Method, device and electronic equipment for realizing in-lane turning of unmanned vehicle

Abstract

The present disclosure provides a method and device for realizing in-lane U-turn of an unmanned vehicle and an electronic device. The method is applied to an autonomous vehicle or an unmanned vehicle, and includes: obtaining driving state information and current road section information of the unmanned vehicle, and determining whether to trigger an in-lane U-turn operation by using a U-turn decision condition; when it is determined to trigger the in-lane U-turn operation, determining whether the unmanned vehicle meets a road environment constraint, checking a road boundary constraint based on a position of the unmanned vehicle in a road coordinate system, and determining an initial scene, calculating a next pose state based on a displacement flag and a current pose state of the unmanned vehicle, determining whether to take a vehicle center point corresponding to the next pose state as a path key point, calculating all path key points, and taking a path composed of the path key points as a driving trajectory of the unmanned vehicle for realizing the in-lane U-turn. The present disclosure guarantees the success rate of the in-lane U-turn operation of the unmanned vehicle, improves the efficiency of the in-lane U-turn, and expands the application range of the unmanned vehicle.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of unmanned driving, and particularly relates to a method and device for realizing lane turning of an unmanned vehicle and an electronic device. BACKGROUND

[0002] An unmanned vehicle is a comprehensive system integrating functions such as environment perception, planning and decision, and multi-level auxiliary driving. The unmanned vehicle is also called an automatic driving vehicle or unmanned vehicle. In the driving process of the unmanned vehicle, turning is required in some scenarios, which is particularly common in the use scenario of a parking lot. Therefore, the unmanned vehicle needs to have the ability to turn in a narrow road, which is also called lane turning or narrow road turning.

[0003] At present, when the existing unmanned vehicle turns in a narrow road, certain prior information is required, such as a preset turning path or pose information when the turning is completed. The current pose, intermediate pose and target pose of the vehicle are determined in real time, so that the vehicle drives along the driving path corresponding to the pose. However, this lane turning method requires the unmanned vehicle to change lanes and adjust the pose multiple times to complete the turning, which reduces the efficiency of the lane turning of the unmanned vehicle. In addition, when the existing unmanned vehicle plans a turning path, how to plan a reasonable turning trajectory under certain constraints is not considered, and the judgment of the current road driving environment is also lacking. Instead of deciding whether the current scenario can execute the lane turning function first, and then executing the lane turning function, the system directly enters the lane turning operation, which leads to the failure of the lane turning and the trapping in a dead zone, resulting in poor lane turning effect. SUMMARY

[0004] Therefore, the embodiments of the present disclosure provide a method and device for realizing lane turning of an unmanned vehicle and an electronic device to solve the problems of low efficiency of lane turning, lack of decision on lane turning operation, easy failure of lane turning, and poor lane turning effect in the prior art.

[0005] In a first aspect, a method for implementing in-lane U-turn by an unmanned vehicle is provided. In the method, during automatic driving of the unmanned vehicle, driving state information and current road segment information of the unmanned vehicle are obtained, and a U-turn decision condition is used to determine whether to trigger an in-lane U-turn operation based on the driving state information and the current road segment information. When it is determined to trigger the in-lane U-turn operation, current road environment information is determined based on the current road segment information, and it is determined whether the unmanned vehicle meets a road environment constraint corresponding to the in-lane U-turn according to the current road environment information. When the unmanned vehicle meets the road environment constraint, a current position of the unmanned vehicle is converted into a road coordinate system, and a road boundary constraint is checked based on the position of the unmanned vehicle in the road coordinate system. When the unmanned vehicle meets the road boundary constraint, an initial scene is determined, and a displacement flag is determined according to a determination result of the initial scene. A next pose state of the unmanned vehicle is calculated based on the displacement flag and a current pose state of the unmanned vehicle. An angle point position of the unmanned vehicle is determined according to the next pose state. A pose of the unmanned vehicle is safety detected based on the angle point position. According to a safety detection result, it is determined whether to take a vehicle center point corresponding to the next pose as a path key point or to take a vehicle center point corresponding to a previous pose as the path key point. All path key points are calculated. A path composed of the all path key points is taken as a driving trajectory of the unmanned vehicle for implementing the in-lane U-turn.

[0006] In a second aspect, an apparatus for implementing in-lane U-turn by an unmanned vehicle is provided. In the apparatus, an obtaining module is configured to obtain driving state information and current road segment information of the unmanned vehicle during automatic driving of the unmanned vehicle, and to determine whether to trigger an in-lane U-turn operation based on the driving state information and the current road segment information by using a U-turn decision condition. A determining module is configured to determine current road environment information based on the current road segment information when it is determined to trigger the in-lane U-turn operation, and to determine whether the unmanned vehicle meets a road environment constraint corresponding to the in-lane U-turn according to the current road environment information. A calculating module is configured to convert a current position of the unmanned vehicle into a road coordinate system when the unmanned vehicle meets the road environment constraint, and to check a road boundary constraint based on the position of the unmanned vehicle in the road coordinate system. When the unmanned vehicle meets the road boundary constraint, the calculating module is configured to determine an initial scene, and to determine a displacement flag according to a determination result of the initial scene. The calculating module is configured to calculate a next pose state of the unmanned vehicle based on the displacement flag and a current pose state of the unmanned vehicle. A planning module is configured to determine an angle point position of the unmanned vehicle according to the next pose state, to safety detect a pose of the unmanned vehicle based on the angle point position, and to calculate all path key points according to a safety detection result. The planning module is configured to determine whether to take a vehicle center point corresponding to the next pose as a path key point or to take a vehicle center point corresponding to a previous pose as the path key point, and to take a path composed of the all path key points as a driving trajectory of the unmanned vehicle for implementing the in-lane U-turn.

[0007] In a third aspect, the present disclosure provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the above method when executing the program.

[0008] The above at least one technical solution adopted by the embodiments of the present disclosure can achieve the following beneficial effects:

[0009] By obtaining the driving state information and the current section information of the unmanned vehicle in the automatic driving process of the unmanned vehicle, determining whether to trigger the lane turning operation based on the turning decision condition based on the driving state information and the current section information, when it is determined to trigger the lane turning operation, determining the current road environment information based on the current section information, determining whether the unmanned vehicle meets the road environment constraint corresponding to the lane turning according to the current road environment information, when the unmanned vehicle meets the road environment constraint, converting the current position of the unmanned vehicle into a road coordinate system, and checking the road boundary constraint based on the position of the unmanned vehicle in the road coordinate system, when the unmanned vehicle meets the road boundary constraint, judging the initial scene, and determining the displacement flag based on the judgment result of the initial scene, calculating the next pose state of the unmanned vehicle based on the displacement flag and the current pose state of the unmanned vehicle, determining the corner point position of the unmanned vehicle according to the next pose state, performing safety detection on the pose of the unmanned vehicle based on the corner point position, determining whether to take the vehicle center point corresponding to the next pose as the path key point or the vehicle center point corresponding to the previous pose as the path key point according to the safety detection result, calculating all path key points, and taking the path composed of all path key points as the driving trajectory of the unmanned vehicle for lane turning. The present disclosure can realize the lane turning operation after decision-making, and fully consider various constraints when planning the lane turning path, thereby ensuring the success rate of the lane turning operation of the unmanned vehicle, improving the efficiency of the lane turning, and expanding the application scenarios and range of the unmanned vehicle. BRIEF DESCRIPTION OF DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present disclosure, and other drawings can be obtained by those skilled in the art without creative labor.

