WO2023092837A1

WO2023092837A1 - Collision detection

Info

Publication number: WO2023092837A1
Application number: PCT/CN2022/071344
Authority: WO
Inventors: 黄超; 叶玥
Original assignee: 上海仙途智能科技有限公司
Priority date: 2021-11-26
Filing date: 2022-01-11
Publication date: 2023-06-01
Also published as: CN115292796B; CN115292796A

Abstract

The present application provides a collision detection method and device. The method comprises: obtaining a self-vehicle body model, a vehicle body model comprised in the self-vehicle body model being a multi-circle model; performing collision detection according to the self-vehicle body model, wherein under the condition that a rectangular vehicle body of the unmanned vehicle is equally divided into n rectangles in a long side direction, the circle centers of n circles comprised in the multi-circle model are respectively center points of the n rectangles, the radius of each circle is a distance from any center point to the vertex of the rectangle where the center point is located, and n is a positive integer larger than 1. By applying this solution, the unmanned vehicle model can be prevented from occupying too many safety areas, and the precision of the unmanned vehicle body model is improved.

Description

Impact checking

technical field

The present application relates to the field of artificial intelligence, in particular to a method and device for collision detection.

Background technique

With the rapid development of artificial intelligence technology, unmanned driving has become a popular research direction of artificial intelligence. The collision detection of moving obstacles is mainly realized based on whether the driving area of the self-vehicle overlaps with the driving area of the obstacle.

The driving track area of an unmanned vehicle is generally related to the shape model of the vehicle. The shape model of a conventional unmanned vehicle is to represent the entire body with a rectangular model, but the length and width of the rectangular model need to be proportional to the length and width of the vehicle Zooming, the accuracy is too poor, and a large number of safe areas around the vehicle will be framed. In order to reduce the safe area occupied by the vehicle shape model, relevant technical personnel proposed the traditional five-circle model, but the traditional five-circle model will cause the vehicle shape model to expand too much in width, and the accuracy is not high. Occupies a lot of safe area.

Contents of the invention

In view of this, this application provides a method and device for collision detection, specifically, this application is achieved through the following technical solutions:

According to the first aspect of the present application, a method for collision detection is provided, the method is applied to an unmanned vehicle, and the method includes: obtaining the body model of the self-vehicle, and the body body model contained in the body model of the self-vehicle is multiple Circle model; Carry out collision detection according to the self-vehicle shape model; Wherein, when the rectangular body of the unmanned vehicle is divided into n rectangles along the long side direction, the n The centers of the circles are respectively the center points of the n rectangles, and the radius of each circle is the distance from any center point to the apex of the rectangle where it is located, and n is a positive integer greater than 1.

Optionally, in the collision detection method, when the unmanned vehicle has one or more accessories beyond the vehicle body, the body model of the self-vehicle also includes a body model of the accessories; the body model of the accessories A single circle model corresponding to each attachment is included; the center of the single circle model is located at the midpoint of the line connecting the two furthest points on the corresponding attachment, and the diameter is the length of the line.

Optionally, in the collision detection method, when the unmanned vehicle has one or more accessories beyond the vehicle body, it also includes: obtaining the boundary of the road terrain within a preset distance in front of the vehicle; calculating In the case that all the accessories of the vehicle are deployed, the distance from the center of each accessory body model to the terrain boundary; if the distance is greater than the radius of the corresponding accessory body model, then expand the accessory corresponding to the accessory body model; if the If the distance is not greater than the radius of the corresponding accessory shape model, then the accessory corresponding to the accessory shape model is put away.

Optionally, in the collision detection method, the collision detection based on the body model of the ego vehicle includes: obtaining the driving parameters of the ego vehicle and the driving parameters of the movable obstacle; the ego vehicle driving parameters include ego vehicle The driving trajectory and the driving speed of the own vehicle, the driving parameters of the movable obstacle include the driving trajectory of the movable obstacle and the driving speed of the movable obstacle; according to the driving parameters of the own vehicle and the driving parameters of the movable obstacle, if It is predicted that the driving area corresponding to the body model of the ego vehicle will overlap with the driving area of the movable obstacle, and then it is determined that there is a collision risk.

Optionally, in the collision detection method, the collision detection based on the ego vehicle shape model further includes: calculating the collision moment between the unmanned vehicle and the movable obstacle, the collision The time is when the driving area corresponding to the body model of the own vehicle overlaps with the driving area of the movable obstacle; the speed of the own vehicle is replanned according to the collision time to eliminate the risk of collision.

