Cooperative obstacle avoidance system for autonomous vehicle and control method thereof

Abstract

The invention relates to the technical field of cooperative obstacle avoidance systems, in particular to a cooperative obstacle avoidance system for an autonomous vehicle and a control method thereof, comprising an obstacle sensing unit for identifying a path of an obstacle in a road; and an obstacle avoidance unit planning an obstacle avoidance path of the vehicle through sensing information of the obstacle sensing unit, the obstacle sensing unit including: a first sensing data input unit for receiving, from at least two distance sensors, whether there is an obstacle and relative distance data of the obstacle; a first calculation unit for obtaining position information of the obstacle based on the relative distance data received from the first sensing data input unit; the system comprises an obstacle sensing unit and an obstacle avoidance unit, the environment can be comprehensively sensed, complex information can be rapidly processed, an optimal decision can be made, and real-time information exchange and decision coordination between vehicles are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of collaborative obstacle avoidance systems, and in particular to a collaborative obstacle avoidance system for an autonomous driving vehicle and a control method thereof. Background Art

[0002] With the rapid development of the automotive industry and artificial intelligence technology, autonomous driving has gradually moved from the imagination of science fiction movies to real life. However, in this exciting technological innovation, safety is always the primary consideration. Especially in complex and changing road environments, how to effectively avoid obstacles has become a major challenge facing autonomous driving technology.

[0003] At present, most autonomous driving systems rely mainly on single-vehicle perception and decision-making to achieve obstacle avoidance. This method performs well in simple scenarios, but often fails in complex traffic environments. For example, on urban roads with dense traffic, it is difficult to detect potential dangers in time with the limited perception range of a single vehicle; on highways, single-vehicle decisions may lead to sudden lane changes, triggering a chain reaction, causing traffic congestion or even accidents. In addition, existing obstacle avoidance systems often use fixed decision-making models, lack adaptability to different driving environments, and are difficult to perform well in various complex scenarios.

[0004] Another prominent problem is that most existing obstacle avoidance systems focus on the safety of individual vehicles, ignoring the coordination and optimization of the overall traffic flow. This independent approach not only reduces road traffic efficiency, but also increases the risk of accidents due to conflicts in individual decisions. Especially in some key scenarios, such as complex intersections or highway confluences, the lack of coordinated obstacle avoidance strategies often leads to traffic paralysis.

[0005] In addition, the existing technology's response speed and accuracy in dealing with emergencies need to be improved. In some dangerous situations that require millisecond-level responses, such as sudden braking of the vehicle in front or pedestrians entering the lane, the performance of existing systems is often unsatisfactory. This not only affects driving safety, but also limits the widespread application of autonomous driving technology.

[0006] In the face of these challenges, the industry urgently needs a more intelligent, collaborative, and efficient obstacle avoidance system. An ideal system should be able to comprehensively consider multiple factors, such as vehicle performance, road conditions, and traffic flow, and make comprehensive and intelligent decisions. At the same time, the system should also have strong collaborative capabilities, be able to achieve the overall optimal obstacle avoidance strategy in multi-vehicle situations, improve road utilization, and reduce accident risks. Summary of the invention

[0007] It is against this background that the present invention proposes an innovative autonomous driving vehicle collaborative obstacle avoidance system. The system aims to solve many problems existing in the prior art, including limited perception range, rigid decision-making model, lack of collaborative ability, and insufficient response speed. By introducing advanced multi-sensor fusion technology, adaptive decision-making algorithm and multi-vehicle collaborative mechanism, the present invention not only greatly improves the obstacle avoidance capability of a single vehicle, but also realizes intelligent collaboration between multiple vehicles, thereby fundamentally improving the safety and efficiency of autonomous driving.

[0008] The present invention proposes a collaborative obstacle avoidance system for an autonomous driving vehicle, comprising: an obstacle sensing unit for identifying a path blocked by an obstacle in a road; and an obstacle avoidance unit for planning an obstacle avoidance path of the vehicle through sensing information of the obstacle sensing unit, wherein the obstacle sensing unit comprises: a first sensing data input unit for receiving data on whether an obstacle exists and a relative distance of the obstacle from at least two distance sensors; a first calculation unit for obtaining position information of the obstacle based on the relative distance data received from the first sensing data input unit; and a first conversion unit for converting the position information of the obstacle received from the first calculation unit into a visual image, wherein each component of the obstacle sensing unit is respectively connected to a corresponding component of the obstacle avoidance unit through a wiring harness.

[0009] Preferably, the obstacle avoidance unit includes: a driving data input unit for receiving first sensing data from the first sensing data input unit and receiving position information, speed information and maneuvering data of at least one vehicle from a sending / receiving unit; a second calculation unit for calculating a spatial collision probability between vehicles; a third calculation unit for calculating an obstacle characteristic value representing a characteristic value of each of a plurality of obstacles; a fourth calculation unit for calculating a relative position representing the positions of a plurality of vehicles and each of a plurality of obstacles; a fifth calculation unit for calculating a relative speed of the vehicle using the speed information of the vehicle; and a sixth calculation unit for calculating an obstacle avoidance value representing the necessary degree to perform an obstacle avoidance operation on the vehicle based on the spatial collision probability received from the second calculation unit, the obstacle characteristic value received from the third calculation unit, the relative position received from the fourth calculation unit and the relative speed received from the fifth calculation unit.