[0011] Figure 1 is a structural schematic diagram of an automatic driving system provided by the embodiments of the present disclosure;

[0012] Figure 2 is a flowchart of a method for realizing lane turning of an unmanned vehicle provided by the embodiments of the present disclosure;

[0013] Figure 3is a structural schematic diagram of a lane U-turn decision maker provided by an embodiment of the present disclosure;

[0014] Figure 4 is an algorithm flow schematic diagram of a Replan state in a lower state machine provided by an embodiment of the present disclosure;

[0015] Figure 5 is a schematic diagram of an initial scene in a Replan state provided by an embodiment of the present disclosure;

[0016] Figure 6 is a flow schematic diagram of a lane U-turn decision process provided by an embodiment of the present disclosure;

[0017] Figure 7 is a structural schematic diagram of an apparatus for realizing lane U-turn by an unmanned vehicle provided by an embodiment of the present disclosure;

[0018] Figure 8 is a structural schematic diagram of an electronic device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0019] In the following description, specific details are set forth in order to provide a thorough understanding of the embodiments of the present disclosure. However, persons skilled in the art will understand that the present disclosure can be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present disclosure with unnecessary detail.

[0020] As described above, an unmanned vehicle, also known as an autonomous vehicle, an unmanned vehicle, or a wheeled mobile robot, is an integrated and intelligent new era technology product that integrates environment perception, path planning, state recognition, and vehicle control. With the rapid development of unmanned driving technology, the application scenarios and scope of autonomous vehicles have gradually expanded, including unmanned delivery vehicles, unmanned retail vehicles, unmanned cleaning vehicles, unmanned patrol vehicles, and the like. During road travel, an unmanned vehicle needs to perform U-turn driving in some scenarios, such as when the road ahead is congested and cannot be passed, or when the road ahead is a dead-end road. The situation of U-turn driving by an unmanned vehicle is particularly common in parking lot scenarios. Therefore, an unmanned vehicle needs to have the ability to U-turn in a narrow road, which is also referred to as lane U-turn or narrow road U-turn.

[0021] It should be noted that the UGV turning around driving is divided into two different situations, i.e., lane-out turning around and lane-in turning around, the difference between the two situations is that the lane-out turning around refers to that in the open road, the UGV often needs to cross the lane to turn around, and after turning around, the UGV enters the reverse lane to drive, so the space for crossing the lane to turn around is relatively wide; while the lane-in turning around is usually relatively narrow, so the planning and control requirements for the UGV are relatively high when the UGV performs the lane-in turning around. It is particularly emphasized that in the embodiments of the present disclosure, the lane-out turning around and the lane-in turning around can also be respectively referred to as "lane-out turning around" and "lane-in turning around", and "turning around" and "turning around" have the same meaning in the embodiments of the present disclosure, and are both used to refer to the UGV turning to the opposite direction on the driving road.

[0022] The following will describe in detail the method for realizing vehicle turning around and the problems existing in the prior art by combining the technical solutions in two prior arts, which can specifically include the following contents:

[0023] The first prior art (publication number: CN112660147A) discloses a method, device and equipment for controlling vehicle turning around and a storage medium, which iteratively performs the following operations at least once until the proximity of the current pose of the vehicle and the target pose of the vehicle when the vehicle completes the turning around exceeds a threshold value: determining an intermediate pose to be reached by the vehicle based at least on the current pose of the vehicle and the target pose; causing the vehicle to drive along a driving path from the current position to the intermediate position; in response to determining that a condition for switching the driving path is met, causing the vehicle to stop; and updating the current pose of the vehicle using the pose when the vehicle stops.

[0024] The second prior art (publication number: CN111891137A) discloses an automatic driving narrow road turning around method, system and vehicle, which first has a preset forward turning around path, mainly judges whether the current forward tracking path is feasible, if yes, the vehicle drives forward along the forward tracking path to realize turning around; otherwise, the vehicle stops and switches the driving direction when the vehicle drives forward along the forward tracking path to a predetermined position; a backward tracking path is generated in real time according to the current position of the vehicle and the backward reference path, so that the vehicle drives backward along the backward tracking path, and a forward tracking path is generated in real time according to the position of the vehicle.

[0025] Based on the content disclosed in the above-mentioned first prior art, the design theory of this technical solution is only applicable to turning around on a relatively wide lane, since it theoretically needs to adjust the pose three times to turn around, if the road section is too narrow, it will lead to failure in the middle of turning around and fall into a dead zone, which needs remote manual intervention; secondly, the technical solution needs to know the vehicle pose (i.e., the target pose) after turning around, which puts high requirements on the decision module, and the technical solution does not mention how to accurately calculate the pose after turning around.

[0026] Based on the above-mentioned content disclosed in the second prior art solution, the technical solution needs to obtain a preset U-turn path, and adjusts the driving direction of the vehicle through the forward tracking path and the backward tracking path, so as to adjust the attitude of the vehicle.

[0027] It can be seen that, whether it is the above-mentioned first prior art solution or the second prior art solution, when the vehicle U-turns, preset prior information such as a preset U-turn path or pose information when the U-turn is completed is needed; and the prior art solution only considers the problem from the perspective of automatic driving planning, without combining vehicle control (i.e., driving according to the planned path, which often has some deviations when executed), so there is a high probability that the U-turn trajectory is generated, and the final U-turn process fails due to the error of the vehicle control following. In addition, the prior art solution does not consider the constraints of road width and road boundary.

[0028] In view of the problems existing in the above-mentioned prior art solutions, it is urgent to provide a method for realizing in-lane U-turn of an unmanned vehicle. The method provided in the embodiments of the present disclosure at least considers the following technical problems:

[0029] How does an automatic driving vehicle judge that the current section cannot continue to drive during driving, make a decision of in-lane U-turn, and reduce the probability of making a wrong decision?

[0030] How does the automatic driving system execute the decision of in-lane U-turn, that is, how to plan the driving trajectory of the U-turn process under the consideration of the constraints of road boundary and vehicle control error?

[0031] How does the automatic driving system decide to exit the decision of in-lane U-turn, and how does the automatic driving system handle after judging that it cannot cope with the current scene?

[0032] The functional modules of the automatic driving system involved in the actual scene of the embodiments of the present disclosure will be described below in conjunction with the accompanying drawings. Figure 1 is a structural schematic diagram of the automatic driving system provided by the embodiments of the present disclosure. As Figure 1 shown, the automatic driving system mainly includes the following contents:

[0033] The automatic driving system comprises a sensor 101, a perception module 102, a positioning module 103, a high-definition map module 104, a decision planning module 105, a control module 106, an unmanned vehicle 107, and a remote control center 108. The sensor 101 comprises a laser radar, a camera and other data acquisition devices installed on the unmanned vehicle. The perception module 102 analyzes the road environment, obstacles and driving scenes around the vehicle according to the data collected by the sensor 101. The high-definition map module 104 stores high-precision map information about the road. The decision planning module 105 comprises a decision module and a motion planning module. The decision planning module 105 is used to make decisions on the driving state of the unmanned vehicle and plan a driving reference line. The lane turning function is realized by the decision planning module 105 and then specifically executed by the control module 106. The unmanned vehicle 107 comprises a vehicle bottom system, such as a vehicle chassis and a driver.

