaller, medium, larger, large}.

[0122] Secondly, formulate fuzzy rules. In the embodiment of the present invention, fuzzy rules are the core of the fuzzy reasoning system, which is a summary of the vehicle lane-changing rules and is based on descriptive language variables. For example, when the speed difference is large and the relative distance is close, the lane-changing demand is strong; when the speed difference is small and the relative distance is far, the lane-changing demand is weak. The corresponding fuzzy rule statement in this article is as follows:

[0123] if It is "big", When it is "near", If is "high", It is "strong";

[0124] if It is "small", When it is "far", If it is "low", It is "weak".

[0125] Based on this, other corresponding fuzzy rules in vehicle lane changing are obtained. The specific fuzzy rules obtained according to the fuzzy subsets are shown in Table 1. According to the fuzzy levels and subsets of the state variables divided above, 8 rules can be obtained.

[0126] Table 1 Fuzzy rule base

[0127]

[0128] Secondly, fuzzy reasoning is performed. Specifically, fuzzy reasoning is the process of reasoning about known facts and rules to obtain conclusions. Descriptive language variables are used instead of precise values, and the Mamdani reasoning method is used for reasoning. It should be understood that the Mamdani reasoning method uses the classic maximum-minimum synthesis operation method as the operation rule of fuzzy relations and fuzzy sets. In the fuzzy reasoning constructed by the Mamdani reasoning method, both the premise and the consequent are fuzzy language values. In essence, fuzzification and defuzzification modules are added to the input and output parts of the pure fuzzy logic system, respectively, and an unknown conclusion can be derived from one or several known judgments.

[0129] Premise 1: If It is big. It is near. It is near. is high, then is strong;

[0130] Premise 2: It is big. It is near. It is near. is high;

[0131] After: It is strong.

[0132] Finally, the defuzzification operation is performed. It should be understood that fuzzy reasoning is to infer fuzzy rules to obtain inference results. In order to obtain the output controlled by the system, the fuzzy set needs to be defuzzified to obtain an accurate output. The embodiment of the present invention specifically uses the centroid method to defuzzify and obtain an accurate lane change intention degree value. The specific calculation is shown in formula (1):

[0133] , (1)

[0134] in, Indicates the precise lane change intention degree value; Indicates A fuzzy value in the fuzzy set of input quantities; represents the membership function in the corresponding fuzzy set.

[0135] If the lane-changing threshold is set to 0.5, when the precise lane-changing intention degree value obtained by fuzzy reasoning is greater than 0.5, the intelligent vehicle will have a lane-changing intention, otherwise it will not.

[0136] In the embodiment of the present invention, the target lane is determined according to the benefit calculation results of the lane where the intelligent vehicle is currently located and the candidate lanes, such as Figure 4 As shown, including:

[0137] S310, respectively calculating the driving benefit of each candidate lane and calculating the driving benefit of the lane where the intelligent vehicle is currently located;

[0138] In the embodiment of the present invention, after the intention to change lanes is generated, the target lane needs to be selected. Since the main purposes of lane changing are generally two-fold: obtaining more driving space and improving efficiency, the target lane can be selected by calculating the benefits of each candidate lane and then comparing their sizes. The original lane is recorded as the lane , the left lane and the right lane as candidate lanes are recorded as lanes .

[0139] Specifically, the driving benefit of each candidate lane and the driving benefit of the lane where the intelligent vehicle is currently located are calculated respectively, such as Figure 5 As shown, including:

[0140] S311, respectively calculating the driving efficiency benefit and driving space benefit of each candidate lane, and calculating the driving efficiency benefit and driving space benefit of the lane where the intelligent vehicle is currently located;

[0141] First, the benefit of driving efficiency is calculated. Specifically, in order to avoid the situation in which the existing technology of calculating driving efficiency benefit usually only considers the instantaneous speed of vehicles in different lanes and ignores the acceleration, resulting in the failure of lane changing to achieve the expected purpose, in the embodiment of the present invention, the target lane vehicle The instantaneous speed of the vehicle is greater than that of the vehicle in front of it in the current lane. , when after a period of time due to Slow down, and When driving at a constant speed or accelerating, the vehicle can achieve a higher driving speed by keeping the current lane and following the vehicle. Therefore, the embodiment of the present invention takes acceleration into account to calculate the driving efficiency of each candidate lane, so as to be more in line with the actual situation, thereby obtaining a more accurate driving efficiency benefit. The specific calculation formula of the driving efficiency benefit is shown in formula (2):

[0142] , (2)

[0143] in, Indicates lane In the prediction time domain The accumulated value of income within is the sampling period; is the desired speed of the vehicle; and are the initial speed and acceleration of the front vehicle in the lane respectively; , for The moment corresponding to the sampling time.