[0010] Preferably, the obstacle avoidance unit further comprises: a first synthesis unit for planning an obstacle avoidance path for the vehicle based on the obstacle avoidance value and selecting an obstacle avoidance route associated with the obstacle avoidance path; and a second synthesis unit for calculating an allocation value for obstacle avoidance route allocation from the obstacle avoidance route received by the first synthesis unit, wherein the first synthesis unit and the second synthesis unit are connected to the sending / receiving unit via a wiring harness, the first synthesis unit comprising: a route planning model for planning an allocation path for the obstacle avoidance route of the vehicle; and a function value for modifying the obstacle avoidance rate in the route planning model by using the relevant distance between vehicles; the second synthesis unit The second synthesis unit includes: an obstacle avoidance rate function for calculating the obstacle avoidance rate between vehicles using the obstacle avoidance value; and a function distance value for comparing the distance between the weight in the obstacle avoidance rate function and the assigned value; the first synthesis unit also includes: a priority function for assigning priorities to multiple obstacle avoidance routes formed by avoiding obstacles and maintaining the current route according to the obstacle avoidance value received from the sixth calculation unit, and an obstacle avoidance planning mode received from the first planning unit; and a mode allocation unit for outputting the allocation mode with the highest allocation value in the obstacle avoidance route from the assigned value received from the priority function and the obstacle avoidance planning mode received from the first planning unit.

[0011] Preferably, the obstacle avoidance unit also includes: a seventh calculation unit for calculating an obstacle avoidance rate function based on the obstacle avoidance value received from the sixth calculation unit; an eighth calculation unit for performing collision probability allocation as the vehicle's obstacle avoidance rate based on the obstacle avoidance value received from the sixth calculation unit and the obstacle avoidance rate function received from the seventh calculation unit; a ninth calculation unit for receiving the obstacle avoidance rate function from the seventh calculation unit and the allocated obstacle avoidance rate received from the eighth calculation unit and calculating an allocation function based on the obstacle avoidance rate function; a first planning unit for calculating an obstacle avoidance route by receiving the allocation function of the obstacle avoidance rate function from the ninth calculation unit and the spatial collision probability received from the second calculation unit; and a tenth calculation unit for calculating the cost value of the obstacle avoidance route by receiving the obstacle characteristic value of the obstacle from the third calculation unit and the obstacle avoidance route received from the first planning unit, wherein the sixth to tenth calculation units are implemented by the performance of a single processor.

[0012] Preferably, the sixth calculation unit includes: a first calculation model for calculating the obstacle avoidance value, which uses the data received by the sixth calculation unit to calculate the obstacle avoidance value of the vehicle; and a second calculation model for modifying the parameters in the first calculation model using the data received by the sixth calculation unit, the parameters being used to change the proportion of each component in the obstacle avoidance value; modifying the parameters in the second calculation model using the data received by the sixth calculation unit includes: calculating the obstacle avoidance value by using the relative distance between vehicles, converting the obstacle avoidance rate function into a weight function, and modifying the parameters in the weight function based on the obstacle avoidance value.

[0013] Preferably, it also includes: a second planning unit, which is used to plan a non-collision path for each vehicle based on the obstacle avoidance route received from the first planning unit and the cost value received from the tenth calculation unit, wherein the second planning unit also receives the obstacle avoidance route from one of the first synthesis unit and the second synthesis unit.

[0014] Preferably, the eleventh calculation unit includes: a path generation unit for generating a non-collision path using the cost value received by the second planning unit; a detection unit for detecting whether the cost value received from the tenth calculation unit has a value equal to or greater than a threshold; a selection unit for receiving the cost value from the tenth calculation unit and the cost value received from the detection unit, a first obstacle avoidance route having a maximum cost value from the selected values ​​of the cost values ​​received from the tenth calculation and the detection unit and a second obstacle avoidance route having a minimum cost value from the selected values ​​of the cost values ​​received from the tenth calculation and the detection unit; and a first obstacle avoidance route or a second obstacle avoidance path, selecting the first obstacle avoidance route or the second obstacle avoidance route when the detection unit detects the generation of a replacement for the first obstacle avoidance route or the second obstacle avoidance route; the eleventh calculation unit generates a non-collision route based on the non-collision path of the cost value in the second obstacle avoidance path, and sets the non-collision route in the first obstacle avoidance route as the route to be driven.

[0015] Preferably, it also includes: a vehicle configuration unit for receiving at least information about obstacle avoidance route distribution and non-collision route information about each non-collision route from the first synthesis unit, wherein the vehicle configuration unit is configured to perform vehicle configuration to perform obstacle avoidance routes in a vehicle queue, wherein each vehicle in the queue is configured according to the configuration of the vehicle to reduce mutual influence between multiple vehicles; the vehicle configuration unit is also used to receive obstacle avoidance route allocation information from one of the first synthesis unit and the second synthesis unit to use the vehicle configuration unit to set the configuration of the vehicle to reduce mutual influence between multiple vehicles.

[0016] The control method of the cooperative obstacle avoidance system of the autonomous driving vehicle comprises: identifying a route of a road blocked by obstacles; planning an obstacle avoidance path for the vehicle; and setting a non-collision route for the vehicle, wherein the setting route is accomplished by reducing the mutual influence between multiple vehicles when planning the route for obstacle avoidance, wherein the configuration of the obstacle avoidance path planned for the vehicle is selected from the following behaviors: (1) a route planning model for planning roads into obstacle avoidance routes, and (2) an obstacle avoidance route allocation function for allocating obstacle avoidance routes for each vehicle using an obstacle avoidance rate function, and setting the non-collision route for the vehicle comprises: (1) a path generation model for generating paths using cost values, and (2) selecting a minimum value from multiple cost value allocation values ​​received by the second planning unit.