[0034] Further, in the embodiment of the present disclosure, the decision module in the decision planning module 105 is mainly used to make decisions on the current driving state and the corresponding driving reference line, such as the cruise function, the lane changing function, the reverse driving function, the cross-lane turning function and the like. The decision module comprises two sub-modules of decision and behavior planning. The lane turning function is mainly completed by the two modules of decision and behavior planning. The decision module is mainly used to determine whether the lane turning is needed at present. The behavior planning module is used to determine whether the current scene meets the conditions for the lane turning and plan a driving guide line (i.e., a driving reference line) for the lane turning. The driving guide line is updated in real time during the lane turning, and the vehicle turn signal is controlled during the lane turning.

[0035] Further, the motion planning module in the decision planning module 105 plans a smooth driving trajectory that meets various vehicle dynamics constraints based on the driving reference line, considers obstacles, vehicle constraints and road environment constraints, and then the control module 106 executes the driving trajectory.

[0036] The embodiment of the present disclosure focuses on the lane turning function in the decision module, specifically introduces how to make decisions on the lane turning, how to plan a driving reference line for the lane turning, and how to maintain the relationship between the lane turning state and other decision states. The technical solution of the present disclosure will be described in detail in combination with the drawings and specific embodiments.

[0037] Figure 2 is a flowchart of a method for realizing lane turning of an unmanned vehicle provided by the embodiment of the present disclosure. Figure 2 The method for realizing lane turning of the unmanned vehicle can be executed by the decision planning module in the automatic driving system. As shown in Figure 2As shown, the method for realizing the U-turn of the unmanned vehicle in the lane can specifically include:

[0038] S201, in the automatic driving process of the unmanned vehicle, obtaining the driving state information and the current road section information of the unmanned vehicle, and determining whether to trigger the operation of the U-turn in the lane based on the driving state information and the current road section information by using the U-turn decision condition;

[0039] S202, when it is determined to trigger the operation of the U-turn in the lane, determining the current road environment information based on the current road section information, and determining whether the unmanned vehicle meets the road environment constraint corresponding to the U-turn in the lane according to the current road environment information;

[0040] S203, when the unmanned vehicle meets the road environment constraint, converting the current position of the unmanned vehicle into a road coordinate system, and checking the road boundary constraint based on the position of the unmanned vehicle in the road coordinate system, when the unmanned vehicle meets the road boundary constraint, judging the initial scene, and determining the displacement flag bit according to the judgment result of the initial scene, calculating the next pose state of the unmanned vehicle based on the displacement flag bit and the current pose state of the unmanned vehicle;

[0041] S204, determining the corner point position of the unmanned vehicle according to the next pose state, performing safety detection on the pose of the unmanned vehicle based on the corner point position, determining whether to take the vehicle center point corresponding to the next pose as the path key point or to take the vehicle center point corresponding to the previous pose as the path key point according to the safety detection result, calculating all path key points, and taking the path composed of all path key points as the driving trajectory of the unmanned vehicle to realize the U-turn in the lane.

[0042] Specifically, the U-turn function in the lane of the unmanned vehicle of the embodiment of the present disclosure includes a decision module (i.e., a decision maker module) and a behavior decision module (i.e., a reference line module), wherein the decision maker module is used to determine whether the current scene can trigger the U-turn function in the lane, and the reference line module is used to specifically plan the driving trajectory of the U-turn in the lane, and then the motion planning module and the control module execute the specific U-turn function in the lane.

[0043] Further, the initial scene of the embodiment of the present disclosure refers to the initial attitude of the unmanned vehicle when performing the U-turn operation in the lane and the corresponding scene of the road direction, different initial scenes can correspond to different initial attitudes and vehicle driving directions, each initial scene corresponds to a respective displacement flag bit, the displacement flag bit can be considered as the vehicle driving direction corresponding to each scene, such as left turn, right turn, forward, backward, etc., the initial scene is to convert the attitude of the unmanned vehicle and the road environment from the Cartesian coordinate system to the Frenet coordinate system, and obtain the scene image by simplification.

[0044] According to the technical scheme provided by the embodiment of the present disclosure, in the automatic driving process of the unmanned vehicle, the driving state information and the current road section information of the unmanned vehicle are acquired, and it is judged whether to trigger the lane turning operation by using the turning decision condition; when it is judged to trigger the lane turning operation, it is judged whether the unmanned vehicle meets the road environment constraint corresponding to the lane turning according to the current road environment information; when the road environment constraint is met, the current position of the unmanned vehicle is converted into the road coordinate system, and the road boundary constraint is checked; when the road boundary constraint is met, the displacement flag bit is determined according to the judgment result of the initial scene, and the next pose state of the unmanned vehicle is calculated based on the displacement flag bit and the current pose state of the unmanned vehicle; the corner point position of the unmanned vehicle is determined according to the next pose state, the pose of the unmanned vehicle is safety detected based on the corner point position, the path key points are updated according to the safety detection result, and all path key points are calculated; the path composed of all path key points is taken as the driving track of the unmanned vehicle to realize the lane turning. The present disclosure can realize the decision before the execution of the turning operation, and fully consider various constraints when planning the turning path, so as to ensure the success rate of the lane turning operation of the unmanned vehicle and improve the efficiency of the lane turning and the application scene and range of the unmanned vehicle.

[0045] In some embodiments, based on the driving state information and the current road section information, it is judged whether to trigger the lane turning operation by using the turning decision condition, including: determining the current pose information of the unmanned vehicle based on the driving state information, determining the lane information of the road where the unmanned vehicle is currently located based on the current pose information and the high-precision map information, judging whether the unmanned vehicle is currently in the turning section according to the lane information and the preset turning section, and triggering the lane turning operation of the unmanned vehicle in the lane when the unmanned vehicle is currently in the turning section and the current road environment meets the turning decision condition; or, judging the traffic state of the road in front of the unmanned vehicle based on the current road section information, when the traffic state of the road in front of the unmanned vehicle is not passable and the parking time of the unmanned vehicle exceeds the preset time, judging whether the unmanned vehicle can plan a path to reach the destination after turning, and triggering the lane turning operation of the unmanned vehicle in the lane when it is judged that the unmanned vehicle can plan a path to reach the destination after turning and the rear of the unmanned vehicle has a turning space.

[0046] Specifically, the embodiment of the present disclosure provides two ways to trigger the lane turning operation of the unmanned vehicle, the first way is to set the turning section in advance, and the lane turning operation is performed when the vehicle enters the turning section and the environmental constraint meets the triggering condition; the second way is to judge the real-time road environment information, and if it is judged that the road in front cannot be passed and a road to the destination can be found after turning, the lane turning operation is performed.

[0047] Further, in the first way of triggering the U-turn operation in the lane, a U-turn area (i.e., an area corresponding to the U-turn section) is first preset, and when the vehicle reaches the U-turn area, the vehicle slows down and stops, and after the road in front of and behind the vehicle meets the U-turn condition, the U-turn is triggered and executed; in the second way of triggering the U-turn operation in the lane, when the road in front of the vehicle is blocked by an obstacle or cannot be passed, the stopping time of the vehicle exceeds the preset time, the task path to the destination can still be planned after the task is re-planned after the U-turn, and the space behind the vehicle is available for the U-turn, the U-turn operation in the lane is triggered.