Optionally, in the method for collision detection, it also includes: acquiring a variety of ego vehicle shape models; the multiple ego vehicle shape models include a plurality of multi-circle models with different numbers of circles; When the current road congestion situation is congested, select a model with a large number of circles for collision detection; if the current road congestion situation is smooth, select a model with a small number of circles for collision detection.

Optionally, in the collision detection method, the multi-circle model is a five-circle model.

According to the second aspect of the present application, a collision detection device is provided, the device is applied to an unmanned vehicle, including:

The model acquisition module is used to obtain the body model of the self-vehicle, and the body shape model contained in the body model of the self-vehicle is a multi-circle model; the collision detection module is used to perform collision detection according to the body model of the self-vehicle; wherein, in the In the case where the rectangular body of the unmanned vehicle is divided into n rectangles along the long side direction, the centers of the n circles included in the multi-circle model are respectively the center points of the n rectangles, and each The radius of a circle is the distance from any center point to the apex of the rectangle where it is located, and n is a positive integer greater than 1.

According to a third aspect of the present application, a storage medium is provided, the storage medium stores computer program instructions, and after the computer program instructions are executed, the collision detection method provided by the embodiment of the foregoing first aspect can be implemented.

According to a fourth aspect of the present application, there is provided an electronic device, including: a memory for storing computer program instructions and a processor for executing the computer program instructions, wherein, when the computer program instructions are executed by the processor, it can realize The collision detection method provided by any of the foregoing embodiments.

Through the technical solution provided by this application, the rectangular body of the unmanned vehicle is divided into multiple rectangles along the long side direction, with the center points of the multiple rectangles as the center, and the distance from the center point to the apex of each corresponding rectangle is Radius, build a multi-circle model, by reducing the radius of the largest circle in the model, avoid the model occupying too much safe area, and improve the accuracy of the unmanned vehicle shape model.

Description of drawings

Fig. 1 is a kind of traditional five-circle model schematic diagram shown in the present application;

Fig. 2 is a schematic diagram of a new five-circle model shown in the present application;

Fig. 3 is a flowchart of a method for collision detection shown in the present application;

Fig. 4 is a schematic diagram of a seven-circle model shown in the present application;

Fig. 5 is a schematic diagram of a three-circle model shown in the present application;

FIG. 6 is a flowchart of specific steps of a collision detection method shown in the present application;

Fig. 7 is a schematic diagram of automatic driving of a cleaning vehicle shown in the present application;

Fig. 8 is a hardware structure diagram of an unmanned vehicle where a collision detection device of the present application is located;

Fig. 9 is a block diagram of a collision detection device shown in the present application.

Detailed ways

Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numerals in different drawings refer to the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with aspects of the present application as recited in the appended claims.

The terminology used in this application is for the purpose of describing particular embodiments only, and is not intended to limit the application. As used in this application and the appended claims, the singular forms "a", "the", and "the" are intended to include the plural forms as well, unless the context clearly dictates otherwise. It should also be understood that the term "and/or" as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

It should be understood that although the terms first, second, third, etc. may be used in this application to describe various information, the information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of the present application, first information may also be called second information, and similarly, second information may also be called first information. Depending on the context, the word "if" as used herein may be interpreted as "at" or "when" or "in response to a determination."

Since the projection of the vehicle body in the vertical direction is usually close to a rectangle, the traditional five-circle model is usually used when constructing a body model for an unmanned vehicle in the related art. Fig. 1 is a schematic diagram of a traditional five-circle model shown in the present application. As shown in Fig. 1, the traditional five-circle model includes a large circle and four small circles, the center H of the large circle is the center point of the rectangular body, and the four small circles The centers of the circles are respectively located on the angle bisectors of the four corners of the rectangle, as shown in Figure 1, the line connecting the center P of the small circle and the vertex Q is the angle bisector of the corners where the vertex Q is located, in order to make the five circles just cover the rectangle , the intersection point between two adjacent circles needs to be on the side of the rectangle, such as J in Figure 1. Taking a vehicle with length a and width b as an example, the radius of the small circle is

The radius of the great circle is

However, as the size of the vehicle to which the five-circle model is applied increases, it is gradually found that the use of this model by medium and large vehicles will lead to a large area of invalid collision area. Taking a medium and large bus with a length of 10 meters and a width of 2.5 meters as an example, the radius of the big circle in the traditional five-circle model reaches 3.95 meters, which will cause the body model to expand by 5.4 meters compared to the original width of the body, occupying a safe area It has reached 31.34 square meters, resulting in a large invalid collision area, and cannot automatically drive in a narrow area.