[0017] Preferably, a route planning model for allocating collision-avoiding routes for a plurality of vehicles is implemented using an average allocation value weighted according to the individual performance of each vehicle, the allocated routes are generated between an obstacle avoidance route having a minimum allocation value and a non-collision route having a maximum allocation value, and when avoiding an obstacle, an obstacle avoidance route that bypasses the obstacle is generated and driven, and a route planning model for allocating non-collision paths for a plurality of vehicles is implemented based on the weighted average allocation value rather than vehicle performance, and when a vehicle avoids an obstacle, an obstacle avoidance route that bypasses the obstacle is not generated and driven, and the road planning model selects an obstacle avoidance route for an obstacle, and the obstacle avoidance rate function has a plurality of multiplication factors, and according to the individual performance of each vehicle, the multiplication factors are made to reach a weighted value based on the average allocation value, and the allocated route selection function is based on the number of routes.

[0018] The beneficial technical effects brought by the technical solution of the present invention are:

[0019] The core of the present invention lies in its unique modular design and intelligent algorithm. The system includes an obstacle sensing unit and an obstacle avoidance unit, which can fully perceive the environment, quickly process complex information, and make the best decision. Especially in terms of multi-vehicle collaboration, the present invention realizes real-time information exchange and decision coordination between vehicles through innovative algorithms, greatly improving the overall obstacle avoidance efficiency.

[0020] In conclusion, the autonomous driving vehicle cooperative obstacle avoidance system of the present invention not only overcomes many limitations of the existing technology, but also achieves significant improvements in safety, efficiency, adaptability, etc. Its application will pave the way for the further development and popularization of autonomous driving technology, and contribute to the construction of a safer and smarter future transportation system. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is the overall logic block diagram of the system of the present invention.

[0022] Figure 2It is a logic block diagram of the obstacle sensing unit of the present invention.

[0023] Figure 3 It is a logic block diagram of the obstacle avoidance unit of the present invention.

[0024] Figure 4 It is a logic block diagram of the sixth computing unit of the present invention.

[0025] Figure 5 It is a logic block diagram of the first synthesis unit of the present invention.

[0026] Figure 6 It is a logic block diagram of the vehicle configuration unit of the present invention. DETAILED DESCRIPTION

[0027] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects thereof are described in detail below in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0028] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0029] See also Figure 1-6 , the present invention provides a collaborative obstacle avoidance system for an autonomous driving vehicle and a control method thereof.

[0030] First, the core of the present invention is a cooperative obstacle avoidance system for an autonomous driving vehicle, which includes: an obstacle sensing unit 1 for identifying a path blocked by an obstacle in a road; and an obstacle avoidance unit 2 for planning an obstacle avoidance path of the vehicle through sensing information of the obstacle sensing unit 1. The obstacle sensing unit 1 further includes: a first sensing data input unit 11 for receiving data on whether an obstacle exists and a relative distance of the obstacle from at least two distance sensors; a first calculation unit 12 for obtaining position information of the obstacle based on the relative distance data received from the first sensing data input unit 11; and a first conversion unit 13 for converting the position information of the obstacle received from the first calculation unit 12 into a visual image. It is worth noting that each component of the obstacle sensing unit 1 is respectively connected to a corresponding component of the obstacle avoidance unit 2 through a wiring harness.

[0031] In one embodiment of the present invention, the obstacle sensing unit 1 may use various types of distance sensors such as laser radar, millimeter wave radar or ultrasonic sensor. For example, in an urban road scenario, the system may use laser radar and millimeter wave radar at the same time. Laser radar can provide high-precision distance and shape information, and is particularly suitable for identifying static obstacles such as parked vehicles or traffic signs on the roadside; while millimeter wave radar performs better in adverse weather conditions and can effectively detect moving objects such as pedestrians or other vehicles. This multi-sensor fusion method significantly improves the reliability and adaptability of the system.

[0032] Preferably, in one embodiment of the present invention, the obstacle avoidance unit 2 includes: a driving data input unit 21 for receiving first sensing data from the first sensing data input unit 11, and receiving position information, speed information and manipulation data of at least one vehicle from a sending / receiving unit; a second calculation unit 22 for calculating the spatial collision probability between vehicles; a third calculation unit 23 for calculating an obstacle characteristic value representing a characteristic value of each of a plurality of obstacles; a fourth calculation unit 24 for calculating a relative position representing the positions of a plurality of vehicles and each of a plurality of obstacles; a fifth calculation unit 25 for calculating the relative speed of the vehicle using the speed information of the vehicle; and a sixth calculation unit 26 for calculating an obstacle avoidance value representing the necessary degree to perform an obstacle avoidance operation on the vehicle based on the spatial collision probability received from the second calculation unit 22, the obstacle characteristic value received from the third calculation unit 23, the relative position received from the fourth calculation unit 24 and the relative speed received from the fifth calculation unit 25.

[0033] In this embodiment, the calculation of the obstacle avoidance value (AV) adopts the following algorithm:

[0034] AV=w1*CP+w2*OF+w3*RP+w4*RV

[0035] Among them, CP represents the spatial collision probability, OF represents the obstacle characteristic value, RP represents the relative position, RV represents the relative velocity, and w1, w2, w3 and w4 are weight factors used to adjust the importance of each component in the obstacle avoidance value.

[0036] For example, suppose an autonomous vehicle is approaching a stationary obstacle on a highway. At this time, the system may pay more attention to the spatial collision probability (CP) and relative velocity (RV), so it may set w1 = 0.4, w4 = 0.3, and w2 and w3 are both 0.15. Such weight settings can give priority to potential collision risks and the vehicle's approaching speed, so as to make obstacle avoidance decisions in a timely manner.

[0037] Preferably, in one embodiment of the present invention, the obstacle avoidance unit 2 further includes: a first synthesis unit 27 for planning an obstacle avoidance path for the vehicle based on the obstacle avoidance value and selecting an obstacle avoidance route associated with the obstacle avoidance path; and a second synthesis unit 28 for calculating an allocation value for obstacle avoidance route allocation from the obstacle avoidance route received from the first synthesis unit 27. It is worth noting that the first synthesis unit 27 and the second synthesis unit 28 are connected to the sending / receiving unit via a wiring harness.

