A monitoring method and system for an autonomous driving system

Abstract

The application discloses a kind of monitoring method and system for automatic driving system, the method includes: by positioning monitoring module, the first position of vehicle and the second position compared to determine positioning alarm information;Determine the area proportion of the perception drivable area and the actual drivable area by the perception monitoring module, and determine the perception alarm information based on the detection of road line;Determine the failure number of trajectory update operation of automatic driving system by the planning monitoring module Determine planning alarm information;By control monitoring module, the error of actual control position and target control position is calculated, to determine control alarm information;By task monitoring module, the execution state information of each execution stage when automatic driving system handles target task is recorded, to determine task alarm information;By interactive monitoring module, the communication state information of front-end interactive interface and back end is detected, to determine interactive alarm information.The present scheme can realize all-around monitoring and problem positioning to automatic driving system.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to a monitoring method and system for an autonomous driving system. Background Art

[0002] Autonomous driving systems utilize multi-sensor technology and complex algorithms for environmental perception, path planning, and vehicle control, enabling automated vehicle control. Errors in these systems can directly lead to errors in vehicle control, or even failure. Therefore, monitoring these systems is crucial. However, existing technologies typically only monitor a single module or link in the autonomous driving system. This makes it difficult to conduct timely and effective fault diagnosis and recovery based on these monitoring results when an autonomous driving system anomalies occur. Summary of the Invention

[0003] The present invention provides a monitoring method and system for an autonomous driving system, which can realize all-round monitoring of the autonomous driving system and can quickly and accurately locate the abnormal link, thereby improving the safety and stability of the autonomous driving system.

[0004] In a first aspect, the present invention provides a monitoring method for an autonomous driving system, comprising:

[0005] A positioning monitoring module compares a first position and a second position of the target vehicle at a next moment to obtain a position deviation result, and determines a positioning alarm message based on the position deviation result; wherein the first position is a position of the target vehicle at the next moment predicted based on historical position information of the target vehicle; and the second position is an actual position of the target vehicle at the next moment obtained based on positioning by an autonomous driving system; and the autonomous driving system is configured on the target vehicle;

[0006] The perception monitoring module determines the area ratio between the perceived drivable area and the actual drivable area; if the area ratio exceeds a preset ratio threshold, the perception monitoring module detects the roadside to obtain a roadside detection result, and determines a perception alarm message based on the roadside detection result; wherein the perceived drivable area is the road area where the target vehicle is navigable as determined by the autonomous driving system; and the actual drivable area is the road area where the target vehicle is actually navigable corresponding to the perceived drivable area in the high-precision map;

[0007] Determining, by a planning monitoring module, the number of failed trajectory update operations of the autonomous driving system within a preset time period, and determining a planning alarm message based on a relationship between the number of failures and a preset threshold; the trajectory update operation is used to update the trajectory points of the target vehicle during driving;

[0008] The control monitoring module calculates the lateral and longitudinal errors between the actual control position and the target control position of the target vehicle, and determines a control alarm message based on the relationship between the lateral and longitudinal errors and the error threshold; wherein the actual control position is the position actually reached by the automatic driving system after controlling the target vehicle according to the vehicle control parameters; and the target control position is the position expected to be reached by the automatic driving system after controlling the target vehicle according to the vehicle control parameters;

[0009] Recording, by means of a task monitoring module, execution status information of the autonomous driving system at each execution stage when processing a target task, and determining task alarm information based on the execution status information;

[0010] The interaction monitoring module detects the communication status information between the front-end interaction interface and the back-end of the autonomous driving system, and determines the interaction alarm information based on the communication status information.

[0011] In a second aspect, the present invention further provides a monitoring system for an autonomous driving system, comprising:

[0012] a positioning monitoring module, configured to compare a first position and a second position of a target vehicle at a next moment to obtain a position deviation result, and to determine a positioning alarm message based on the position deviation result; wherein the first position is a position of the target vehicle at the next moment predicted based on historical position information of the target vehicle; and the second position is an actual position of the target vehicle at the next moment obtained based on positioning by an autonomous driving system, wherein the autonomous driving system is configured on the target vehicle;