In order to prevent the body model from expanding too much as the size of the vehicle increases, this application considers reducing the radius of the great circle in Figure 1, and constructs a new five-circle model as the body model of the unmanned vehicle, as shown in Figure 2 Fig. 2 is a schematic diagram of a new five-circle model shown in the present application.

A new five-circle model is constructed based on the plane rectangular body projected by the length and width of the unmanned vehicle in China. Specifically, the rectangular body is evenly divided into five rectangles along the long side, and the center point of each rectangle is is the center of the circle, and the distance from the center point to the apex is the radius to construct a circle. After combining five circles, a new five-circle model is obtained. As shown in Figure 2, the body is evenly divided into five rectangles. Taking the rectangle above the body as an example, the center of the circle is the center point N of the rectangle, and the radius is the distance from the center point N to the vertex M. The intersection point D of each circle is located on the long side, and the whole model can well cover the whole bus body. When the length of the bus is a and the width of the bus is b, the radius of each circle is

Still taking a medium-to-large bus with a length of 10 meters and a width of 2.5 meters as an example, the radius of the circle in the new five-circle model is 1.6 meters, which expands by 0.7 meters from the original width of the bus body, occupying a safe area of 6.86 square meters , compared with the traditional five-circle model, it takes up less safety space and has higher accuracy in covering the body, enabling unmanned vehicles to better adapt to narrow passages.

The vehicle body shape model in this application is a multi-circle model, the number of circles is n, and n is a positive integer greater than 1. The center point of these n rectangles is the center of the circle, and the distance from each center point to the corresponding rectangle vertex is the radius. Circle, the shape model of the unmanned vehicle is obtained by combining n circles.

Based on the new polygon model, the embodiment of the present invention provides a collision detection method applied to unmanned vehicles, please refer to Fig. 3, Fig. 3 is a flow chart of a collision detection method shown in this application As shown in the figure, the specific steps are as follows: Step S102: Obtain the body model of the self-vehicle, and the body model of the vehicle body contained in the body model of the self-vehicle is a multi-circle model.

The body model of the ego vehicle includes at least the body shape model. When the unmanned vehicle has one or more accessories beyond the body, the body model of the ego vehicle also includes the body model of the accessories. The body model of the accessories is generally a single circular body model.

The body shape model of an unmanned vehicle is a multi-circle model, and the number of circles in the multi-circle model can be set by yourself. In order to avoid the body model taking up too much safe space, you can increase the number of circles in the multi-circle model To cover the body, thereby reducing the radius of the circle and improving the accuracy of the body model, but if the number of circles is too large, it will increase the calculation pressure of the unmanned vehicle when performing collision detection, so in order to balance the body model In terms of accuracy and calculation pressure, a multi-circle model of five circles is generally selected as the body model.

The present application does not limit the number of circles contained in the adopted multi-circle model. For example, if you pay more attention to the accuracy of the body model, you can increase the number of circles. For example, you can choose the seven-circle model as shown in Figure 4. Figure 4 is a schematic diagram of a seven-circle model shown in this application. ; Or, if you pay more attention to the working pressure of unmanned vehicles on collision detection, you can relatively reduce the number of circles. For example, you can choose the three-circle model as shown in Figure 5. Figure 5 is a kind of Schematic diagram of the three-circle model.

Step S104: Perform collision detection according to the body model of the ego vehicle.

Specifically, please refer to FIG. 6 for the steps of collision detection based on the body model of the ego vehicle. FIG. 6 is a flow chart of specific steps of a collision detection method shown in this application. parameters and driving parameters of movable obstacles.

The traveling parameters of the self-vehicle include the traveling trajectory of the self-vehicle and the traveling speed of the self-vehicle, and the traveling parameters of the movable obstacle include the traveling trajectory of the movable obstacle and the traveling speed of the movable obstacle.

During the driving process of the unmanned vehicle, the driving parameters of the self-driving vehicle and the driving parameters of the movable obstacle can be obtained every first preset time period t1, and the specific method can refer to related technologies.