[0144] Secondly, calculate the driving space benefit. Specifically, the specific calculation formula of the driving space benefit is shown in formula (3):

[0145] , (3)

[0146] in, It indicates the gain of space; and Respectively represent the distance between the vehicle in front and behind on the candidate lane and the vehicle; Indicates the sensing range of the sensor.

[0147] S312. Calculate the driving benefit of each candidate lane according to the driving efficiency benefit and driving space benefit of each candidate lane, and calculate the driving benefit of the intelligent vehicle according to the driving efficiency benefit and driving space benefit of the intelligent vehicle.

[0148] In the embodiment of the present invention, the driving benefit of each candidate lane is calculated according to the driving efficiency benefit and the driving space benefit of each candidate lane. The specific calculation formula is shown in formula (4):

[0149] , (4)

[0150] in, Candidate lane The total lane change benefit; , is the weight of the corresponding sub-item.

[0151] S320, comparing the driving benefit of the lane where the intelligent vehicle is currently located with the driving benefit of each candidate lane;

[0152] S330: If the driving benefit of the lane where the intelligent vehicle is currently located is less than the driving benefits of each candidate lane, the candidate lane with the maximum driving benefit among the driving benefits of all candidate lanes is used as the target lane.

[0153] S340: If the driving benefit of the lane where the intelligent vehicle is currently located is not less than the driving benefit of each candidate lane, the lane where the intelligent vehicle is currently located is used as the target lane.

[0154] The calculated benefits of each lane are compared. If the benefit of the lane where the intelligent vehicle is currently located is less than the benefits of other candidate lanes, the candidate lane with the largest benefit is selected as the target lane for lane change; if the benefits of the lane where the intelligent vehicle is currently located are greater than the benefits of other candidate lanes, the vehicle maintains its current lane and drives. The logical expression is shown in formula (5):

[0155] , (5)

[0156] in, Indicates the target lane number.

[0157] In the embodiment of the present invention, after determining the target lane, it is necessary to further determine whether the target lane has a perception blind spot. Specifically, when determining whether the target lane has a perception blind spot, first determine the driving interest space of the target lane; then determine whether the driving interest space of the target lane has a perception blind spot.

[0158] In the embodiment of the present invention, the driving space of interest during the lane change process is determined within the target lane to facilitate focusing on the road conditions. The driving space of interest is related to the vehicle speed. For example, the higher the vehicle speed, the longer the distance traveled during the lane change process. The calculation formula of the driving space of interest is shown in formula (6):

[0159] , (6)

[0160] in, Indicates the driving interest space; It indicates the adjustment factor, which can be used to deal with speeding situations on some roads; Indicates the maximum speed limit of the target lane. During the lane change process, the intelligent vehicle and other vehicles generally will not exceed this speed; Indicates the maximum lane change time; Indicates the width of a single lane.

[0161] It should be understood that when there is a large vehicle at a certain distance in front of the smart vehicle, it will block the smart vehicle itself, and the curve will also affect the perception range of the vehicle. There is a perception blind spot in front of the target lane, which makes the smart vehicle itself unable to observe the situation of the vehicle in front of it in the driving space of interest in the target lane, causing potential safety risks to the smart vehicle itself, and it is necessary to detect the perception blind spot.

[0162] The perception blind area in the space of interest in front of the target lane is defined as the space that is invisible to the intelligent vehicle itself and the space occupied by obstacles. In the perception blind spot The calculation is shown in formula (7):

[0163] , (7)

[0164] in, Indicates the space occupied by an object; Indicates time Intelligent vehicle’s own vision The visible space in and represents its corresponding complement; Indicates the driving interest space ahead of the target lane; Represents the overall map space.

[0165] The space occupied by the object and the visible space of the vehicle It can be obtained through the analysis and processing of positioning map and environmental perception module data, usually using point cloud clustering or target recognition neural network algorithm to identify , It is calculated by a ray tracing algorithm. It should be noted that the detection of the perception blind spot is well known to those skilled in the art and will not be described in detail here.

[0166] Specifically, when it is determined that the target lane has a perception blind spot, it is determined whether the lane change condition is met according to the potential risk level of the perception blind spot, such as Figure 6 As shown, including:

[0167] S410: When it is determined that the target lane has a perception blind spot, determine the potential risk level of the perception blind spot according to speed information of the vehicle closest to the smart vehicle in the target lane;

[0168] In the embodiment of the present invention, when the intelligent vehicle changes from the current lane to the target lane, it is necessary to consider avoiding collision with the vehicle in front or behind the target lane. Since there is a perception blind spot in front of the target lane, it is impossible to determine whether there is an obstacle vehicle, and potential risks need to be predicted and evaluated.