[0048] Further, the embodiments of the present disclosure do not need to preset the prior knowledge of the pose after the U-turn is completed, and by presetting the U-turn section, the section information of the lane / road corresponding to the preset U-turn section is set to id, start s, end s, etc. in the conventional opendrive map format; the preset section id, start s, and end s are used as the U-turn section, and when the vehicle enters the range, it is determined whether the U-turn action can be performed.

[0049] In some embodiments, based on the driving state information and the current section information, it is determined whether to trigger the U-turn operation in the lane by using the U-turn decision condition, including: determining the automatic driving state of the vehicle based on the driving state information, when the automatic driving state is in a preset state in which the U-turn operation cannot be triggered, the U-turn operation of the vehicle in the lane is not triggered; determining the section in which the vehicle currently locates based on the current section information, when the section is a preset section in which the U-turn operation cannot be triggered, the U-turn operation of the vehicle in the lane is not triggered; and determining whether the vehicle is in the U-turn operation state based on the current operation state of the vehicle, when the vehicle is in the U-turn operation state, the U-turn operation of the vehicle in the lane is not triggered.

[0050] Specifically, when it is determined whether the current state of the vehicle meets the condition of the U-turn operation in the lane (i.e., whether to trigger the U-turn operation in the lane), the following methods can be used for the determination: first, determining according to the state of the vehicle, when the vehicle is in a parking state, a pose adjustment state, or the like, the U-turn operation cannot be triggered; second, determining according to the section information, when the vehicle is in a highway section, a no-parking zone section, or the like, the U-turn operation cannot be triggered; and third, determining according to the current operation state of the vehicle, for example, when the vehicle itself is in the U-turn state in the lane, the operation is not triggered any more.

[0051] In some embodiments, determining whether the unmanned vehicle meets the road environment constraint corresponding to the in-lane U-turn according to the current road environment information comprises: when triggering the in-lane U-turn operation of the unmanned vehicle, acquiring the current road environment information corresponding to the unmanned vehicle by using the laser radar, judging the road environment constraint when the unmanned vehicle performs the in-lane U-turn based on the front road information, the rear road information, and the obstacle information in the current road environment information, and adjusting the state of the unmanned vehicle to the operation state of the in-lane U-turn according to the judgment result, so as to enable the unmanned vehicle to start the in-lane U-turn function.

[0052] Specifically, after determining the operation of triggering the in-lane U-turn, it is determined whether to start the in-lane U-turn function based on the upper state machine in the reference line module. In actual application, the reference line module is mainly used to generate a specific U-turn reference line path, real-time update of the reference line, and jump between states, for example, internal state jump from Doing to Finish, and external state jump from in-lane U-turn to cruise state; the Behavior module (i.e., behavior decision module) is internally composed of two state machines to complete the in-lane U-turn function. The state between the decision maker module and the reference line module is maintained by the in-lane U-turn decision maker, which is used for state judgment and conversion of the in-lane U-turn function. The specific content of the in-lane U-turn decision maker will be described below in combination with the accompanying drawings and specific embodiments, Figure 3 is a structural schematic diagram of the in-lane U-turn decision maker provided by the embodiments of the present disclosure. As Figure 3 indicated, the in-lane U-turn decision maker mainly includes the following contents:

[0053] The internal state of the in-lane U-turn decision maker includes Init, Doing, and Finish, which correspond to the initialization state, the execution U-turn state, and the complete U-turn state, respectively; in the Init state, the decision of whether the current U-turn function can be triggered is executed; in the Doing state, it is indicated that the unmanned vehicle is currently in the U-turn state, at this time the U-turn reference line module works to plan a specific U-turn path; in the Finish state, it is indicated that the current U-turn operation is completed, and then the state is converted to the Init state.

[0054] Further, the upper state machine includes three states, namely, an INIT state, a DOING state, and a FINISH state; the INIT state is used to determine whether the current road environment meets the lane turning operation, if the current time does not meet the requirement, the state remains in the INIT state, if the lane turning constraint is met, the vehicle position coordinates are converted to the road coordinate system (frenet coordinate system), and then the DOING state is entered; the reason for performing the coordinate system conversion is to project the curve onto the straight road, so that the turning path is planned and then projected back, which can reduce the complexity of turning on the curve. The DOING state includes a lower state machine, which is mainly used to plan the real-time guide line of the turning operation and control the turn signal of the vehicle during the turning process; if the turning operation is finally completed, the FINISH state is entered, if the turning operation fails, the INIT state is returned; the FINISH state represents that the turning operation is completed, if there is a new turning task, the INIT state is entered, and a new round of lane turning operation is started.

[0055] In some embodiments, the current position of the unmanned vehicle is converted to the road coordinate system, and the road boundary constraint is checked based on the position of the unmanned vehicle in the road coordinate system, including: converting the current position corresponding to the unmanned vehicle and the current road segment information from the Cartesian coordinate system to the road coordinate system, and based on the position of the unmanned vehicle in the road coordinate system, calculating the road width corresponding to the current road segment, and determining whether the road width meets the width requirement of the lane turning operation, so as to check the road boundary constraint; wherein the width requirement is the road width determined based on the vehicle body length of the unmanned vehicle and the automatic driving longitudinal control accuracy, and the road coordinate system adopts the Frenet coordinate system.

[0056] Specifically, the checking operation of the road boundary constraint is completed in the lower state machine, which is a subdivision state of the DOING state of the upper state machine, and mainly includes the following six sub-states: a Replan state, a Start state, an InnerPoints state, a FinalPoint state, a Finish state, and a Failed state. The lower state machine enters from the Replan state and ends in the Finish state or the Failed state; after triggering the lane turning operation, the upper state machine enters the DOING state from the INIT state, and first cuts in from the Replan state. The function of the behavior decision module in the embodiment of the disclosure mainly calls the internal algorithm of the Replan state.

[0057] In some embodiments, the initial scene is judged, and a displacement flag bit is determined according to a judgment result of the initial scene; a next pose state of the unmanned vehicle is calculated based on the displacement flag bit and a current pose state of the unmanned vehicle, including: in a road coordinate system, judging the initial scene of the unmanned vehicle based on current pose information and current road segment information of the unmanned vehicle, and obtaining a displacement flag bit corresponding to the initial scene; taking a coordinate corresponding to the current pose state as a starting point, calculating a coordinate position of the next pose state in the road coordinate system under a preset search precision according to a direction in the displacement flag bit at a fixed curvature, and determining the next pose state of the unmanned vehicle based on the coordinate position.