Step S204: According to the driving parameters of the ego vehicle and the driving parameters of the movable obstacle, if it is predicted that the driving area corresponding to the body model of the ego vehicle will overlap with the driving area of the movable obstacle, then it is determined that there is a risk of collision.

After the unmanned vehicle obtains the driving parameters of the own vehicle and the driving parameters of the movable obstacle, it calculates the driving area and the movable obstacle of the own vehicle within the second preset time period t2 respectively according to the driving parameters of the own vehicle and the driving parameters of the movable obstacle. The driving area of the obstacle, and judge whether there will be overlap between the two driving areas. The driving area refers to the area where the body shape model has traveled within the time t2 under the current speed plan, and t2 is not greater than t1.

Judging whether two driving areas overlap can be calculated by the Sutherland-Hodgman algorithm (edge-by-edge clipping algorithm), and the specific calculation method can refer to relevant literature.

If the calculation result is that the two driving areas overlap, it means that there is a risk of collision between the own vehicle and the movable obstacle.

By judging whether the driving area of the unmanned vehicle overlaps with the object driving area of the movable obstacle, the movable obstacle without collision risk can be removed, the workload of the unmanned vehicle can be reduced, and the efficiency of collision detection can be improved.

Step S206: Calculate the collision time between the unmanned vehicle and the movable obstacle.

The collision time is the time when the driving area corresponding to the body model of the ego vehicle overlaps with the driving area of the movable obstacle. After it is determined that there is a risk of collision between the ego vehicle and the movable obstacle, their collision time is further calculated.

Specifically, when the ego vehicle and the movable obstacle maintain a constant speed, the distance from the edge of the shape model of the movable obstacle to the center of all circles of the body shape model of the unmanned vehicle is calculated to avoid the collision between the movable obstacle and the unmanned vehicle. When any part of the body of the driving vehicle collides, when the vehicle has accessories that exceed the body of the vehicle, the distance from the edge of the shape model of the movable obstacle to the center of the shape model of the accessory needs to be calculated. When the distance is the same as the radius of the circle to which the corresponding center belongs, it is considered that the vehicle will collide with the movable obstacle, and the collision time at this time is recorded; if the collision time is not calculated, it is considered that the vehicle collides with the movable obstacle. Obstacles do not collide.

By calculating the distance from the movable obstacle with collision risk to the center of the self-vehicle body model, it is not necessary to calculate the distance with all points on the side of the self-vehicle body model, which reduces the load pressure on the unmanned vehicle.

Step S208: Replan the vehicle's speed according to the collision time to eliminate the collision risk.

After calculating the collision time between the ego vehicle and the movable obstacle, the ego vehicle speed can be increased or decreased. If the collision cannot be avoided by changing the speed for the preset number of times, the ego vehicle speed will be reduced to 0 and stop moving. By changing The speed of the ego vehicle can be used to eliminate the risk of collision; or, when the collision cannot be avoided by changing the speed for a predetermined number of times, the driving path of the ego vehicle can be changed without violating the road driving rules, so that the movable obstacle can be adjusted to the shape model of the ego vehicle. The distances between all the centers of the circles are larger than the radius of the circles to which the corresponding centers belong, so as to realize adaptive collision detection.

In addition to the aforementioned vehicle body body model, the vehicle body model of the present application may also include an accessory body model. When an unmanned vehicle has one or more attachments beyond the vehicle body, a single circle model can be set for these vehicle attachments as an attachment shape model. Taking a sweeper as an example, as shown in Figure 7, Figure 7 is an illustration of the application. A schematic diagram of automatic driving of a sweeping vehicle is presented, and the body shape model of the sweeping vehicle is a new five-circle model. A sweeper generally has two brushes, as shown in Figure 7, brush A and brush B of the sweeper. The user can pre-configure the shape model of the attachment based on the size of the brush. An attachment shape model generally only corresponds to a single working attachment. The shape model is generally a single circle. The midpoint of the line connecting the two furthest points on the brush is used as the center point, and half the length of the line is used as the radius to construct a circle. The radius can also be appropriately increased according to requirements.

The accessories are not limited to the brush of the sweeper, but also include the sprinkler of the sprinkler, the blade of the bulldozer, etc. An accessory includes two states of unfolding and retracting, such as the sweep of the sweeper, the unfolded state is to move the brush Put it down to clean the road, and put it away to put the brush away and wait for it to stand by.