[0013] A perception monitoring module is configured to determine a ratio between the perceived drivable area and the actual drivable area; if the ratio exceeds a preset ratio threshold, the module detects the roadside to obtain a roadside detection result, and determines a perception alarm message based on the roadside detection result; wherein the perceived drivable area is the road area perceived by the autonomous driving system as passable by the target vehicle; and the actual drivable area is the road area in the high-precision map corresponding to the perceived drivable area and actually passable by the target vehicle;

[0014] a planning monitoring module, configured to determine the number of failed trajectory update operations by the autonomous driving system within a preset time period, and to determine a planning alarm message based on a relationship between the number of failures and a preset threshold; the trajectory update operation is configured to update the trajectory points of the target vehicle during driving;

[0015] a control monitoring module, configured to calculate the lateral and longitudinal errors between the actual control position of the target vehicle and the target control position, and to determine a control alarm message based on the relationship between the lateral and longitudinal errors and the error thresholds; wherein the actual control position is the position actually reached by the automatic driving system after controlling the target vehicle according to the vehicle control parameters; and the target control position is the position expected to be reached by the automatic driving system after controlling the target vehicle according to the vehicle control parameters;

[0016] A task monitoring module is used to record execution status information of the autonomous driving system at each execution stage when processing the target task, and determine task alarm information based on the execution status information;

[0017] The interaction monitoring module is used to detect the communication status information between the front-end interactive interface and the back-end of the autonomous driving system, and determine the interaction alarm information based on the communication status information.

[0018] The technical solution provided by the present invention comprises: a positioning monitoring module comparing a first position and a second position of a target vehicle at a next moment to obtain a position deviation result, and determining a positioning alarm message based on the position deviation result; a perception monitoring module determining an area ratio between a perceived drivable area and an actual drivable area; if the area ratio exceeds a preset ratio threshold, the perception monitoring module detecting a roadside line to obtain a roadside line detection result, and determining a perception alarm message based on the roadside line detection result; a planning monitoring module determining the number of failed trajectory update operations by the autonomous driving system within a preset time period, and determining a planning alarm message based on the relationship between the number of failures and a preset number threshold; a control monitoring module calculating the lateral and longitudinal errors between the actual control position and the target control position of the target vehicle, and determining a control alarm message based on the relationship between the lateral and longitudinal errors and the error threshold; a task monitoring module recording execution status information at each execution stage when the autonomous driving system processes a target task, and determining a task alarm message based on the execution status information; and an interaction monitoring module detecting communication status information between a front-end interactive interface and a back-end of the autonomous driving system, and determining an interaction alarm message based on the communication status information. The solution of the present invention can monitor the positioning, perception, planning, control, tasks and interaction links of the autonomous driving system separately. Through multi-level, multi-dimensional and all-round monitoring, it can promptly detect anomalies in the autonomous driving system and quickly and accurately locate the link where the anomaly occurs, thereby improving the safety and stability of the autonomous driving system.

[0019] The above content of the invention is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are specifically listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The above and other features, advantages, and aspects of the various embodiments of the present invention will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and that the originals and elements are not necessarily drawn to scale.

[0021] Figure 1 This is a flow chart of a monitoring method for an autonomous driving system provided by an embodiment of the present invention;

[0022] Figure 2 This is a flow chart of another monitoring method for an autonomous driving system provided by an embodiment of the present invention;

[0023] Figure 3 Schematic diagram of a monitoring system for an autonomous driving system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0024] Embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0025] It should be understood that the various steps described in the method embodiments of the present invention can be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect. The terms "including" and their variations used herein are open inclusions, i.e., "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one other embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the description below.

[0026] It should be noted that the concepts of "first" and "second" mentioned in the present invention are only used to distinguish different modules or units, and are not used to limit the order or interdependence of the functions performed by these modules or units. The modifications of "one" and "multiple" mentioned in the present invention are illustrative and not restrictive. Those skilled in the art should understand that unless the context clearly indicates otherwise, it should be understood as "one or more". The names of the messages or information exchanged between multiple modules in the embodiments of the present invention are for illustrative purposes only and are not used to limit the scope of these messages or information.