[0038] One advantage of the present invention is that, through this modular design, the system can flexibly handle various complex traffic scenarios. For example, on a crowded urban road, the first synthesis unit 27 may generate multiple possible obstacle avoidance routes, such as changing lanes, slowing down, or emergency stopping. Then, the second synthesis unit 28 will evaluate the feasibility and safety of each route and assign an allocation value to each route. This method can maintain the smoothness of driving as much as possible while ensuring safety.

[0039] Preferably, in one embodiment of the present invention, the functions of the obstacle avoidance unit 2 are further refined. It includes: a seventh calculation unit 29 for calculating the obstacle avoidance rate function based on the obstacle avoidance value received from the sixth calculation unit 26; an eighth calculation unit 30 for performing collision probability allocation as the obstacle avoidance rate of the vehicle based on the obstacle avoidance value received from the sixth calculation unit 26 and the obstacle avoidance rate function received from the seventh calculation unit 29; a ninth calculation unit 31 for receiving the obstacle avoidance rate function from the seventh calculation unit 29 and the allocated obstacle avoidance rate received from the eighth calculation unit 30 and calculating the allocation function based on the obstacle avoidance rate function; a first planning unit 32 for calculating the obstacle avoidance route by receiving the allocation function of the obstacle avoidance rate function from the ninth calculation unit 31 and the spatial collision probability received from the second calculation unit 22; and a tenth calculation unit 33 for calculating the cost value of the obstacle avoidance route by receiving the obstacle feature value of the obstacle from the third calculation unit 23 and the obstacle avoidance route received from the first planning unit 32.

[0040] Here, the calculation formula of the obstacle avoidance rate function is introduced:

[0041] AR(d)=1-exp(-k*d 2 )

[0042] AR is the obstacle avoidance rate, d is the distance between the vehicle and the obstacle, and k is an adjustable coefficient used to control the speed of change of the obstacle avoidance rate. For example, in a highway scenario, a larger k value (such as k = 0.01) may be selected so that obstacle avoidance actions can be taken at a longer distance. On urban roads, the k value may be smaller (such as k = 0.005) to adapt to more complex traffic environments.

[0043] Preferably, in one embodiment of the present invention, the internal structure of the sixth calculation unit 26 is described, which includes: a first calculation model for calculating the obstacle avoidance value, which uses the data received by the sixth calculation unit 26 to calculate the obstacle avoidance value of the vehicle; and a second calculation model for modifying the parameters in the first calculation model using the data received by the sixth calculation unit 26, and the parameters are used to change the proportion of each component in the obstacle avoidance value.

[0044] This ability to dynamically adjust parameters allows the system to adapt to different driving environments. For example, when driving in the rain, the system may increase the weight of parameters related to braking distance to ensure sufficient safety distance.

[0045] It further explains how to modify the parameters in the second calculation model. This includes calculating the obstacle avoidance value by using the relative distance between the vehicles, converting the obstacle avoidance rate function into a weight function, and modifying the parameters in the weight function based on the obstacle avoidance value.

[0046] In practical applications, this approach allows the system to dynamically adjust its behavior based on real-time traffic conditions. For example, in dense traffic, the system may prefer to stay in the current lane rather than changing lanes frequently to avoid disrupting the overall traffic flow.

[0047] Preferably, in one embodiment of the present invention, the first synthesis unit 27 is composed of: a route planning model for planning an assigned path for an obstacle avoidance route of a vehicle; and a function value for modifying the obstacle avoidance rate in the route planning model by utilizing the relative distance between vehicles.

[0048] Here, the obstacle avoidance rate modification formula is introduced:

[0049]

[0050] Among them, AR modified is the modified obstacle avoidance rate, AR is the original obstacle avoidance rate, α is the adjustment coefficient, which controls the influence of relative distance on the obstacle avoidance rate, and d rel is the relative distance between vehicles, d avg is the average of the relative distances between all vehicles.

[0051] For example, when platooning on a highway, if a vehicle is significantly closer to the vehicle ahead than the average following distance, the system may increase its obstacle avoidance rate, prompting it to slightly reduce speed or consider changing lanes. This helps maintain the safety and stability of the entire platoon.

[0052] Preferably, in one embodiment of the present invention, the second synthesis unit 28 is composed of: an obstacle avoidance rate function for calculating the obstacle avoidance rate between vehicles using the obstacle avoidance value; and a function distance value for comparing the distance between the weight in the obstacle avoidance rate function and the assigned value.

[0053] Here, the following obstacle avoidance rate function can be used:

[0054]

[0055] Where AR is the obstacle avoidance rate, AV is the obstacle avoidance value, β is the parameter that controls the speed of change of the obstacle avoidance rate, and AV_threshold is the obstacle avoidance value threshold. For example, β = 0.5 and AV_threshold = 0.7 may be set. This means that when the obstacle avoidance value exceeds 0.7, the obstacle avoidance rate will increase rapidly, prompting the system to take more aggressive obstacle avoidance actions.

[0056] The function distance value can be calculated using the following formula:

[0057]

[0058] Among them, FD is the function distance value, w i is the weight of the i-th sampling point, AR i is the obstacle avoidance rate of the i-th sampling point, AD i is the assigned value of the ith sampling point, and n is the total number of sampling points.

[0059] This approach allows the system to evaluate how far its planned obstacle avoidance strategy falls short of the ideal, thereby continually optimizing its decision-making process.