Since the conditions of the road on which unmanned vehicles drive will not remain unchanged, for example, bicycles may be parked on both sides of the road, or road construction signs occupy the road, etc., which will affect the originally preset driving route of unmanned vehicles. Accessories for driverless vehicles do not work properly.

At this time, the unmanned vehicle can keep the attachment in the closed state in the area where the attachment is not needed; in the area where the attachment is required, the boundary of the road terrain within a certain distance in front of the vehicle can be obtained, because the body model of the vehicle is covered when the attachment is deployed The area is the largest, so regardless of whether the vehicle accessories are currently unfolded or retracted, the unmanned vehicle can calculate the distance from the center of each accessory shape model to the terrain boundary under the assumption that all vehicle accessories are deployed.

When the distance from the center of each accessory shape model to the terrain boundary is greater than the radius of the corresponding accessory shape model, it indicates that the road terrain in front of the vehicle will not collide with the vehicle accessory, so if the accessory corresponding to the accessory shape model is currently in the In the unfolded state, the current state will continue to be maintained. If the attachment is currently in the retracted state, the attachment will be expanded; when the distance from the center of each attachment body model to the terrain boundary is not greater than the radius of the corresponding attachment body model, it indicates that the vehicle The road terrain ahead will collide with the accessory of the vehicle, so if the accessory corresponding to the accessory shape model is currently in the retracted state, keep the current state, and if the accessory is currently in the unfolded state, then retract the accessory.

After the unmanned vehicle obtains the boundary of the road terrain within a certain distance in front of the vehicle, if it detects that there is an attachment that will collide with the road terrain in front of the vehicle, and the attachment cannot be deployed, it can also calculate the distance between the front road boundaries If the distance between the road boundaries is greater than the width of the self-vehicle when the self-vehicle accessories are all deployed, the driving path can be re-planned and the self-vehicle accessories can continue to be in the unfolded state for work.

Taking the sweeper as an example, as shown in Figure 7, the radius of the sweeping brush body model corresponding to the sweeping brush of the sweeper is r. When all the sweeping brushes are in the unfolded state, the width of the body is m. After the terrain boundary of the road, you can calculate the width n of the road, and the distance d from the center of the sweeping shape model in the unfolded state to the terrain boundary. If d>r, it is considered that the sweeper has enough space to the terrain ahead Put down the sweeping brush to clean the ground, and keep the sweeping brush in the unfolded state; if d≤r, it is considered that the terrain ahead of the sweeper is too narrow. In order to avoid collision between the sweeping brush and the terrain, it is necessary to keep the sweeping brush in the retracted state at this time .

Since each brush is independent, for example, when the terrain on the side of brush A in Figure 7 is narrow, while the terrain on the side of brush B is wide, you can only put away brush A to avoid collisions, and brush B remains The ground is cleaned in the unfolded state; or when the width of the road is n≥m, the driving route can be re-planned to shift the driving route of the sweeper to the side of the brush B, so that the distance between the road terrain boundary and the brush A is greater than that of the brush A The radius of the attachment model, to avoid the brush A from colliding with the road terrain.

An independent accessory shape model is set for each accessory of the unmanned vehicle. When the preset driving road changes, a single accessory can be folded or unfolded to avoid collision between the accessory and the changed road, and dynamically adapt to the changed environment. road conditions.

In the aforementioned step S208, after the collision time between the movable obstacle and the own vehicle is calculated, it can also be judged whether it is an accessory of the own vehicle that collides with the movable obstacle, and if not, re-plan the driving speed or trajectory of the own vehicle; If so, when calculating the retracted state of the accessory, whether the distance from the center of the accessory shape model to the side of the movable obstacle shape model is greater than the radius of the accessory shape model, if greater, then put the accessory away and keep the current self The driving parameters of the vehicle remain unchanged. If it is not greater than , then re-plan the driving speed or trajectory of the own vehicle.

By judging whether the shape model of the accessory collides with the movable obstacle, and then determining whether to put away the accessory to avoid the collision, the driving efficiency of the unmanned vehicle can be improved, and it will not be slowed down due to changing the driving speed of the vehicle or replanning the driving route. The journey of a human-driven vehicle.

Since the congestion on the road is different at different times, in order to adapt to different road conditions, the administrator can pre-configure a variety of ego vehicle body models on the unmanned vehicle, and each ego vehicle body model includes a different number of circles.