[0169] S420: If the latest position of the vehicle closest to the intelligent vehicle in the target lane is within the perception blind spot, determine that the potential risk level of the perception blind spot is high risk;

[0170] S430: If the latest position of the vehicle closest to the intelligent vehicle in the target lane is outside the perception blind spot, determine that the potential risk level of the perception blind spot is low risk;

[0171] Since vehicles traveling in the same lane have similar patterns, the intelligent vehicle can observe the information of the nearest vehicle in the target lane and use the constant speed prediction model to predict its position after a certain period of time. The calculation formula is shown in formula (8):

[0172] , (8)

[0173] in, Indicates the initial speed of the nearest vehicle in the target lane; Indicates its initial position; Indicates the maximum lane change time; Indicates the elapsed time The latest position after.

[0174] if In the blind spot If the potential risk level of the perception blind spot is within , it is determined that the potential risk level of the perception blind spot is high risk, and certain measures need to be taken to reduce the risk; if In the blind spot In addition, the potential risk level of the perception blind spot is determined to be low risk, and the smart vehicle itself can calculate whether the lane change safety meets the requirements based on the real-time updated environmental information.

[0175] S440: When the potential risk level of the perception blind spot is high risk, an active adjustment instruction of the intelligent vehicle is generated to adjust the lateral position or longitudinal position of the intelligent vehicle, and obtain the latest real-time environmental status of the target lane;

[0176] In the embodiments of the present invention, it should be understood that when the potential risk level of the perception blind spot is high risk, the risk level can be reduced by actively adjusting the smart vehicle. Specifically, when there is a potential collision risk in the blind spot, the smart vehicle can take active adjustment measures to optimize perception to obtain more environmental information and reduce the impact of occlusion.

[0177] Specifically, when the potential risk level of the perception blind spot is high risk, an active modulation instruction of the intelligent vehicle is generated to adjust the lateral position or longitudinal position of the intelligent vehicle, such as Figure 7 As shown, including:

[0178] S441, when the potential risk level of the perception blind spot is high risk, respectively calculating the lateral action benefit and the longitudinal action benefit;

[0179] Before active adjustment, it is necessary to select horizontal and vertical actions in order to more effectively reduce potential risks. The potential benefits of horizontal or vertical actions are compared to determine which action to choose. The potential benefits mainly consider the increase in visibility and the cost of the corresponding action. The calculation formula of the defined benefit function is shown in formula (9):

[0180] , (9)

[0181] in, It indicates the total benefit of horizontal or vertical action; Indicates an increase in visibility, i.e., a new blind spot With the original blind area The difference between Indicates the cost of horizontal and vertical motion time and energy consumption; and is the weight of the corresponding item.

[0182] S442: if the lateral motion gain is greater than or equal to the longitudinal motion gain, generating a lateral motion adjustment instruction to adjust the lateral position of the intelligent vehicle;

[0183] In the embodiment of the present invention, the benefits of horizontal or vertical actions are compared to select whether to adopt horizontal or vertical actions. If the movement is horizontal, it will move horizontally; otherwise, it will move vertically.

[0184] S443: If the lateral motion gain is less than the longitudinal motion gain, a longitudinal motion adjustment instruction is generated to adjust the longitudinal position of the intelligent vehicle.

[0185] It should be understood that lateral action refers to changing the lateral position of the vehicle, and the specific process is as follows:

[0186] First, assume that the distance from the intelligent vehicle to the edge of the target lane is , discretize the distance into multiple fixed distance intervals. Choose to discretize every 0.2 meters and get the following discretized points:

[0187] Point 1: Distance from lane edge Department;

[0188] Point 2: Distance from lane edge Meter place;

[0189] Point 3: Distance from lane edge Meter place;

[0190] point : Distance from lane edge Meters.

[0191] Then calculate the visibility of each discrete point , and then compare to get the optimal lateral displacement.

[0192] Similarly, the expected speed of the longitudinal action can be discretized and its visibility can be calculated to obtain the optimal longitudinal deceleration. Through active adjustment strategies, the intelligent vehicle can obtain vehicle information inside the original blind spot to better determine whether it is safe to change lanes.

[0193] S450: when the potential risk level of the perception blind spot is low risk, obtaining the real-time environmental status of the target lane;

[0194] S460: Determine whether a lane change condition is met based on the latest real-time environmental status of the target lane.

[0195] In the embodiment of the present invention, the judgment may be made based on the Gipps safety distance model, the principle of which is that when the front vehicle brakes urgently and the rear vehicle needs reaction time, the safety distance between the two vehicles can stop the two vehicles without rear-end collision.