[0058] Specifically, the planning of the U-turn path is implemented in the Replan state algorithm in the behavior decision module. The internal algorithm of the Replan state in the lower state machine is described in detail below in combination with the drawings and specific embodiments. Figure 4 is a schematic diagram of the algorithm of the Replan state in the lower state machine provided by the embodiments of the present disclosure. As shown in Figure 4 , the algorithm of the Replan state mainly includes the following contents:

[0059] 1) Road constraint check

[0060] The position of the unmanned vehicle and the road information are projected from the Cartesian coordinate system to the Frenet coordinate system, which not only facilitates subsequent calculation, but also reduces the difficulty of processing curved roads; by calculating the minimum road width of the current road segment where the unmanned vehicle is located, it is judged whether the road width meets the width requirement of the in-lane U-turn function. In actual application, the width requirement of the in-lane U-turn function is the vehicle length + automatic driving longitudinal control accuracy x 2; 2 is the minimum value, and the larger the value, the wider the road required; for example: the vehicle length of the unmanned vehicle is 4.6 m, and the automatic driving longitudinal control accuracy is 0.5 m, so the road width must be at least 5.6 m to perform in-lane U-turn;

[0061] 2) Initial scene judgment

[0062] Figure 5 is a schematic diagram of the initial scene under the Replan state provided by the embodiments of the present disclosure. As shown in Figure 5 , there are 16 possible cases of the initial scene, and the arrows represent the driving direction of the vehicle at the end of the final U-turn or the initial road direction. The opposite direction of the arrow can represent the driving direction of the vehicle at the end of the final U-turn; each scene corresponds to the flag bits of left turn, right turn, forward and backward; for example Figure 5 , the first initial scene 1F-L in corresponds to the flag bits representing forward and left turn; in actual application, these initial scenes are simplified by converting from the Cartesian coordinate system to the Frenet coordinate system, so the road directions are all horizontal;

[0063] 3) Search with given displacement flag

[0064] In the search of the next pose state, a fixed curvature can be used, and the position coordinates (x, y, theta) corresponding to the current pose can be calculated according to the left / right, front / back flags, and the preset search accuracy, to obtain the (x', y', theta') corresponding to the next pose state;

[0065] 4) Safety detection

[0066] According to the new pose (i.e. the next pose state), the position coordinates of the four corners of the vehicle are calculated, and then according to the position of the vehicle driving road in the Frenet coordinate system, it is judged whether the shape of the vehicle is all within the road;

[0067] 5) Give up the last search result and update the key point

[0068] If the vehicle center point of the current pose fails the safety detection, it means that this pose needs to be given up; at this time, the vehicle center point corresponding to the last pose is recorded as the key point, and the left / right, front / back state of the next step is adjusted according to the key point, for example, the left / right, front / back flags are opposite to those of the last pose;

[0069] 6) Dead loop detection

[0070] From the historical key points and the point set between the key points, it can be judged whether it is trapped in a dead loop, for example, if the same key point appears, it is considered to be trapped in a dead loop; in addition, it can also be judged whether the distance between each key point is greater than 2 times the automatic driving longitudinal control accuracy, if the distance between the key points does not meet the condition, it is considered to be trapped in a dead loop;

[0071] 7) U-turn completion

[0072] This step is mainly used to judge whether the vehicle is in the state of the third initial scene or the fourth initial scene in Figure 5 ;

[0073] 8) Coordinate conversion and key point correction

[0074] The results planned in the Frenet coordinate system are converted to the Cartesian coordinate system;

[0075] 9) Effectiveness check 1

[0076] Check whether the key points and the point set between the key points are all within the road, and the total displacement of the point set between the key points is greater than 2 times the automatic driving longitudinal control accuracy;

[0077] 10) Trajectory smoothing processing

[0078] Since the aforementioned operation is a turning trajectory calculated by fixed curvature, if the lateral tracking accuracy of the vehicle is to be improved, the trajectory needs to be smoothed in curvature; for example, a conventional spline curve, a clothoid curve, etc. are used for trajectory smoothing between key points;

[0079] 11) Effectiveness check 2

[0080] After the trajectory smoothing, the key points and the point set between the key points are checked again to see if they are all within the road, and the total displacement of the point set between the key points is greater than 2 times the longitudinal control accuracy of the autonomous vehicle.

[0081] In some embodiments, the corner point position of the unmanned vehicle is determined according to the next pose state, the pose of the unmanned vehicle is safety detected based on the corner point position, and according to the safety detection result, it is judged whether to take the vehicle center point corresponding to the next pose as the path key point or take the vehicle center point corresponding to the previous pose as the path key point, which comprises: determining the corner point position of the unmanned vehicle in the next pose state based on the coordinate position in the road coordinate system corresponding to the next pose state, judging whether the unmanned vehicle is in the lane based on the corner point position, so as to safety detect the pose according to the judgment result; when it is judged that the unmanned vehicle is in the lane, the vehicle center point corresponding to the next pose is taken as the path key point, and when it is judged that the unmanned vehicle is not in the lane, the next pose state is discarded, and the vehicle center point corresponding to the previous pose is taken as the path key point; the path is updated according to the path key point until the turning path is planned.

[0082] Specifically, when safety detecting the new pose (i.e. the searched next pose), the positions of the four corner points of the vehicle are calculated according to the position of the new pose in the Frenet coordinate system, and then it is judged whether the vehicle shape is all within the current lane according to the position information of the road where the unmanned vehicle is currently located. The functions of other sub-states in the lower state machine are briefly described below, which can specifically include the following contents:

[0083] In the Start state, the start point and the end point are extended, so that the motion planning module can better consider the obstacle constraint when planning the path; the first start point in the trajectory group is taken out, and the end point is extended, and it is judged whether it is the last one. If it is, enter the Final Point state.

[0084] In the Final Point state, it is used to judge whether the current segment in the path is completed, and when it is completed, it enters the Finish state, and when it is not completed, it is judged whether the parking time is too long, and when the parking time is too long, it reenters the Replan state for new path planning, otherwise it judges whether the current position deviates from the previous segment too much, and if not, it issues the current segment.

[0085] In the Inner Points state, whether the current segment is completed is judged, and the judgment basis is whether the current vehicle pose is within the longitudinal error range, the current position deviates too much from the current segment, and the judgment is made through the heading deviation, lateral distance error, and longitudinal distance error.

[0086] In some embodiments, after the path composed of all the path key points is taken as the driving trajectory of the unmanned vehicle for the U-turn in the lane, the method further includes: converting the driving trajectory from the road coordinate system to the Cartesian coordinate system, performing validity checking on the driving trajectory in the Cartesian coordinate system, performing smoothing processing on the driving trajectory, performing boundary collision detection on the driving trajectory after the smoothing processing, and determining whether to make the U-turn in the lane according to the result of the boundary collision detection; wherein the validity checking includes checking whether the point set between the path key points is in the lane, and whether the total displacement corresponding to the point set between the path key points is greater than the preset automatic driving longitudinal control accuracy.

[0087] Specifically, after the U-turn operation in the lane is completed, the planning module triggers task re-planning, generates the next segment of the task, and continues driving; if the behavior planning module triggers the U-turn failure state in the middle of the U-turn, the decision module initiates a task failure to the remote control center and requests takeover. In the same road segment, the U-turn can be triggered repeatedly only n times in a short time, and if the n U-turn decisions are not finally successful, the autonomous vehicle is parked for protection and no longer triggers the U-turn state, and a request for remote takeover is sent to the cloud; to avoid falling into a dead loop, n is a preset value.

[0088] The overall decision-making process of the U-turn in the lane will be described in detail below in combination with the accompanying drawings and specific embodiments, Figure 6 is a flowchart of the U-turn decision-making process provided by the embodiments of the present disclosure. As Figure 6 shown, the U-turn decision-making process mainly includes the following contents:

[0089] 1) Determine whether the current vehicle state meets the conditions for the U-turn in the lane, such as: according to the vehicle state, the U-turn cannot be triggered when the vehicle is in the parking state, the attitude adjustment state, etc.; according to the road segment information, the U-turn cannot be triggered when the vehicle is in the highway segment, the no-parking zone segment, etc.; the U-turn is not continued to be judged when the vehicle itself is in the U-turn state in the lane.