During the driving process of the unmanned vehicle, the current road congestion under the driving trajectory can also be obtained. If the current road is congested, the self-vehicle shape model with a large number of circles is selected to improve the accuracy of the self-vehicle shape model completely covering the vehicle. Then reduce the safe area occupied by the ego vehicle shape model to avoid invalid collision detection; if the current road is smooth, choose the ego vehicle shape model with a small number of circles, so that the unmanned vehicle can reduce the calculation of road terrain when performing collision detection The pressure of the distance between the shape model of the boundary or the movable obstacle and the body model of the ego vehicle makes a faster response to the impending collision and improves the performance of the unmanned vehicle in dealing with unexpected situations.

For example: the administrator can pre-configure three self-vehicle body models on the unmanned vehicle, one is a body model of three circles, one is a body model of five circles, and the other is a body model of seven circles Body model, after the unmanned vehicle obtains the current road congestion situation under the trajectory of the vehicle, select the body model of the vehicle according to the situation: if the road congestion is serious, select the body model of seven circles to reduce the radius of each circle , improve the accuracy of the model, and avoid misjudgment of collision detection when the unmanned vehicle is driving; if the road is very smooth, choose a three-circle body shape model to reduce the computing pressure of the unmanned vehicle and improve the ability to deal with emergencies; In general, the body model of five circles is selected to reduce the calculation pressure of unmanned vehicles while maintaining the accuracy of the model, and maintain a balance between the two.

Through the technical solution provided by this application, using a new shape model, multiple identical circles are evenly distributed up and down on the rectangular body of the vehicle, reducing the excessive lateral expansion of the model due to the enlarged body and occupying too much safe area. problem, improving the accuracy of body models for unmanned vehicles. At the same time, using the technical solution proposed in this application, it is also possible to detect the terrain environment around the vehicle in real time, and adaptively expand or retract the accessories carried by the vehicle according to the width of the terrain in front of the vehicle, so as to improve the ability of the vehicle to adapt to the environment .

Corresponding to the foregoing embodiment of a collision detection method, the present application also provides an embodiment of a collision detection device.

An embodiment of a collision detection device of the present application can be applied to unmanned vehicles. The device embodiments can be implemented by software, or by hardware or a combination of software and hardware. Taking software implementation as an example, as a logical device, it is formed by reading the corresponding computer program instructions in the non-volatile memory into the memory through the processor of the unmanned vehicle where it is located. From the hardware level, as shown in Figure 8, it is a hardware structure diagram of an unmanned vehicle where a collision detection device of the present application is located, except for the processor, memory, network interface, and non-volatile memory shown in Figure 8 In addition to the volatile memory, the unmanned vehicle where the device in the embodiment is located usually may also include other hardware according to the actual function of the unmanned vehicle, which will not be repeated here.

Please refer to FIG. 9. FIG. 9 is a block diagram of a collision detection device shown in the present application. The collision detection device can be applied to the aforementioned unmanned vehicle, including: a model acquisition module 902, used to obtain the body of the vehicle Model, used to obtain the body model of the self-vehicle, the vehicle body body model contained in the body model of the self-vehicle is a multi-circle model; the collision detection module 904 is used to perform collision detection according to the body model of the self-vehicle; wherein, in the In the case where the rectangular body of the unmanned vehicle is equally divided into n rectangles along the long side direction, the centers of the n circles contained in the multi-circle model are respectively the center points of the n rectangles, and each circle The radius of a shape is the distance from any center point to the apex of the rectangle where it is located, and n is a positive integer greater than 1.

Optionally, in the model acquisition module 902, when the unmanned vehicle has one or more accessories beyond the vehicle body, the vehicle body model also includes an accessory body model; the accessory body model includes A single-circle model corresponding to each accessory; the center of the single-circle model is located at the midpoint of the line connecting the two furthest points on the corresponding accessory, and the diameter is the length of the line.

Optionally, when the unmanned vehicle has one or more accessories beyond the vehicle body, it also includes: a terrain acquisition module 906, which is used to acquire the boundary of the road terrain within a preset distance in front of the vehicle; a distance calculation module 908. It is used to calculate the distance from the center of each accessory body model to the terrain boundary when all the accessories of the vehicle are deployed; if the distance is greater than the radius of the corresponding accessory body model, expand the corresponding Attachment; if the distance is not greater than the radius of the corresponding attachment body model, then put away the attachment corresponding to the attachment body model.