[0196] Specifically, the real-time environment status of the target lane is used to determine whether the lane change condition is met, such as Figure 8 As shown, including:

[0197] S461. Determine a safety distance constraint condition, wherein the expression of the safety distance constraint condition is:

[0198] ,

[0199] in, Indicates the position of the vehicle in front of the target lane, Indicates the position of the vehicle behind the target lane, Indicates the safe distance between the smart vehicle and the vehicle ahead in the target lane without collision. Indicates the safe distance between the smart vehicle and the vehicle behind the target lane to avoid collision;

[0200] S462, obtaining the real-time position of the front vehicle in the target lane and the real-time position of the rear vehicle in the target lane;

[0201] S463, determining whether the distance between the real-time position of the front vehicle in the target lane and the real-time position of the rear vehicle in the target lane satisfies the safety distance constraint condition;

[0202] S464: If the condition is satisfied, it is determined that the intelligent vehicle meets the lane change condition;

[0203] S465: If not, it is determined that the intelligent vehicle does not meet the lane changing condition.

[0204] In the embodiment of the present invention, first, the safe distance between the intelligent vehicle itself and the vehicle in front of the target lane without collision is calculated. , the calculation formula is shown in formula (10):

[0205] , (10)

[0206] in, Indicates the reaction time of the smart vehicle itself; Indicates the longitudinal speed of the smart vehicle itself; Indicates the longitudinal speed of the vehicle ahead in the target lane; and Respectively represent the maximum braking acceleration of the smart vehicle itself and the preceding vehicle.

[0207] Secondly, calculate the safe distance between the vehicle and the vehicle behind the target lane without collision , the calculation formula is shown in formula (11):

[0208] , (11)

[0209] in, Indicates the reaction time of the vehicle behind in the target lane; Indicates the longitudinal speed of the smart vehicle itself; Indicates the longitudinal speed of the following vehicle; and Respectively represent the maximum braking acceleration of the smart vehicle itself and the following vehicle.

[0210] Finally, the safe distance condition required to be met at the lane change time is calculated as shown in formula (12):

[0211] , (12)

[0212] in, and Respectively represent the positions of the front and rear vehicles in the target lane.

[0213] In an embodiment of the present invention, a lane change path planning result is determined based on real-time environment perception information, such as Fig. 9 As shown, including:

[0214] S510, determining the obstacle ellipse scope and obstacle risk factor of the target lane;

[0215] S520, determining the obstacle repulsion potential field of the target lane according to the obstacle ellipse scope and the obstacle hazard coefficient of the target lane;

[0216] S530, generating a resultant potential field of the target lane according to the obstacle gravitational potential field of the target lane, the blind spot repulsive potential field, the obstacle velocity repulsive potential field, and the obstacle repulsive duration;

[0217] S540, determining a curvature speed constraint condition according to the curvature of the target lane;

[0218] S550: Determine a lane change path according to the curvature speed constraint and the resultant potential field of the target lane.

[0219] It should be noted that the commonly used methods for lane change path planning include methods based on polynomial curve interpolation, methods based on sampling optimization, and methods based on learning. Compared with these methods, the artificial potential field method has the advantages of simplicity, stability, and convenience for underlying real-time control. Based on the improved artificial potential field method, a collision-free trajectory that meets the safety and stability requirements is planned during the lane change process of the intelligent vehicle. The improved artificial potential field method model architecture is as follows Fig.10 shown.

[0220] Specifically, the first is the obstacle scope. In view of the problem that in the traditional artificial potential field method, since the virtual repulsive field of the obstacle is usually defined as a circular area, its effective radius is too large or too small, which may pose a risk to the safe driving of other vehicles. The embodiment of the present invention is based on the road condition restrictions and the difference in the longitudinal and lateral speeds of the vehicle. The longitudinal driving distance and speed are larger, while the lateral driving distance and speed are smaller. The principle optimizes the scope of the obstacle repulsive field, that is, the effective distance is increased in the longitudinal direction so that the intelligent vehicle can turn to avoid obstacles; the effective range is reduced in the lateral direction so that the intelligent vehicle can quickly leave the dangerous area. Therefore, the obstacle scope of the embodiment of the present invention is elliptical, which is more suitable for large vehicle obstacles. The length calculation of the major axis of the elliptical scope is shown in formula (13):

[0221] , (13)

[0222] in, Indicates the length of the obstacle vehicle; represents the relative speed between the smart vehicle and the obstacle vehicle, is the speed of the smart vehicle itself; Indicates the braking reaction time; Indicates the mass of the vehicle; Indicates the maximum braking force.

[0223] The length of the minor axis of the ellipse scope is calculated as shown in formula (14):

[0224] , (14)

[0225] in, Indicates a vehicle; Indicates the expansion factor. When approaching a large vehicle, To increase the lateral safety distance.