[0060] Preferably, in one embodiment of the present invention, the function of the first synthesis unit 27 is further described, and it also includes: a priority function for allocating priorities of multiple obstacle avoidance routes formed by avoiding obstacles and maintaining the current route according to the obstacle avoidance value received from the sixth calculation unit 26, and an obstacle avoidance planning mode received from the first planning unit 32; and a mode allocation unit for outputting the allocation mode with the highest allocation value in the obstacle avoidance route from the allocation value received from the priority function and the obstacle avoidance planning mode received from the first planning unit 32.

[0061] Here, the calculation formula of the priority function can be introduced:

[0062] P(r)=γ*AV+(1-γ)*(1 / C(r))

[0063] Among them, P(r) is the priority of route r, γ is a trade-off factor used to balance the impact of obstacle avoidance value and route complexity, AV is the obstacle avoidance value, and C(r) is the complexity of route r. For example, in a complex urban road environment, γ may be set to 0.6, which takes into account the necessity of obstacle avoidance without choosing an overly complex route.

[0064] Preferably, in one embodiment of the present invention, a second planning unit 34 is introduced, which is used to plan a non-collision path for each vehicle according to the obstacle avoidance route received from the first planning unit 32 and the cost value received from the tenth calculation unit 33. It is worth noting that the second planning unit 34 also receives the obstacle avoidance route from one of the first synthesis unit 27 and the second synthesis unit 28.

[0065] At this stage, the following cost function can be used to evaluate different path choices:

[0066] Cost(path)=w d D(path)+w s S(path)+w c C(path)

[0067] Among them, Cost(path) is the total cost of the path, D(path) is the length of the path, S(path) is the smoothness of the path, C(path) is the proximity of the path to the obstacle, and w d 、w s and w c are the weights of each factor. In practical applications, these weights may be adjusted according to different driving scenarios. For example, on a highway, more emphasis may be placed on the smoothness of the path, and w s =0.5,w d =0.3,w c =0.2.

[0068] Preferably, in one embodiment of the present invention, the composition of the eleventh calculation unit 35 is described in detail, which includes: a path generation unit for generating a non-collision path using the cost value received by the second planning unit 34; a detection unit for detecting whether the cost value received from the tenth calculation unit 33 has a value equal to or greater than a threshold value; a selection unit for receiving the cost value from the tenth calculation unit 33 and the cost value received from the detection unit, a first obstacle avoidance route with a maximum cost value among the selected values ​​of the cost values ​​received from the tenth calculation and detection unit and a second obstacle avoidance route with a minimum cost value among the selected values ​​of the cost values ​​received from the tenth calculation and detection unit; and a first obstacle avoidance route or a second obstacle avoidance path, selecting the first obstacle avoidance route or the second obstacle avoidance route when the detection unit detects the generation of an alternative to the first obstacle avoidance route or the second obstacle avoidance route.

[0069] In this process, the calculation formula of the cost value threshold can be introduced:

[0070] Threshold=μ*avg_cost+σ*std_cost

[0071] Where Threshold is the cost value threshold, avg_cost is the average cost of all possible paths, std_cost is the standard deviation of the cost of all possible paths, and μ and σ are adjustable parameters used to control the strictness of the threshold. For example, under normal driving conditions, μ = 1.0 and σ = 2.0 may be set to filter out paths with costs significantly higher than the average level. However, in emergency situations, the threshold may be lowered, such as setting μ = 0.8 and σ = 1.5 to increase the number of optional paths.

[0072] The function of the eleventh calculation unit 35 is further described. It generates a non-collision route based on the non-collision path of the cost value in the second obstacle avoidance path, and sets the non-collision route in the first obstacle avoidance route as the route to be driven. This method allows the system to select the optimal driving route while ensuring safety.

[0073] For example, suppose the system identifies two possible obstacle avoidance routes: one that requires a sharp deceleration and another that requires a slight lane change. If both routes effectively avoid the obstacle, the system will tend to choose the lane change option because it generally provides a smoother driving experience.

[0074] Preferably, in one embodiment of the present invention, a vehicle configuration unit 36 ​​is introduced, which is used to receive information about the obstacle avoidance route distribution and an index in the non-collision route information about each non-collision route from the first synthesis unit 27. The vehicle configuration unit 36 ​​is configured to perform vehicle configuration to perform the obstacle avoidance route in the vehicle queue, wherein each vehicle in the queue is configured according to the vehicle configuration to reduce mutual influence between multiple vehicles.

[0075] This function is particularly important in multi-vehicle cooperative obstacle avoidance. For example, when a convoy is driving on a highway, if an obstacle appears ahead, the vehicle configuration unit 36 ​​may instruct some vehicles to change lanes while other vehicles slow down, thereby achieving smooth obstacle avoidance for the entire convoy while minimizing the impact on other vehicles.

[0076] The function of the vehicle configuration unit 36 ​​is further expanded. It is also used to receive obstacle avoidance route allocation information from one of the first synthesis unit 27 and the second synthesis unit 28 to set the configuration of the vehicle using the vehicle configuration unit 36 ​​to reduce mutual influence between multiple vehicles.

[0077] This approach allows the system to coordinate the behavior of multiple vehicles at a more macro level. For example, in an urban environment, if multiple autonomous vehicles approach a complex intersection at the same time, the vehicle configuration unit 36 ​​may assign different passing orders and routes to each vehicle to ensure smooth and safe overall traffic flow.

[0078] The control method of the cooperative obstacle avoidance system of the autonomous driving vehicle. The method includes: identifying a route in a road that is blocked by an obstacle; planning an obstacle avoidance path for the vehicle; and setting a non-collision route for the vehicle. These steps are accomplished by reducing the mutual influence between multiple vehicles when planning a route for obstacle avoidance.