Optionally, the collision detection module 904 is specifically used for: a parameter acquisition unit, configured to acquire the driving parameters of the own vehicle and the driving parameters of the movable obstacle; the driving parameters of the own vehicle include the driving trajectory of the own vehicle and the driving speed of the own vehicle, and the The moving parameters of the movable obstacle include the moving trajectory of the moving obstacle and the moving speed of the moving obstacle; the collision determination unit, according to the driving parameters of the own vehicle and the driving parameters of the moving obstacle, if the prediction of the shape of the own vehicle If the driving area corresponding to the model overlaps with the driving area of the movable obstacle, it is determined that there is a risk of collision.

Optionally, the device further includes: a collision time calculation module 910, configured to calculate the collision time between the unmanned vehicle and the movable obstacle, and the collision time is the driving time corresponding to the ego vehicle shape model. The time when the area overlaps with the driving area of the movable obstacle; the speed planning module 912 is used to re-plan the vehicle speed according to the collision time to eliminate the collision risk.

For the implementation process of the functions and effects of each unit in the above device, please refer to the implementation process of the corresponding steps in the above method for details, and will not be repeated here.

As for the device embodiment, since it basically corresponds to the method embodiment, for related parts, please refer to the part description of the method embodiment. The device embodiments described above are only illustrative, and the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in One place, or it can be distributed to multiple network elements. Part or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this application. It can be understood and implemented by those skilled in the art without creative effort.

Embodiments of the subject matter and functional operations described in this specification can be implemented in digital electronic circuitry, tangibly embodied computer software or firmware, computer hardware including the structures disclosed in this specification and their structural equivalents, or in A combination of one or more of . Embodiments of the subject matter described in this specification can be implemented as one or more computer programs, that is, one or more of computer program instructions encoded on a tangible, non-transitory program carrier for execution by or to control the operation of data processing apparatus. Multiple modules. Alternatively or additionally, the program instructions may be encoded on an artificially generated propagated signal, such as a machine-generated electrical, optical or electromagnetic signal, which is generated to encode and transmit information to a suitable receiver device for transmission by the data The processing means executes. A computer storage medium may be a machine-readable storage device, a machine-readable storage substrate, a random or serial access memory device, or a combination of one or more of them.

The processes and logic flows described in this specification can be performed by one or more programmable computers executing one or more computer programs to perform corresponding functions by operating on input data and generating output. The processes and logic flows can also be performed by, and apparatus can also be implemented as, special purpose logic circuitry, such as an FPGA (Field Programmable Gate Array) or an ASIC (Application Specific Integrated Circuit).

Computers suitable for the execution of a computer program include, for example, general and/or special purpose microprocessors, or any other type of central processing unit. Generally, a central processing unit will receive instructions and data from a read only memory and/or a random access memory. The essential components of a computer include a central processing unit for implementing or executing instructions and one or more memory devices for storing instructions and data. Typically, a computer will also include, or be operatively coupled to, one or more mass storage devices for storing data, such as magnetic or magneto-optical disks, or optical disks, to receive data therefrom or to It transmits data, or both. However, a computer is not required to have such a device. In addition, a computer may be embedded in another device such as a mobile phone, a personal digital assistant (PDA), a mobile audio or video player, a game console, a Global Positioning System (GPS) receiver, or a device such as a Universal Serial Bus (USB) ) portable storage devices like flash drives, to name a few.

Computer-readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media, and memory devices, including, for example, semiconductor memory devices (such as EPROM, EEPROM, and flash memory devices), magnetic disks (such as internal hard disks or removable disks), magneto-optical disks, and CD ROM and DVD-ROM disks. The processor and memory can be supplemented by, or incorporated in, special purpose logic circuitry.

While this specification contains many specific implementation details, these should not be construed as limitations on the scope of any inventions or of what may be claimed, but rather as primarily describing features of particular embodiments of particular inventions. Certain features that are described in this specification in multiple embodiments can also be implemented in combination in a single embodiment. On the other hand, various features that are described in a single embodiment can also be implemented in multiple embodiments separately or in any suitable subcombination. Furthermore, although features may function in certain combinations as described above and even be initially so claimed, one or more features from a claimed combination may in some cases be removed from that combination and the claimed A protected combination can point to a subcombination or a variant of a subcombination.