[0027] Figure 1 This is a flow chart of a method for monitoring an autonomous driving system provided by an embodiment of the present invention. The embodiment of the present invention is applicable to monitoring an autonomous driving system configured on a vehicle. The method can be executed by a monitoring system for an autonomous driving system. The monitoring system for an autonomous driving system can be implemented in the form of software and / or hardware. The monitoring system for an autonomous driving system can be deployed on a vehicle and run based on an independent thread. Figure 1 As shown, the monitoring method for the automatic driving system according to the embodiment of the present invention may include the following process:

[0028] S101. Compare the first position and the second position of the target vehicle at the next moment through the positioning monitoring module to obtain a position deviation result, and determine the positioning alarm information based on the position deviation result; wherein the first position is the position of the target vehicle at the next moment predicted based on the historical position information of the target vehicle; the second position is the actual position of the target vehicle at the next moment based on the positioning of the automatic driving system; the automatic driving system is configured on the target vehicle.

[0029] The target vehicle is a vehicle that needs to be monitored by the autonomous driving system. The autonomous driving system is configured on the target vehicle and can be used to control the target vehicle to perform autonomous driving.

[0030] The target vehicle's historical location information is obtained by the autonomous driving system. This information specifically refers to the location information determined before the next moment. For example, this information can include both the current and previous locations. The vehicle's historical location information essentially reflects the vehicle's movement between moments.

[0031] The first position is the target vehicle's predicted position at the next moment based on its historical position information. In other words, the first position essentially reflects the position the target vehicle could theoretically reach at the next moment if it continued traveling at the speed corresponding to the current relevant historical position information. Correspondingly, the second position is the target vehicle's position at the next moment based on the positioning of the autonomous driving system. The position deviation can be calculated by subtracting the first and second positions.

[0032] Specifically, in actual scenarios, the positioning capability of the autonomous driving system is crucial for the target vehicle, but the positioning capability of the autonomous driving system may also have problems. Therefore, the positioning monitoring module can compare the first and second positions of the target vehicle at the next moment to obtain a position deviation result. Since the first position reflects the predicted position of the target vehicle at the next moment, and the second position is the position actually located by the autonomous driving system at the next moment, based on the comparison of the two, it is possible to determine whether the positioning capability of the autonomous driving system is accurate. The positioning alarm information is determined based on the position deviation result. The positioning alarm information is used to reflect the existence of positioning anomalies in the autonomous driving system.

[0033] S102. Determine the area ratio relationship between the perceived drivable area and the actual drivable area through the perception monitoring module; if the area ratio relationship exceeds a preset ratio threshold, detect the roadside through the perception monitoring module to obtain a roadside detection result, and determine the perception alarm information based on the roadside detection result; wherein, the perceived drivable area is the road area where the target vehicle is passable as determined by the automatic driving system; the actual drivable area is the road area where the target vehicle is actually passable that corresponds to the perceived drivable area in the high-precision map.

[0034] Among them, the autonomous driving system can rely on the integration of multiple sensors and technologies to perceive the environment around the target vehicle. The perceived drivable area is the road area that the autonomous driving system perceives and determines is passable by the target vehicle. Correspondingly, the actual drivable area is the road area in the HD map that is actually passable by the target vehicle and corresponds to the perceived drivable area. HD maps are high-precision maps that have been pre-configured in the autonomous driving system. HD maps can have an accuracy of up to centimeters and contain detailed lane line information, traffic sign information, road feature information, etc., thus providing the autonomous driving system with accurate and reliable map information. In other words, the perceived drivable area is essentially the drivable area determined by the autonomous driving system after perception, while the actual drivable area is the area that is actually drivable in the actual geographical environment. Roadside lines refer to various boundary lines on the road, such as lane lines.

[0035] In actual scenarios, the automatic driving system can perceive the road area in front of the target vehicle, so as to provide the target vehicle with the ability to understand the surrounding environment, and facilitate the automatic driving system to make relevant path planning and decision. Then, the perception ability of the automatic driving system can be detected by means of the perception monitoring module, so as to determine whether the perception ability of the automatic driving system is abnormal.