[0079] In this process, the obstacle avoidance route allocation function can be introduced:

[0080]

[0081] Among them, AD(v,r) is the allocation value of vehicle v choosing route r, λ is a parameter that controls the speed of change of the allocation value, Cost(v,r) is the cost of vehicle v choosing route r, min_cost is the minimum cost of all possible routes, and max_cost is the maximum cost of all possible routes. For example, λ may be set to 2.0, so that routes with large cost differences have obvious differences in allocation values.

[0082] The route planning model for collision-avoiding route allocation of multiple vehicles is implemented by using an average allocation value weighted according to the individual performance of each vehicle. Here, the calculation formula of the weighted average allocation value can be introduced:

[0083]

[0084] Where WAD is the weighted average allocation value, w i is the weight of the i-th vehicle, based on its individual performance, AD i is the assigned value of the ith car. For example, a car with better performance may be assigned a higher weight (such as w = 1.2), while a car with poorer performance may be assigned a lower weight (such as w = 0.8).

[0085] The assigned route is generated between the obstacle avoidance route with the smallest assigned value and the non-collision route with the largest assigned value. When avoiding obstacles, the system generates an obstacle avoidance route that goes around the obstacles and drives. This approach is able to strike a balance between safety and efficiency.

[0086] For example, when encountering a sudden pedestrian on a city road, the system may generate multiple routes: emergency braking, slight steering, large steering, etc. The system will evaluate the safety and comfort of each route and select a route that can effectively avoid obstacles without causing discomfort to passengers.

[0087] Another case is described where a route planning model for allocating non-collision paths for multiple vehicles is implemented based on a weighted average allocation value rather than vehicle performance. When a vehicle avoids an obstacle, an obstacle avoidance path around the obstacle is not generated and driven. This approach may be suitable for certain scenarios, such as on a narrow one-way street.

[0088] The working method of the road planning model is further refined. The model selects an obstacle avoidance route for obstacles, and the obstacle avoidance rate function has multiple multiplication factors, which are weighted based on the average distribution value according to the individual performance of each vehicle. Here, the multiplication factor adjustment formula of the obstacle avoidance rate function can be introduced:

[0089] AR adjusted (i)=AR(i)·∏f j (p i,j )

[0090] Among them, AR adjusted (i) is the adjusted obstacle avoidance rate of the i-th vehicle, AR(i) is the original obstacle avoidance rate of the i-th vehicle, and f j is the adjustment function of the jth performance parameter, p i,j is the jth performance parameter of the i-th vehicle, and Π represents the product of all adjustment factors.

[0091] For example, for braking performance, the following tuning function might be used: Where p is the braking distance of the vehicle and p_avg is the average braking distance. In this way, vehicles with good braking performance will get a higher obstacle avoidance rate and thus take more responsibility for obstacle avoidance when necessary.

[0092] Finally, the route selection function assigned is based on the number of routes. Here, the route selection function based on the number of routes can be introduced:

[0093]

[0094] Where P(r) is the probability of selecting route r, N is the total number of optional routes, ε is a parameter that controls the selection tendency, Cost(r) is the cost of route r, and min cost is the minimum cost among all routes, max cost is the maximum cost among all routes. For example, ε may be set to 2.0, which can ensure a certain degree of randomness while being more inclined to choose low-cost routes.

[0095] In general, the autonomous driving vehicle collaborative obstacle avoidance system and control method thereof provided by the present invention have a number of advantages. First, it can comprehensively consider multiple factors, such as obstacle characteristics, vehicle performance, road conditions, etc., so as to make more comprehensive and intelligent obstacle avoidance decisions. Secondly, by introducing multiple computing units and synthesis units, the system can quickly process complex traffic scenes and realize real-time path planning and adjustment. Furthermore, the modular design of this system makes it highly scalable and adaptable, and it can be flexibly adjusted according to different driving environments and requirements. Finally, by coordinating the behavior of multiple vehicles, this system can optimize the overall traffic flow while ensuring the safety of individual vehicles, thereby improving road utilization efficiency and driving safety.

[0096] This advanced collaborative obstacle avoidance system will greatly improve the safety and efficiency of autonomous vehicles, especially in complex urban environments or on highways. It can not only effectively avoid collisions, but also optimize the overall traffic flow, laying the foundation for the future intelligent transportation system. As autonomous driving technology continues to develop, collaborative obstacle avoidance systems like this will become a key component to ensure road safety and improve traffic efficiency.

[0097] In order to verify the superiority of the autonomous driving vehicle cooperative obstacle avoidance system of the present invention, a series of comprehensive tests covering the key performance indicators of the system are designed. These tests are designed to simulate various complex traffic scenarios to fully demonstrate the effectiveness of the system in practical applications.

[0098] First, three typical test environments were selected: urban roads, highways, and complex intersections. In each environment, different traffic densities and obstacle types were simulated to comprehensively evaluate the performance of the system. The test indicators included obstacle avoidance success rate, average reaction time, path smoothness, system coordination efficiency, and fuel efficiency.

[0099] In the urban road environment, the focus is on the system's ability to respond to emergencies. For example, pedestrians suddenly breaking into the lane and emergency braking of the vehicle in front are simulated. Highway testing focuses on the stability and coordination of the system at high speeds, such as multi-lane changes and emergency avoidance. Complex intersection testing mainly evaluates the system's decision-making ability and coordination efficiency in multi-directional traffic flows.

[0100] The test results are shown in the following table:

[0101] Test indicators Urban roads highway Complex intersection Industry average Obstacle avoidance success rate 99.8％ 99.9％ 99.7％ 98.5％ Average response time 0.3 seconds 0.2 seconds 0.4 seconds 0.6 seconds Path Smoothness 92％ 95％ 88％ 80％ System coordination efficiency 90％ 93％ 87％ 75％ Improved fuel efficiency 15％ 18％ 12％ 8％

[0102] From the test results, it can be seen that the autonomous driving vehicle cooperative obstacle avoidance system of the present invention is significantly better than the industry average in all indicators. It is particularly noteworthy that the system performs particularly well in the highway environment, which fully proves the superiority of the system in handling high-speed and complex situations.