[0226] Secondly, determine the danger level of the obstacle. In view of the problem that the repulsive potential field function in the prior art only considers the distance between the vehicle and the obstacle, which leads to the safety risk of lane change due to incomplete consideration of factors, the embodiment of the present invention is based on the consideration that the danger level of the obstacle is also closely related to factors such as its speed, mass and category, that is, the faster the speed and the greater the mass, the higher the danger, especially the vehicles carrying dangerous goods are more dangerous. Therefore, the embodiment of the present invention introduces the danger coefficient To consider the danger level of different vehicles, The calculation formula is shown in formula (15):

[0227] , (15)

[0228] in, Indicates the mass of the preceding vehicle; It indicates the adjustment base to prevent the repulsion coefficient from being too large; represents a natural constant; Indicates the type of vehicle in front. The value is determined by the danger level of the vehicle in front and is amplified by an exponential function.

[0229] Based on the introduced elliptical scope and hazard coefficient, the calculation formula of the repulsive potential field function of the embodiment of the present invention is shown in formula (16):

[0230] , (16)

[0231] Next, determine the velocity potential field. The intelligent vehicle of the embodiment of the present invention also needs to consider dynamic obstacles, especially fast-moving vehicles, when avoiding static obstacles. At this time, the speed of the dynamic obstacle is the key factor. For this purpose, a velocity potential field is established, and its calculation formula is shown in formula (17):

[0232] , (17)

[0233] in, represents the gain factor of the velocity potential field; represents the relative speed between the smart vehicle and the obstacle, represents the speed of the vehicle, Indicates the speed of the obstacle car; Indicates speed Direction and vector The angle between them, where the vector It is the vector composed of the smart vehicle and the current obstacle point.

[0234] The repulsive force calculation formula of the velocity potential field is shown in formula (18):

[0235] , (18)

[0236] Finally, determine the blind spot potential field function. If the intelligent vehicle has a blind spot due to the obstruction of a large vehicle or the shape of the road during the lane change process, there may be other obstacles in the blind spot. Therefore, setting the sensor blind spot repulsive potential field can improve the safety of the host vehicle's lane change path planning. Specifically, the calculation formula for the blind spot potential field function is shown in formula (19):

[0237] , (19)

[0238] in, represents the blind zone repulsion coefficient; Indicates the relative distance between the blind spot boundary and the smart vehicle itself, which is obtained by analyzing perception and high-definition map data; Indicates the speed of the smart vehicle itself; Indicates whether a blind spot exists or not.

[0239] The repulsive force calculation of the blind zone potential field is shown in formula (20):

[0240] , (20)

[0241] In summary, the calculation formula of the total potential field function is shown in formula (21):

[0242] ,(twenty one)

[0243] The calculation formula of the combined force acting on the vehicle is shown in formula (22):

[0244] ,(twenty two)

[0245] In addition, the safety speed constraint of the road curvature needs to be considered. Because in the process of changing lanes, due to the curve with a certain curvature, the speed of each lane change point needs to be constrained to ensure safety. The calculation formula is shown in formula (23):

[0246] ,(twenty three)

[0247] in, Indicates the maximum speed limit of the target lane; represents the friction coefficient; represents the acceleration due to gravity; represents the road curvature, which can be obtained from formula (24):

[0248] ,(twenty four)

[0249] Wherein, y represents the curve equation of the lane centerline, which can be obtained through the perceived lane line detection, for example, by using some polynomial curve fitting methods, such as using trinomial curve fitting, then y=a+bx+cx 2 +dx 3 , where a, b, c and d are coefficients to be determined, which can be obtained by solving the discrete points of the lane centerline. It should be noted that there are many ways to obtain the lane centerline curve equation, and only one example is given here, which is well known to those skilled in the art and will not be described here.

[0250] The corresponding algorithm program is constructed through the above expressions, and the lane change path planning result can be obtained by applying the improved artificial potential field method. The lane change path planning result is input into the control execution module to complete the lane change operation.

[0251] In summary, the intelligent vehicle lane changing decision planning method provided by the present invention comprehensively considers the occlusion factors in the lane changing decision, which is closer to the actual driving condition; by adding the active adjustment strategy for the perception blind spot, it helps to reduce the perception blind spot and reduce the potential risk; by introducing the obstacle elliptical scope, hazard coefficient, obstacle speed and blind spot repulsive potential field, and considering the curvature to constrain the speed, the improved artificial potential field method is more suitable for the lane changing scenario with large vehicle occlusion, thereby improving the safety of path planning.