[0036] Specifically, the perception monitoring module can project the perception drivable area determined by the automatic driving system in the high-definition map, so as to determine the area proportion relationship between the perception drivable area and the actual drivable area. The obtained area proportion relationship can be compared with the preset proportion threshold. Generally, the preset proportion threshold is set as 1 by default. If the area proportion relationship exceeds the preset proportion threshold, that is, the area proportion of the perception drivable area to the actual drivable area is greater than 1, it indicates that the perception drivable area determined by the automatic driving system is larger than the actual drivable area, which means that the current perception ability of the automatic driving system is obviously problematic. Therefore, the perception monitoring module can further detect the road edge line. The detection of the road edge line refers to the detection of the road edge line in the automatic driving system perception result and the corresponding road edge line in the high-definition map, so as to determine the deviation between the two, thereby obtaining the road edge line detection result, and determining the perception alarm information based on the road edge line detection result. The perception alarm information is used to reflect the perception abnormality of the automatic driving system.

[0037] S103, determining, by the planning monitoring module, the number of failures of the trajectory updating operation of the automatic driving system in a preset time period, and determining the planning alarm information based on the relationship between the number of failures and the preset number threshold; the trajectory updating operation is used to update the trajectory point of the target vehicle in the driving process.

[0038] The preset time period can be set based on actual needs, for example, the preset time period can be set as 50 ms. The trajectory updating operation of the automatic driving system is to update the trajectory point of the target vehicle in the driving process. The trajectory point can indicate the position point that the target vehicle needs to reach at each time point in the driving process.

[0039] The preset number threshold is also set based on actual needs. Based on the preset number threshold, the range of the number of failures of the trajectory updating operation that can be accepted can be reflected. For example, the preset number threshold can be set as 3 times. If the number of failures is within this range, it may not cause problems to the driving of the vehicle, which is acceptable; if the number of failures exceeds this range, it will cause great problems to the driving of the vehicle, for example, it may cause the movement of the vehicle to stop, which is unacceptable.

[0040] Specifically, the automatic driving system can plan and update the driving trajectory of the target vehicle, that is, update the trajectory points of the target vehicle in the driving process. However, the automatic driving system may also fail to successfully update the trajectory points, that is, the trajectory updating operation fails. Therefore, the planning monitoring module can determine the number of times of failure of the trajectory updating operation of the automatic driving system within a preset time period, and determine the planning alarm information based on the relationship between the number of failures and the preset number threshold, so as to determine whether the automatic driving system has a serious problem in the trajectory updating link of the vehicle. The planning alarm information is used to reflect that the automatic driving system has a planning abnormality.

[0041] S104, calculating, by the control monitoring module, a lateral and longitudinal error between an actual control position and a target control position of the target vehicle, and determining a control alarm information based on a relationship between the lateral and longitudinal error and an error threshold; wherein the actual control position is a position actually reached by the automatic driving system after controlling the target vehicle according to the vehicle control parameter; and the target control position is a position expected to be reached by the automatic driving system after controlling the target vehicle according to the vehicle control parameter.

[0042] The vehicle control parameter can be used to adjust the motion behavior of the vehicle. The vehicle control parameter includes parameters such as speed, acceleration, angular velocity, angular acceleration, steering angle, and steering rate. The actual control position is a position actually reached by the automatic driving system after controlling the target vehicle according to the vehicle control parameter; and the target control position is a position expected to be reached by the automatic driving system after controlling the target vehicle according to the vehicle control parameter. It should be noted that the actual control position and the target control position are essentially two positions corresponding to the same vehicle control parameter after control, and the actual control position is the position that can actually be reached, and the target control position is the position expected to be reached.

[0043] The lateral and longitudinal error includes a lateral error and a longitudinal error. The lateral and longitudinal error reflects the difference between the actual control position and the target control position.

[0044] Specifically, after the automatic driving system determines the specific vehicle control parameter for controlling the target vehicle, the vehicle can be controlled based on these vehicle control parameters. However, there may be control errors in the process of controlling the vehicle, resulting in the vehicle failing to reach the expected position. Therefore, the control monitoring module can calculate the lateral and longitudinal error between the actual control position and the target control position of the target vehicle, so as to reflect the controllability of the target vehicle, and then determine the control alarm information based on the relationship between the lateral and longitudinal error and the error threshold; the control alarm information reflects that the automatic driving system has an abnormality in controlling the vehicle. For example, if the lateral and longitudinal error exceeds the error threshold, it indicates that the controllability of the vehicle is abnormal, and the control alarm information can be prompted for subsequent adjustment.