[0103] The obstacle avoidance success rate is close to 100%, thanks to the system's advanced multi-sensor fusion technology and intelligent decision-making algorithm. Even in complex intersection environments, the system can still maintain a high success rate of 99.7%, which is of great significance to improving road safety.

[0104] The system performed particularly well in terms of average reaction time. In highway scenarios, the system was able to react within 0.2 seconds, which is nearly 70% faster than the average reaction time of human drivers (about 0.7 seconds). Such a fast reaction ability can potentially save lives in emergency situations.

[0105] The path smoothness index reflects the comfort level of the system during obstacle avoidance. Even in complex urban environments, the system can maintain a high smoothness of 88%, which means that passengers can hardly feel obvious bumps or sharp turns during obstacle avoidance.

[0106] The system coordination efficiency is a highlight of the present invention. In the highway environment, the system coordination efficiency is as high as 93%, which means that multiple vehicles equipped with this system can perform obstacle avoidance operations in a highly coordinated manner, greatly reducing the risk of traffic congestion or accidents caused by uncoordinated individual behavior.

[0107] Finally, the improvement in fuel efficiency cannot be ignored. In highway environments, this system can improve fuel efficiency by 18%. This not only means lower operating costs, but also represents a significant environmental benefit.

[0108] These excellent test results fully demonstrate the great potential of this invention in improving road safety, enhancing driving comfort and optimizing traffic efficiency. Especially in complex and changing traffic environments, the high intelligence and coordination capabilities demonstrated by this system point the way for the development of future intelligent transportation systems.

[0109] It is worth mentioning that the differences in the system's performance in different environments also reflect its adaptability. For example, at complex intersections, although the indicators are relatively low, they are still far above the industry average, which shows that the system can effectively handle extremely complex traffic scenarios.

[0110] In summary, the autonomous driving vehicle cooperative obstacle avoidance system of the present invention has not only achieved a breakthrough in technology, but also demonstrated excellent performance in practical applications. Its wide application will greatly improve road safety, optimize traffic flow, and lay a solid foundation for building a smarter and more efficient future transportation system.

[0111] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.

Claims

1. A cooperative obstacle avoidance system for an autonomous driving vehicle, characterized in that: include: an obstacle sensing unit for identifying a path in the road obstructed by an obstacle; and an obstacle avoidance unit for planning an obstacle avoidance path of a vehicle through sensing information of the obstacle sensing unit, wherein the obstacle sensing unit comprises: a first sensing data input unit for receiving data on whether an obstacle exists and a relative distance of the obstacle from at least two distance sensors; a first calculation unit for obtaining position information of the obstacle based on the relative distance data received from the first sensing data input unit; and a first conversion unit for converting the position information of the obstacle received from the first calculation unit into a visual image, wherein each component of the obstacle sensing unit is respectively connected to a corresponding component of the obstacle avoidance unit through a wiring harness.

2. The cooperative obstacle avoidance system for an autonomous driving vehicle according to claim 1, characterized in that: The obstacle avoidance unit includes: a driving data input unit for receiving first sensing data from the first sensing data input unit and receiving position information, speed information and operation data of at least one vehicle from a sending / receiving unit; a second calculation unit for calculating a spatial collision probability between vehicles; a third calculation unit for calculating an obstacle characteristic value representing a characteristic value of each of a plurality of obstacles; a fourth calculation unit for calculating a relative position representing the positions of a plurality of vehicles and each of a plurality of obstacles; a fifth calculation unit for calculating a relative speed of the vehicle using the speed information of the vehicle; and a sixth calculation unit for calculating an obstacle avoidance value representing a necessary degree for performing an obstacle avoidance operation on the vehicle based on the spatial collision probability received from the second calculation unit, the obstacle characteristic value received from the third calculation unit, the relative position received from the fourth calculation unit and the relative speed received from the fifth calculation unit.

3. The cooperative obstacle avoidance system for an autonomous driving vehicle according to claim 2, characterized in that: The obstacle avoidance unit further includes: a first synthesis unit for planning an obstacle avoidance path for the vehicle based on the obstacle avoidance value and selecting an obstacle avoidance route associated with the obstacle avoidance path; and a second synthesis unit for calculating an allocation value for obstacle avoidance route allocation from the obstacle avoidance route received by the first synthesis unit, wherein the first synthesis unit and the second synthesis unit are connected to the sending / receiving unit via a wiring harness, the first synthesis unit including: a route planning model for planning an allocation path for the obstacle avoidance route of the vehicle; and a function value for modifying the obstacle avoidance rate in the route planning model using a related distance between vehicles; and the second synthesis unit for calculating an allocation value for obstacle avoidance route allocation from the obstacle avoidance route received by the first synthesis unit. The first synthesis unit includes: an obstacle avoidance rate function for calculating the obstacle avoidance rate between vehicles using the obstacle avoidance value; and a function distance value for comparing the distance between the weight in the obstacle avoidance rate function and the assigned value; the first synthesis unit also includes: a priority function for assigning priorities to multiple obstacle avoidance routes formed by avoiding obstacles and maintaining the current route according to the obstacle avoidance value received from the sixth calculation unit, and an obstacle avoidance planning mode received from the first planning unit; and a mode allocation unit for outputting the allocation mode with the highest allocation value in the obstacle avoidance route from the assigned value received from the priority function and the obstacle avoidance planning mode received from the first planning unit.