[0090] 2) Maintain information of the historical triggered road segment and the number of triggers, and determine whether the information needs to be cleared according to the current position; mainly used to prevent the U-turn from falling into a dead loop in a certain road segment; for example, the vehicle is always in a certain road segment and cannot successfully make a U-turn, and the historical number of triggers is recorded through this step, and then the remote control center cloud can be requested to intervene.

[0091] 3) Determine whether the current lane U-turn has triggered n times or more, which sets the maximum number of triggers for the same section to prevent the vehicle from triggering U-turns but failing to adjust the end, resulting in a dead loop.

[0092] 4) Determine whether the current section belongs to the preset U-turn section. Based on the current vehicle pose information, the lane information of the current road is calculated based on the high-precision map information, such as id, s, etc. Determine whether the current vehicle is in the preset U-turn section.

[0093] 5) Issue a stop command. If the vehicle has stopped in the preset U-turn section, first issue a stop command to the speed planning submodule in the motion planning module to control the vehicle to stop. If the vehicle has stopped, maintain the stop command and prepare to call the lane U-turn reference line module to calculate whether a U-turn path can be planned.

[0094] 6) Determine whether the time when the vehicle stops is greater than the preset time. By presetting the maximum parking time of the vehicle, the reason for the inability to pass through the front road is detected. The historical motion planning module contains the parking reason. If the vehicle cannot move for a long time, such as more than 5 minutes, it indicates that the road in front is not passable. After trying the U-turn function, it can be determined whether the task path to the destination can be re-planned.

[0095] 7) Call the reference line module to determine whether an effective lane U-turn trajectory can be generated. The path generation function of the reference line module is called. The reference line module will determine whether a reasonable U-turn path can be planned based on the road information of the current section, the control accuracy of the vehicle, and other constraints. After generating an initial trajectory, boundary collision detection is performed again in the Cartesian coordinate system. If the collision detection passes, it indicates that the U-turn can be performed. Otherwise, it indicates that the U-turn cannot be performed.

[0096] 8) Maintain a trigger counter. If the validity detection in the previous step passes, the count is 1, otherwise the count is 0. The trigger counter records the success and failure status (success is 1 and failure is 0) within a period of time, and then the average success probability, kurtosis, skewness, etc. information can be calculated.

[0097] 9) Determine whether the total number of trigger counter counts exceeds the threshold.

[0098] 10) Determine whether the probability and skewness calculated by the counter are greater than the respective thresholds. Calculate the trigger probability mean and skewness of the trigger counter. If both are satisfied, the U-turn is triggered. At the same time, maintain the historical information of the trigger, and the vehicle state machine jumps to maintain. Otherwise, end the current round of judgment.

[0099] According to the technical scheme provided by the embodiment of the present disclosure, the unmanned vehicle often needs to make a U-turn on a narrow road during driving due to the actual road or cloud task. Since the road width does not meet the constraint of the turning radius of the unmanned vehicle, the unmanned vehicle needs to change lanes multiple times to adjust the posture to complete the U-turn. The embodiment of the present disclosure first uses the decision maker module to determine whether the current scene can trigger the in-lane U-turn function. When it is determined that the U-turn can be made in the lane, the reference line module is used to specifically plan the in-lane U-turn trajectory. The road boundary constraint, vehicle control error constraint, and environmental constraint are fully considered when planning the U-turn path, so that the unmanned vehicle can complete the U-turn operation in the narrow lane. In addition, the in-lane U-turn function can make the unmanned vehicle be applied to a wider range of scenarios, expand the area where the unmanned vehicle can drive, and improve the operation efficiency, while reducing the frequency of intervention by the driver or the remote cloud. The vehicle does not need to detour from an additional road section to realize the U-turn, or needs remote intervention to realize the U-turn, thereby improving the operation efficiency of the unmanned vehicle.

[0100] The following is an apparatus embodiment of the present disclosure, which can be used to execute the method embodiments of the present disclosure. For details not disclosed in the apparatus embodiment of the present disclosure, please refer to the method embodiments of the present disclosure.

[0101] Figure 7 FIG. 1 is a structural schematic diagram of an apparatus for realizing in-lane U-turn of an unmanned vehicle provided by an embodiment of the present disclosure. As shown in FIG. 1, the apparatus for realizing in-lane U-turn of the unmanned vehicle includes: Figure 7

[0102] The acquisition module 701 is configured to acquire driving state information and current road section information of the unmanned vehicle during automatic driving of the unmanned vehicle, and determine whether to trigger the in-lane U-turn operation based on the driving state information and the current road section information by using a U-turn decision condition;

[0103] The judgment module 702 is configured to determine current road environment information based on the current road section information when it is determined to trigger the in-lane U-turn operation, and determine whether the unmanned vehicle meets the road environment constraint corresponding to the in-lane U-turn according to the current road environment information;

[0104] The calculation module 703 is configured to convert the current position of the unmanned vehicle into a road coordinate system when the unmanned vehicle meets the road environment constraint, check the road boundary constraint based on the position of the unmanned vehicle in the road coordinate system, determine an initial scene when the unmanned vehicle meets the road boundary constraint, determine a displacement flag based on the judgment result of the initial scene, and calculate a next pose state of the unmanned vehicle based on the displacement flag and the current pose state of the unmanned vehicle;

[0105] ​The planning module 704 is configured to determine the corner point position of the unmanned vehicle according to the next pose state, perform safety detection on the pose of the unmanned vehicle based on the corner point position, determine, according to the safety detection result, whether to take the vehicle center point corresponding to the next pose as a path key point or take the vehicle center point corresponding to the previous pose as the path key point, calculate all path key points, and take the path composed of all path key points as the driving track of the lane U-turn of the unmanned vehicle.

[0106] In some embodiments, Figure 7 The acquisition module 701 determines the current pose information of the unmanned vehicle based on the driving state information, determines the lane information of the road currently traveled by the unmanned vehicle based on the current pose information and the high-precision map information, determines whether the unmanned vehicle is currently in a U-turn road section according to the lane information and the preset U-turn road section, triggers the operation of the lane U-turn of the unmanned vehicle when the unmanned vehicle is currently in the U-turn road section and the current road environment meets the U-turn decision condition; or determines the traffic state of the road in front of the unmanned vehicle based on the current road section information, determines whether the unmanned vehicle can plan a path to the destination after the U-turn when the traffic state of the road in front of the unmanned vehicle is impassable and the parking time of the unmanned vehicle exceeds the preset time, and triggers the operation of the lane U-turn of the unmanned vehicle when the unmanned vehicle can plan a path to the destination after the U-turn and the rear of the unmanned vehicle has a U-turn space.

[0107] In some embodiments, Figure 7 The acquisition module 701 determines the automatic driving state of the unmanned vehicle based on the driving state information, does not trigger the operation of the lane U-turn of the unmanned vehicle when the automatic driving state is in a preset state in which the U-turn operation cannot be triggered, determines the road section currently traveled by the unmanned vehicle based on the current road section information, and does not trigger the operation of the lane U-turn of the unmanned vehicle when the road section is a preset road section in which the U-turn operation cannot be triggered, and determines whether the unmanned vehicle is in a U-turn operation state based on the current operation state of the unmanned vehicle, and does not trigger the operation of the lane U-turn of the unmanned vehicle when the unmanned vehicle is in the U-turn operation state.