[0045] S105. Record, through the task monitoring module, execution status information of the autonomous driving system at each execution stage when processing the target task, and determine task alarm information based on the execution status information.

[0046] The target task is the task that the autonomous driving system is to process and advance. For example, the target task can be a perception task, a positioning task, a planning task, etc. The execution phase includes task start, task in progress, and task end. Execution status information includes normal execution and abnormal execution.

[0047] Specifically, when an autonomous driving system processes a target task, problems or anomalies may occur at various stages of the task. Therefore, a task monitoring module can record the execution status information of the autonomous driving system at each stage of the task and determine task alarm information based on this execution status information. Task alarm information indicates anomalies in the autonomous driving system's task execution.

[0048] Optionally, the mission alarm information may also include specific exception information. Based on the mission alarm information, the autonomous driving system can take appropriate countermeasures. For example, it can restart the mission or display the mission alarm information on the target vehicle's interface.

[0049] S106. Detect the communication status information between the front-end interactive interface and the back-end of the autonomous driving system through the interaction monitoring module, and determine the interaction alarm information based on the communication status information.

[0050] The front-end interactive interface not only displays various parameter and status information for users, but also supports different user operations. For example, the front-end interactive interface can display the task execution status, the health of the autonomous driving system, map information, vehicle location, vehicle chassis information, prompt information, and various alarm information to facilitate user understanding and control; users can also perform related operations on the front-end interface.

[0051] Communication status information refers to the communication status during the request and response process between the front-end interactive interface and the autonomous driving system back-end. For example, the front-end interactive interface and the autonomous driving system back-end can communicate via streaming.

[0052] Specifically, the interaction monitoring module can monitor the communication status between the front-end interactive interface and the back-end of the autonomous driving system and determine interaction alarm information based on this communication status information. This interaction alarm information is used to indicate interaction anomalies in the autonomous driving system. For example, if it is detected that the front-end and back-end cannot communicate normally within the SSE (Server-SentEvents) cycle, the autonomous driving system is considered to have an interaction anomaly.

[0053] The technical solution provided by the present invention comprises: a positioning monitoring module comparing a first position and a second position of a target vehicle at a next moment to obtain a position deviation result, and determining a positioning alarm message based on the position deviation result; a perception monitoring module determining an area ratio between a perceived drivable area and an actual drivable area; if the area ratio exceeds a preset ratio threshold, the perception monitoring module detecting a roadside line to obtain a roadside line detection result, and determining a perception alarm message based on the roadside line detection result; a planning monitoring module determining the number of failed trajectory update operations by the autonomous driving system within a preset time period, and determining a planning alarm message based on the relationship between the number of failures and a preset number threshold; a control monitoring module calculating the lateral and longitudinal errors between the actual control position and the target control position of the target vehicle, and determining a control alarm message based on the relationship between the lateral and longitudinal errors and the error threshold; a task monitoring module recording execution status information at each execution stage when the autonomous driving system processes a target task, and determining a task alarm message based on the execution status information; and an interaction monitoring module detecting communication status information between a front-end interactive interface and a back-end of the autonomous driving system, and determining an interaction alarm message based on the communication status information. The solution of the present invention can monitor the positioning, perception, planning, control, tasks and interaction links of the autonomous driving system separately. Through multi-level, multi-dimensional and all-round monitoring, it can promptly detect anomalies in the autonomous driving system and quickly and accurately locate the link where the anomaly occurs, thereby improving the safety and stability of the autonomous driving system.

[0054] Figure 2 This is a flow chart of another monitoring method for an autonomous driving system provided by an embodiment of the present invention. The technical solution of this embodiment further optimizes the process of determining positioning alarm information based on the position deviation result by the positioning monitoring module in the above embodiment. This embodiment can be combined with various optional solutions in one or more of the above embodiments. Figure 2 As shown, the monitoring method for the automatic driving system according to the embodiment of the present invention may include the following process:

[0055] S201. Compare the first position and the second position of the target vehicle at the next moment through the positioning monitoring module to obtain a position deviation result; wherein, the first position is the position of the target vehicle at the next moment predicted based on the historical position information of the target vehicle; the second position is the actual position of the target vehicle at the next moment based on the positioning of the automatic driving system; the automatic driving system is configured on the target vehicle.