[0252] As another embodiment of the present invention, a smart vehicle lane change decision planning device is provided, which is used to implement the smart vehicle lane change decision planning method described above, wherein: Fig.11 As shown, the intelligent vehicle lane change decision planning device 100 includes:

[0253] An acquisition module 110 is used to acquire data required for lane change decision making, wherein the data required for lane change decision making includes at least environmental perception data, map positioning data, and trajectory prediction data;

[0254] A lane-changing intention determination module 120, configured to determine a lane-changing intention based on the data required for the lane-changing decision;

[0255] A target lane determination module 130, configured to determine a target lane according to a benefit calculation result between a current lane of the intelligent vehicle and a candidate lane when the lane change intention requires changing the road;

[0256] A lane-changing condition determination module 140 is used to determine whether a lane-changing condition is met according to a potential risk level of the perception blind spot when it is determined that the target lane has a perception blind spot;

[0257] A path planning module 150 is used to determine a lane change path planning result according to the real-time environment perception information when it is determined that the lane change condition is met;

[0258] The output module 160 is used to output the lane change path planning result to the control execution module.

[0259] The intelligent vehicle lane change decision planning device provided by the present invention determines the lane change intention through the data required for the lane change decision, and determines the target lane when the lane change intention requires changing roads, and determines whether the lane change conditions are met based on the perception blind spot of the target lane, and determines the lane change path planning results based on the real-time environmental perception information when it is determined that the lane change conditions are met. This intelligent vehicle lane change decision planning method comprehensively considers the occlusion factors in the lane change decision, which is closer to the actual driving conditions. Therefore, the lane change path planning results obtained based on the intelligent vehicle lane change decision planning device can be more suitable for lane changes in the presence of large vehicle occlusion. In summary, the intelligent vehicle lane change decision planning device provided by the present invention is suitable for lane change scenarios under large vehicle occlusion, and can achieve safe lane changes.

[0260] The specific working principle of the intelligent vehicle lane changing decision planning device of the present invention can be referred to the description of the intelligent vehicle lane changing decision planning method above, which will not be repeated here.

[0261] As another embodiment of the present invention, a lane changing system 10 is provided, wherein: Fig.12 As shown, it includes: an environment perception module 200, a trajectory prediction module 300, a map positioning module 400, a control execution module 500 and the intelligent vehicle lane change decision planning device 100 described in Article 9 above, and the environment perception module 200, the trajectory prediction module 300, the map positioning module 400 and the control execution module 500 are all connected to the intelligent vehicle lane change decision planning device 100 in communication;

[0262] The environment perception module 200 is used to perceive the road environment in real time to obtain environment perception data;

[0263] The trajectory prediction module 300 is used to predict the driving trajectory of the intelligent vehicle to obtain trajectory prediction data;

[0264] The map positioning module 400 is used to map the driving route of the intelligent vehicle to obtain map positioning data;

[0265] The intelligent vehicle lane change decision planning device 100 is used to determine the lane change intention according to the lane change decision required data, and determine the target lane when the lane change intention is to change the road and determine whether the lane change condition is met according to the perception blind spot of the target lane, and determine the lane change path planning result according to the real-time environment perception information when it is determined that the lane change condition is met;

[0266] The control execution module 500 is used to control the intelligent vehicle to change lanes according to the lane change path planning result.

[0267] The lane changing system in the embodiment of the present invention comprises an environment perception module 200, a trajectory prediction module 300, a map positioning module 400, an intelligent vehicle lane changing decision planning device 100 and a control execution module 500. The environment perception module 200 is responsible for collecting real-time perception data of the entire driving environment, such as the position, speed and acceleration of surrounding vehicles; the trajectory prediction module 300 can realize the trajectory prediction of the intelligent vehicle; the positioning map module 400 provides a high-precision map including information such as road geometry, lane markings and traffic signs, and determines the precise position and orientation of vehicles, obstacles, etc. on the map; before changing lanes, the intelligent vehicle lane changing decision planning device 100 can make reasonable decisions and obtain lane changing instructions based on the current driving environment and vehicle status by analyzing real-time perception data, map information and positioning results; and can generate an optimal lane changing path using a planning algorithm based on the instruction output of the lane changing decision module and considering factors such as the current position of the vehicle and the driving status of surrounding vehicles. Finally, the control execution module 500 controls the vehicle to execute the lane changing action by adjusting the steering angle, acceleration and braking force according to the lane changing path provided by the planning module to ensure that the vehicle completes the lane changing operation safely and smoothly.

[0268] The intelligent vehicle lane change decision planning device 100 of the embodiment of the present invention needs to make reasonable decisions based on the current driving environment and vehicle status by analyzing real-time perception, prediction and positioning map data information to obtain accurate lane change instructions. The process of lane change decision is as follows: Fig.13 As shown, the process includes the generation of lane change intention, selection of target lane, perception of blind spot detection, active adjustment and lane change safety verification.