Similarly, while operations are depicted in the figures in a particular order, this should not be construed as requiring that those operations be performed in the particular order shown, or sequentially, or that all illustrated operations be performed, to achieve the desired result. In some cases, multitasking and parallel processing may be advantageous. Furthermore, the separation of various system modules and components in the above-described embodiments should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems can often be integrated together in a single software product in, or packaged into multiple software products.

Thus, certain embodiments of the subject matter have been described. Other embodiments are within the scope of the following claims. In some cases, the actions recited in the claims can be performed in a different order and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some implementations, multitasking and parallel processing may be advantageous.

The above is only a preferred embodiment of the application, and is not intended to limit the application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the application should be included in the application. within the scope of protection.

Claims

A collision detection method applied to unmanned vehicles, the method comprising:

Obtaining the body model of the self-vehicle, the body body model contained in the body model of the self-vehicle is a multi-circle model;

Collision detection is performed according to the vehicle shape model; wherein, when the rectangular body of the unmanned vehicle is divided into n rectangles along the long side direction, the n circles contained in the multi-circle model The centers of the circles are the center points of the n rectangles, and the radius of each circle is the distance from any center point to the apex of the rectangle where it is located, and n is a positive integer greater than 1.
According to the method according to claim 1, when the unmanned vehicle has one or more accessories beyond the vehicle body, the body model of the self-vehicle also includes an accessory body model;

The accessory body model includes a single-circle model corresponding to each accessory; the center of the single-circle model is located at the midpoint of the line connecting the two furthest points on the corresponding accessory, and the diameter is the length of the line.
The method according to claim 2, further comprising:

Obtain the boundary of the road terrain within the preset distance in front of the ego vehicle;

Calculate the distance from the center of each accessory shape model to the terrain boundary when all the accessories of the vehicle are unfolded;

If the distance is greater than the radius of the corresponding accessory body model, then expand the accessory corresponding to the accessory body model;

If the distance is not greater than the radius of the corresponding accessory body model, the accessory corresponding to the accessory body model is put away.
The method according to claim 2, wherein said performing collision detection according to said self-vehicle body model comprises:

Obtain the driving parameters of the own vehicle and the driving parameters of the movable obstacle; the driving parameters of the own vehicle include the driving trajectory of the own vehicle and the driving speed of the own vehicle, and the driving parameters of the movable obstacle include the driving trajectory of the movable obstacle and the moving obstacle the speed of the object;

According to the driving parameters of the own vehicle and the driving parameters of the movable obstacle, if it is predicted that the driving area corresponding to the body model of the ego vehicle will overlap with the driving area of the movable obstacle, it is determined that there is a collision risk.
The method according to claim 4, characterized in that the method further comprises:

Calculating the collision time between the unmanned vehicle and the movable obstacle, the collision time being the time when the driving area corresponding to the body model of the ego vehicle overlaps with the driving area of the movable obstacle;

The vehicle speed is replanned according to the collision time to eliminate the collision risk.
The method according to claim 1, further comprising:

Obtain a variety of vehicle body models; the multiple vehicle body models include multiple circle models with different numbers of circles;

If the current road congestion under the driving trajectory of the self-vehicle is congestion, select the model with a large number of circles for collision detection;

If the current road congestion is smooth, select a model with a small number of circles for collision detection.
The method according to claim 1, wherein the multi-circle model is a five-circle model.
A collision detection device applied to unmanned vehicles, the device comprising:

A model acquisition module, used to acquire the body model of the self-vehicle, and the body body model contained in the body model of the self-vehicle is a multi-circle model;

A collision detection module, configured to perform collision detection according to the self-vehicle shape model; wherein, when the rectangular body of the unmanned vehicle is equally divided into n rectangles along the long side direction, the multi-circle model The centers of the n circles included are the center points of the n rectangles, and the radius of each circle is the distance from any center point to the apex of the rectangle where it is located, and n is a positive integer greater than 1.
A storage medium, on which computer program instructions are stored, and the computer program instructions can implement the method according to any one of claims 1 to 7 after being executed.
An electronic device, characterized in that it comprises:

memory for storing computer program instructions; and

a processor for executing computer program instructions,

Wherein, the method described in any one of claims 1 to 7 can be realized when the computer program instructions are executed by the processor.