[0108] In some embodiments, Figure 7 The judgment module 702 acquires the current road environment information corresponding to the unmanned vehicle by using the laser radar when triggering the operation of the lane U-turn of the unmanned vehicle, judges the road environment constraint when the unmanned vehicle performs the lane U-turn based on the front road information, the rear road information, and the obstacle information in the current road environment information, adjusts the state of the unmanned vehicle to the operation state of the lane U-turn according to the judgment result, and enables the lane U-turn function of the unmanned vehicle.

[0109] In some embodiments, Figure 7The calculation module 703 of the unmanned vehicle converts the current position of the unmanned vehicle and the current road section information from the Cartesian coordinate system to the road coordinate system, and calculates the road width corresponding to the current road section based on the position of the unmanned vehicle in the road coordinate system, to check the road boundary constraint, and determine whether the road width meets the width requirement of the turning operation in the lane.

[0110] In some embodiments, Figure 7 The calculation module 703 judges the initial scene of the unmanned vehicle in the road coordinate system based on the current pose information and the current road section information of the unmanned vehicle, and obtains the displacement flag corresponding to the initial scene. According to the fixed curvature, the coordinates corresponding to the current pose state are taken as the starting point, and the coordinate position of the next pose state in the road coordinate system is calculated under the preset search accuracy according to the direction in the displacement flag, and the next pose state of the unmanned vehicle is determined based on the coordinate position.

[0111] In some embodiments, Figure 7 The planning module 704 determines the corner point position of the unmanned vehicle in the next pose state based on the coordinate position in the road coordinate system corresponding to the next pose state, judges whether the unmanned vehicle is in the lane based on the corner point position, and performs safety detection on the pose according to the judgment result. When it is judged that the unmanned vehicle is in the lane, the vehicle center point corresponding to the next pose is taken as the path key point, and when it is judged that the unmanned vehicle is not in the lane, the next pose state is discarded, and the vehicle center point corresponding to the previous pose is taken as the path key point. The path is updated according to the path key point until the turning path is planned.

[0112] In some embodiments, Figure 7 After the planning module 704 takes the path composed of all path key points as the driving trajectory of the unmanned vehicle to realize the turning in the lane, the driving trajectory is converted from the road coordinate system to the Cartesian coordinate system, the effectiveness of the driving trajectory is checked in the Cartesian coordinate system, and the driving trajectory is smoothed. The driving trajectory after smoothing is subjected to boundary collision detection, and it is judged whether to turn in the lane according to the result of boundary collision detection. The effectiveness check includes checking whether the point set between the path key points is in the lane, and whether the total displacement corresponding to the point set between the path key points is greater than the preset automatic driving longitudinal control accuracy.

[0113] It should be understood that the size of the serial number of each step in the above embodiments does not mean the order of execution, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present disclosure.

[0114] Figure 8FIG. 1 is a structural schematic diagram of an electronic device according to an embodiment of the present disclosure. As shown in FIG. 1, the electronic device 8 according to the embodiment of the present disclosure includes a processor 801, a memory 802, and a computer program 803 stored in the memory 802 and capable of running on the processor 801. The processor 801 implements the steps in the above method embodiments when executing the computer program 803. Alternatively, the processor 801 implements the functions of the modules / units in the above device embodiments when executing the computer program 803. Figure 8

[0115] By way of example, the computer program 803 can be divided into one or more modules / units, which are stored in the memory 802 and executed by the processor 801 to complete the present disclosure. The one or more modules / units can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program 803 in the electronic device 8.

[0116] The electronic device 8 can be a desktop computer, a notebook computer, a palm computer, a cloud server, and the like. The electronic device 8 can include but is not limited to the processor 801 and the memory 802. Those skilled in the art can understand that the electronic device 8 can include more or fewer components, or combine certain components, or different components, for example, the electronic device can also include an input / output device, a network access device, a bus, and the like. Figure 8 The electronic device 8 is merely an example and does not constitute a limitation on the electronic device 8, which can include more or fewer components than those shown, or combine certain components, or different components, for example, the electronic device can also include an input / output device, a network access device, a bus, and the like.

[0117] The processor 801 can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic, discrete hardware components, or the like. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0118] ​The memory 802 can be an internal storage unit of the electronic device 8, for example, a hard disk or a memory of the electronic device 8. The memory 802 can also be an external storage device of the electronic device 8, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device 8. Further, the memory 802 can include both the internal storage unit and the external storage device of the electronic device 8. The memory 802 is used to store computer programs and other programs and data required by the electronic device. The memory 802 can also be used to temporarily store data that has been output or will be output.

[0119] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above-mentioned division of each functional unit and module is exemplified, and in actual application, the above-mentioned functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be realized in the form of hardware or software function unit. In addition, the specific name of each functional unit and module is only for easy distinction, and does not limit the protection scope of the present application. The specific working process of the unit and module in the above system can refer to the corresponding process in the foregoing method embodiments, which will not be described here.

[0120] In the above embodiments, the description of each embodiment has its own emphasis, and the parts not described or recorded in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0121] Those of ordinary skill in the art can appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether the functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. A person skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present disclosure.

[0122] In the embodiments of the present disclosure, it should be understood that the disclosed apparatus / computer device and method can be implemented in other manners. For example, the described apparatus / computer device embodiments are merely schematic. For example, the division of the modules or units is merely logical function division. There can be another division manner for the actual implementation, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between different units, can be indirect couplings or communication connections through some interfaces, devices or units, and can be electrical, mechanical or in other forms.

[0123] The units described as separated components can or can not be physically separated, and the components displayed as units can or can not be physical units, i.e., can be located in one place, or can be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purposes of the embodiments.

[0124] In addition, each functional unit in the various embodiments of the present disclosure can be integrated in one processing unit, or each unit can exist physically as separate units, or two or more units can be integrated in one unit. The integrated unit can be implemented in the form of hardware, or in the form of software functional units.

[0125] If the integrated module / unit is implemented in the form of software functional units and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on such understanding, the present disclosure implements all or part of the processes in the above-described embodiment methods, which can also be completed by computer programs instructing related hardware. The computer program can be stored in a computer readable storage medium, and when the processor executes the computer program, the steps of the above-described various method embodiments can be implemented. The computer program can include computer program code, which can be in the form of source code, object code, executable file or some intermediate form. The computer readable medium can include any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier wave signal, telecommunication signal and software distribution medium, etc. It should be noted that the computer readable medium can include appropriate contents according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to the legislation and patent practice, the computer readable medium does not include electric carrier wave signal and telecommunication signal.

[0126] The above examples are only used to illustrate the technical solutions of the present disclosure, rather than limit the same; although the present disclosure has been described in detail with reference to the foregoing examples, it should be understood by those of ordinary skill in the art that the technical solutions recorded in the foregoing examples can still be modified, or some technical features thereof can be replaced by equivalent replacements; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure, and should be included in the protection scope of the present disclosure.