[0269] In summary, the lane changing system provided by the present invention takes into account the lane changing decision of intelligent vehicles in the presence of large vehicle obstructions, reduces the perception blind spots and potential risks by considering the obstruction factors in the lane changing decision and adding active adjustment strategies, and uses an improved artificial potential field method to plan the lane changing path. This makes it more suitable for lane changing scenarios where intelligent vehicles encounter obstructions from large vehicles, thereby improving the safety of path planning.

[0270] The specific working process of the lane changing system provided by the present invention can refer to the description of the intelligent vehicle lane changing decision planning method above, which will not be repeated here.

[0271] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present invention, but the present invention is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A lane-changing decision-making and planning method for an intelligent vehicle, characterized in that: include: Acquiring data required for lane-changing decision making, wherein the data required for lane-changing decision making includes at least environmental perception data, map positioning data, and trajectory prediction data; determining a lane-changing intention according to the lane-changing decision-making required data; When the lane change intention is to change the road, the target lane is determined according to the benefit calculation result of the current lane of the intelligent vehicle and the candidate lanes; When it is determined that the target lane has a perception blind spot, determining whether a lane change condition is met according to a potential risk level of the perception blind spot; When it is determined that the lane changing conditions are met, the lane changing path planning result is determined based on the real-time environment perception information; Outputting the lane change path planning result to a control execution module; The target lane is determined based on the benefit calculation results of the current lane of the intelligent vehicle and the candidate lanes, including: Calculate the driving benefit of each candidate lane and the driving benefit of the lane where the intelligent vehicle is currently located; Compare the driving benefit of the current lane of the intelligent vehicle with the driving benefit of each candidate lane; If the driving benefit of the lane where the intelligent vehicle is currently located is less than the driving benefits of each candidate lane, the candidate lane with the maximum driving benefit among all the candidate lanes is taken as the target lane; If the driving benefit of the lane where the intelligent vehicle is currently located is not less than the driving benefit of each candidate lane, the lane where the intelligent vehicle is currently located is taken as the target lane; Calculate the driving benefit of each candidate lane and the driving benefit of the lane where the intelligent vehicle is currently located, including: The driving efficiency benefit and driving space benefit of each candidate lane are calculated respectively, and the driving efficiency benefit and driving space benefit of the lane where the intelligent vehicle is currently located are calculated. The calculation formula of the driving efficiency benefit is: ， in, Indicates lane In the prediction time domain The accumulated value of income within is the sampling period; is the desired speed of the vehicle; and are the initial speed and acceleration of the front vehicle in the lane respectively; , for The moment corresponding to the sampling time; The calculation formula of the driving space benefit is: ， in, It indicates the gain of space; and Respectively represent the distance between the vehicle in front and behind on the candidate lane and the vehicle; Indicates the sensing range of the sensor; The driving benefit of each candidate lane is calculated according to the driving efficiency benefit and driving space benefit of each candidate lane, and the driving benefit of the intelligent vehicle is calculated according to the driving efficiency benefit and driving space benefit of the intelligent vehicle. The calculation formula of the driving benefit of each candidate lane is: ， in, Candidate lane The total lane change benefit; , is the weight of the corresponding sub-item; When it is determined that the target lane has a perception blind spot, determining whether a lane change condition is met according to a potential risk level of the perception blind spot includes: When it is determined that the target lane has a perception blind spot, determining the potential risk level of the perception blind spot according to speed information of the vehicle closest to the intelligent vehicle in the target lane; If the latest position of the vehicle closest to the intelligent vehicle in the target lane is within the perception blind area, determining that the potential risk level of the perception blind area is high risk; If the latest position of the vehicle closest to the intelligent vehicle in the target lane is outside the perception blind zone, the potential risk level of the perception blind zone is determined to be low risk; When the potential risk level of the perception blind spot is high risk, an active adjustment instruction of the intelligent vehicle is generated to adjust the lateral position or longitudinal position of the intelligent vehicle, and obtain the latest real-time environmental status of the target lane; When the potential risk level of the perception blind spot is low risk, obtaining the real-time environmental status of the target lane; Determine whether lane change conditions are met based on the latest real-time environmental status of the target lane; When the potential risk level of the perception blind spot is high risk, an active modulation instruction of the intelligent vehicle is generated to adjust the lateral position or longitudinal position of the intelligent vehicle, including: When the potential risk level of the perception blind spot is high risk, the lateral action benefit and the longitudinal action benefit are calculated respectively, wherein the action to be selected is determined by comparing the potential benefits of the lateral or longitudinal actions. The calculation formula of the benefit function is: ， in, It indicates the total benefit of horizontal or vertical action; Indicates an increase in visibility, i.e., a new blind spot With the original blind area The difference between Indicates the cost of horizontal and vertical motion time and energy consumption; and is the weight of the corresponding item; If the lateral motion gain is greater than or equal to the longitudinal motion gain, generating a lateral motion adjustment instruction to adjust the lateral position of the intelligent vehicle; If the lateral motion gain is less than the longitudinal motion gain, a longitudinal motion adjustment instruction is generated to adjust the longitudinal position of the intelligent vehicle.