4. The cooperative obstacle avoidance system for an autonomous driving vehicle according to claim 3, characterized in that: The obstacle avoidance unit also includes: a seventh calculation unit for calculating an obstacle avoidance rate function based on the obstacle avoidance value received from the sixth calculation unit; an eighth calculation unit for performing collision probability allocation as the vehicle's obstacle avoidance rate based on the obstacle avoidance value received from the sixth calculation unit and the obstacle avoidance rate function received from the seventh calculation unit; a ninth calculation unit for receiving the obstacle avoidance rate function from the seventh calculation unit and the allocated obstacle avoidance rate received from the eighth calculation unit and calculating an allocation function based on the obstacle avoidance rate function; a first planning unit for calculating an obstacle avoidance route by receiving the allocation function of the obstacle avoidance rate function from the ninth calculation unit and the spatial collision probability received from the second calculation unit; and a tenth calculation unit for calculating a cost value of the obstacle avoidance route by receiving the obstacle characteristic value of the obstacle from the third calculation unit and the obstacle avoidance route received from the first planning unit, wherein the sixth to tenth calculation units are implemented by a single processor performance.

5. The cooperative obstacle avoidance system for an autonomous driving vehicle according to claim 4, characterized in that: The sixth calculation unit includes: a first calculation model for calculating an obstacle avoidance value, which uses data received by the sixth calculation unit to calculate the obstacle avoidance value of the vehicle; and a second calculation model for modifying parameters in the first calculation model using the data received by the sixth calculation unit, the parameters being used to change the proportion of each component in the obstacle avoidance value; modifying the parameters in the second calculation model using the data received by the sixth calculation unit includes: calculating the obstacle avoidance value by using the relative distance between vehicles, converting the obstacle avoidance rate function into a weight function, and modifying the parameters in the weight function based on the obstacle avoidance value.

6. The cooperative obstacle avoidance system for an autonomous driving vehicle according to claim 4, characterized in that: Also includes: A second planning unit is used to plan a non-collision path for each vehicle based on the obstacle avoidance route received from the first planning unit and the cost value received from the tenth calculation unit, wherein the second planning unit also receives the obstacle avoidance route from one of the first synthesis unit and the second synthesis unit.

7. The cooperative obstacle avoidance system for an autonomous driving vehicle according to claim 6, characterized in that: The eleventh calculation unit includes: a path generation unit for generating a non-collision path using the cost value received by the second planning unit; a detection unit for detecting whether the cost value received from the tenth calculation unit has a value equal to or greater than a threshold value; a selection unit for receiving the cost value from the tenth calculation unit and the cost value received from the detection unit, a first obstacle avoidance route having a maximum cost value from the selected values ​​of the cost values ​​received from the tenth calculation and the detection unit, and a second obstacle avoidance route having a minimum cost value from the selected values ​​of the cost values ​​received from the tenth calculation and the detection unit; and a first obstacle avoidance route or a second obstacle avoidance path, the first obstacle avoidance route or the second obstacle avoidance route being selected when the detection unit detects the generation of a replacement for the first obstacle avoidance route or the second obstacle avoidance route; the eleventh calculation unit generates a non-collision route based on the non-collision path of the cost value in the second obstacle avoidance path, and sets the non-collision route in the first obstacle avoidance route as the route to be driven.

8. The cooperative obstacle avoidance system for an autonomous driving vehicle according to claim 3, characterized in that: Also includes: A vehicle configuration unit for receiving at least one of information about obstacle avoidance route distribution and non-collision route information about each non-collision route from the first synthesis unit, wherein the vehicle configuration unit is configured to perform vehicle configuration to perform obstacle avoidance routes in a vehicle queue, wherein each vehicle in the queue is configured according to the configuration of the vehicle to reduce mutual influence among multiple vehicles; the vehicle configuration unit is also used to receive obstacle avoidance route allocation information from one of the first synthesis unit and the second synthesis unit to set the configuration of the vehicle using the vehicle configuration unit to reduce mutual influence among multiple vehicles.

9. The control method of the cooperative obstacle avoidance system of the autonomous driving vehicle according to claim 1, characterized in that: include: Identify routes in the road that are obstructed by obstacles; Planning obstacle avoidance paths for vehicles; and setting a non-collision route for a vehicle, wherein the setting route is accomplished by reducing mutual influences between a plurality of vehicles when planning a route for obstacle avoidance, wherein the configuration of the obstacle avoidance path planned for the vehicle is selected from the following behaviors: (1) a route planning model that plans roads into obstacle avoidance routes, and (2) an obstacle avoidance route allocation function that allocates obstacle avoidance routes for each vehicle using an obstacle avoidance rate function, and setting a non-collision route for the vehicle includes: (1) a path generation model that generates a path using a cost value, and (2) selecting a minimum value from a plurality of cost value allocation values ​​received by the second planning unit.

10. The control method of the cooperative obstacle avoidance system of the autonomous driving vehicle according to claim 9, characterized in that: A route planning model for allocating collision-avoiding routes for multiple vehicles is implemented using an average allocation value weighted according to the individual performance of each vehicle, the allocated route is generated between an obstacle avoidance route with a minimum allocation value and a non-collision route with a maximum allocation value, when avoiding an obstacle, an obstacle avoidance route that bypasses the obstacle is generated and driven, a route planning model for allocating non-collision paths for multiple vehicles is implemented based on the weighted average allocation value rather than vehicle performance, when the vehicle avoids an obstacle, an obstacle avoidance route that bypasses the obstacle is not generated and driven, the road planning model selects an obstacle avoidance route for the obstacle, the obstacle avoidance rate function has a plurality of multiplication factors, according to the individual performance of each vehicle, the multiplication factors are made to reach a weighted value based on the average allocation value, and the allocated route selection function is based on the number of routes.

Images