Claims

1. A method for realizing in-lane U-turn by an unmanned vehicle, characterized in that, The application relates to an unmanned vehicle lane turning method and device. In the automatic driving process of the unmanned vehicle, driving state information and current road section information of the unmanned vehicle are acquired, and whether the operation of lane turning is triggered is determined by using a turning decision condition based on the driving state information and the current road section information. When it is determined that the operation of lane turning is triggered, current road environment information is determined based on the current road section information, and whether the unmanned vehicle meets the road environment constraint corresponding to the lane turning is determined according to the current road environment information. When the unmanned vehicle meets the road environment constraint, the current position of the unmanned vehicle is converted into a road coordinate system, and the road boundary constraint is checked based on the position of the unmanned vehicle in the road coordinate system. When the unmanned vehicle meets the road boundary constraint, an initial scene is judged, and a displacement flag is determined according to the judgment result of the initial scene.

2. The method of claim 1, wherein, The next pose state of the unmanned vehicle is calculated based on the displacement flag and the current pose state of the unmanned vehicle. The coordinate position of the unmanned vehicle in the next pose state in the road coordinate system is determined based on the coordinate position corresponding to the next pose state in the road coordinate system. Whether the unmanned vehicle is in the lane is determined based on the corner point position, so as to perform safety detection on the pose according to the judgment result.

3. The method of claim 2, wherein, When it is determined that the unmanned vehicle is in the lane, the vehicle center point corresponding to the next pose is taken as a path key point. When it is determined that the unmanned vehicle is not in the lane, the next pose state is discarded, and the vehicle center point corresponding to the previous pose is taken as a path key point. The path is updated according to the path key point until a turning path is planned, all path key points are calculated, and the path composed of all path key points is taken as the driving track of the unmanned vehicle for lane turning. The operation of lane turning is triggered when the unmanned vehicle is currently in the turning road section and the current road environment meets the turning decision condition. The passing state of the road in front of the unmanned vehicle is determined based on the current road section information. When the passing state of the road in front of the unmanned vehicle is impassable, and the parking time of the unmanned vehicle exceeds the preset time, whether the unmanned vehicle can plan a path to reach the destination after turning is determined. When it is determined that the unmanned vehicle can plan a path to reach the destination after turning, and the rear of the unmanned vehicle has a turning space, the operation of lane turning of the unmanned vehicle is triggered. The operation of lane turning is triggered when the unmanned vehicle is currently in the turning road section and the current road environment meets the turning decision condition. The passing state of the road in front of the unmanned vehicle is determined based on the current road section information. determining an automatic driving state of the unmanned vehicle based on the driving state information, and not triggering a lane-turning operation of the unmanned vehicle when the automatic driving state is in a preset state in which the lane-turning operation cannot be triggered; determining a current road section in which the unmanned vehicle is located based on the current road section information, and not triggering the lane-turning operation of the unmanned vehicle when the current road section is a preset road section in which the lane-turning operation cannot be triggered; determining whether the unmanned vehicle is in a lane-turning operation state based on a current operation state of the unmanned vehicle, and not triggering the lane-turning operation of the unmanned vehicle when the unmanned vehicle is in the lane-turning operation state.

4. The method of claim 1, wherein, The determining whether the unmanned vehicle meets the road environment constraint corresponding to the lane-turning operation based on the current road environment information includes: When triggering the lane-turning operation of the unmanned vehicle, acquiring the current road environment information corresponding to the unmanned vehicle by using a laser radar, determining a road environment constraint when the unmanned vehicle is lane-turning based on front road information, rear road information and obstacle information in the current road environment information, and adjusting a state of the unmanned vehicle to a lane-turning operation state according to a determination result, so as to enable the unmanned vehicle to start the lane-turning function.

5. The method of claim 1, wherein, The converting the current position of the unmanned vehicle into a road coordinate system and checking a road boundary constraint based on the position of the unmanned vehicle in the road coordinate system includes: converting the current position of the unmanned vehicle and the current road section information from a Cartesian coordinate system into a road coordinate system, and calculating a road width corresponding to the current road section based on the position of the unmanned vehicle in the road coordinate system, to check the road boundary constraint by determining whether the road width meets a width requirement of the lane-turning operation; wherein the width requirement is a road width determined based on a vehicle body length of the unmanned vehicle and an automatic driving longitudinal control accuracy, and the road coordinate system adopts a Frenet coordinate system.

6. The method of claim 5, wherein, The determining an initial scene, determining a displacement flag based on a determination result of the initial scene, calculating a next pose state of the unmanned vehicle based on the displacement flag and a current pose state of the unmanned vehicle, includes: determining the initial scene of the unmanned vehicle based on the current pose information of the unmanned vehicle and the current road section information in the road coordinate system, and obtaining a displacement flag corresponding to the initial scene; calculating a coordinate position of the next pose state in the road coordinate system according to a fixed curvature, a coordinate corresponding to the current pose state as a starting point, a direction in the displacement flag and a preset search accuracy, and determining the next pose state of the unmanned vehicle based on the coordinate position.

7. The method of claim 1, wherein, After the path composed of all the path key points is taken as the driving track of the unmanned vehicle for realizing the lane-turning operation, the method further includes: Convert the driving trajectory from the road coordinate system to a Cartesian coordinate system, perform validity checking on the driving trajectory in the Cartesian coordinate system, and perform smoothing processing on the driving trajectory, perform boundary collision detection on the smoothed driving trajectory, and determine whether to make a U-turn in the lane according to the result of the boundary collision detection. The validity checking includes checking whether the point set between the path key points is in the lane and whether the total displacement corresponding to the point set between the path key points is greater than a preset automatic driving longitudinal control accuracy.

8. An apparatus for realizing in-lane U-turn by an unmanned vehicle, characterized in that, The method comprises: An acquisition module configured to, in an automatic driving process of the unmanned vehicle, acquire driving state information and current section information of the unmanned vehicle, and determine whether to trigger a lane U-turn operation based on a U-turn decision condition and the driving state information and the current section information; A determination module configured to, when it is determined to trigger the lane U-turn operation, determine current road environment information based on the current section information, and determine whether the unmanned vehicle meets a road environment constraint corresponding to the lane U-turn according to the current road environment information; A calculation module configured to, when the unmanned vehicle meets the road environment constraint, convert a current position of the unmanned vehicle into a road coordinate system, and check a road boundary constraint based on the position of the unmanned vehicle in the road coordinate system, determine an initial scene when the unmanned vehicle meets the road boundary constraint, determine a displacement flag based on a result of the determination of the initial scene, calculate a next pose state of the unmanned vehicle based on the displacement flag and a current pose state of the unmanned vehicle, and the displacement flag includes a left turn / right turn direction and a forward / backward direction corresponding to each scene; A planning module configured to determine an angle point position of the unmanned vehicle in the next pose state based on a coordinate position in the road coordinate system corresponding to the next pose state, determine whether the unmanned vehicle is in the lane based on the angle point position, so as to perform safety detection on the pose according to a result of the determination, determine a vehicle center point corresponding to the next pose as a path key point when it is determined that the unmanned vehicle is in the lane, discard the next pose state and determine a vehicle center point corresponding to a previous pose as a path key point when it is determined that the unmanned vehicle is not in the lane, update a path based on the path key points, until a U-turn path is planned, all path key points are calculated, and the path composed of all path key points is taken as a driving trajectory of the unmanned vehicle for realizing a lane U-turn. 9.An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of any one of claims 1 to 7 when executing the program.

Images