2. The intelligent vehicle lane change decision planning method according to claim 1, characterized in that: Determining a lane-changing intention according to the lane-changing decision-making required data includes: determining a lane-changing intention influencing parameter according to the lane-changing decision-making required data; Performing fuzzy processing on the lane-changing intention influencing parameter to obtain a fuzzy processing result; Reasoning the fuzzification processing result according to the fuzzy rule base and fuzzy reasoning to obtain a fuzzy reasoning result; Defuzzifying the fuzzy reasoning result to obtain a lane-changing intention degree value; comparing the lane-changing intention degree value with a lane-changing threshold; If the lane change intention degree value is greater than the lane change threshold, determining that the lane change intention requires changing roads; If the lane-changing intention degree value is not greater than the lane-changing threshold value, it is determined that the lane-changing intention does not require changing roads.

3. The intelligent vehicle lane change decision planning method according to claim 1, characterized in that: Determine whether the lane change conditions are met based on the real-time environmental status of the target lane, including: Determine the safety distance constraint condition, the expression of the safety distance constraint condition is: ， in, Indicates the position of the vehicle in front of the target lane, Indicates the position of the vehicle behind the target lane, Indicates the safe distance between the smart vehicle and the vehicle ahead in the target lane without collision. Indicates the safe distance between the smart vehicle and the vehicle behind the target lane to avoid collision; Obtain the real-time position of the vehicle in front of the target lane and the real-time position of the vehicle behind the target lane; Determine whether the distance between the real-time position of the front vehicle in the target lane and the real-time position of the rear vehicle in the target lane satisfies the safety distance constraint condition; If so, it is determined that the intelligent vehicle meets the lane-changing condition; If not, it is determined that the intelligent vehicle does not meet the lane changing conditions.

4. The intelligent vehicle lane change decision planning method according to claim 1, characterized in that: The lane change path planning results are determined based on real-time environmental perception information, including: Determine the obstacle ellipse scope and obstacle hazard factor of the target lane; Determine the obstacle repulsion potential field of the target lane according to the obstacle ellipse scope and obstacle hazard coefficient of the target lane; Generate the resultant potential field of the target lane according to the obstacle gravitational potential field, blind spot repulsive potential field, obstacle velocity repulsive potential field and obstacle repulsive duration of the target lane; Determine the curvature speed constraint condition according to the curvature of the target lane; A lane changing path is determined according to the curvature speed constraint condition and the resultant potential field of the target lane.

5. An intelligent vehicle lane-changing decision-making and planning device, used to implement the intelligent vehicle lane-changing decision-making and planning method according to any one of claims 1 to 4, characterized in that: include: An acquisition module, used to acquire data required for lane-changing decision-making, wherein the data required for lane-changing decision-making at least includes environmental perception data, map positioning data, and trajectory prediction data; A lane-changing intention determination module, used to determine the lane-changing intention according to the data required for the lane-changing decision; A target lane determination module, used for determining a target lane according to a benefit calculation result between a current lane of the intelligent vehicle and a candidate lane when the lane change intention requires changing the road; A lane-changing condition determination module, for determining whether a lane-changing condition is met according to a potential risk level of the perception blind spot when it is determined that the target lane has a perception blind spot; A path planning module is used to determine a lane change path planning result according to real-time environment perception information when it is determined that the lane change conditions are met; The output module is used to output the lane change path planning result to the control execution module.

6. A lane changing system, characterized in that: include: An environment perception module, a trajectory prediction module, a map positioning module, a control execution module and the intelligent vehicle lane change decision planning device according to claim 5, wherein the environment perception module, the trajectory prediction module, the map positioning module and the control execution module are all in communication connection with the intelligent vehicle lane change decision planning device; The environment perception module is used to perceive the road environment in real time to obtain environment perception data; The trajectory prediction module is used to predict the driving trajectory of the intelligent vehicle to obtain trajectory prediction data; The map positioning module is used to map the driving route of the intelligent vehicle to obtain map positioning data; The intelligent vehicle lane change decision planning device is used to determine the lane change intention according to the data required for the lane change decision, and determine the target lane when the lane change intention is to change the road, and determine whether the lane change condition is met according to the perception blind spot of the target lane, and determine the lane change path planning result according to the real-time environment perception information when it is determined that the lane change condition is met; The control execution module is used to control the intelligent vehicle to change lanes according to the lane change path planning result.

Images