[0056] As an optional but non-limiting implementation, the historical location information includes the location information of the target vehicle at the current moment and the location information at the previous moment; accordingly, the process of determining the first location includes the following steps A1-A2:

[0057] Step A1: Obtain the current position information and the previous position information of the target vehicle obtained by the automatic driving system through the positioning monitoring module.

[0058] Among them, the location information is the vehicle positioning data obtained by the automatic driving system.

[0059] Specifically, the positioning monitoring module can obtain the target vehicle's current and previous location information by monitoring the message data in the automatic driving system.

[0060] It should be noted that in an embodiment of the present invention, the various parameters or data determined by the autonomous driving system can be transmitted in the form of topic messages. Then, each monitoring module in the monitoring method can monitor these topic information to obtain the required parameter information or data information from the autonomous driving system. This will not be described in detail elsewhere in this embodiment.

[0061] Step A2: Input the current position information and the previous position information into the trajectory prediction model through the positioning monitoring module, and output the first position; wherein the trajectory prediction model can predict the position information that the target vehicle will reach at the next moment based on the target vehicle's current position information.

[0062] The trajectory prediction model can be obtained by training an LSTM (Long Short-Term Memory) neural network model as the initial network model. Specifically, during the training process, the historical location information of the target vehicle can be used as training data, and each historical location information is input into the neural network model for training. The following formula is used in the training process:

[0063] h t =f(W h ·[h t-1 ,x t ]+bh

[0064] wherein h t denotes the hidden state at time step t; x t denotes the input historical position information; W h denotes the weight; and b h denotes the bias term.

[0065] The model output y can be calculated by the following formula:

[0066]

[0067] wherein y denotes the predicted position; W0denotes the initial weight; and b 0h denotes the initial bias term.

[0068] After the training of the LSTM neural network model is completed, the trained neural network model can be used as a trajectory prediction model. Then, the position information at the current time and the position information at the previous time are input to the trajectory prediction model through the positioning monitoring module, and the first position can be output.

[0069] S202, if the position deviation result exceeds the preset deviation threshold, the point cloud data of the target vehicle at the current time and the point cloud data of the target vehicle at the next time are obtained through the positioning monitoring module.

[0070] wherein the point cloud data can be determined based on the laser radar configured on the target vehicle.

[0071] Specifically, if the position deviation result exceeds the preset deviation threshold, it indicates that the actual position of the vehicle positioned by the autonomous driving system and the predicted position are quite different at the next time, and the positioning ability of the autonomous driving system may be problematic. Therefore, it is necessary to continue to make more detailed and accurate judgments on the target vehicle. Specifically, the point cloud data of the target vehicle at the current time and the point cloud data of the target vehicle at the next time can be obtained through the positioning monitoring module, and the two frames of point cloud data essentially reflect the real motion of the target vehicle from the current time to the next time in the actual scene, and then the positioning ability of the autonomous navigation system can be further judged from the actual point cloud data.

[0072] S203, a point cloud matching algorithm is started to determine the corresponding matching point group in the point cloud data at the current time and the point cloud data at the next time; wherein the matching point group includes a first measured point and a second measured point, the first measured point belongs to the point cloud data at the current time, and the second measured point belongs to the point cloud data at the next time.

[0073] ​Among them, the matching point group essentially reflects the corresponding groups of corresponding points in two adjacent frames of point cloud data. The matching point group includes the first measured point and the second measured point. The first measured point belongs to the point cloud data at the current moment, and the second measured point belongs to the point cloud data at the next moment.

[0074] Specifically, after obtaining the point cloud data of the target vehicle at the current moment and the point cloud data at the next moment, the point cloud matching algorithm ICP (Iterative Closest Point) can be started to determine the corresponding matching point groups in the point cloud data at the current moment and the point cloud data at the next moment.

[0075] S204 : Determine a translation vector between the first measured point and the second measured point in the matching point group, and convert the translation vector into a translation distance.

[0076] The translation vector reflects the distance and direction of movement between two frames of point cloud data. In other words, the translation vector essentially reflects the actual movement of the target vehicle from one moment to the next. The translation distance, on the other hand, reflects the actual distance the target vehicle has traveled from one moment to the next.

[0077] Specifically, the translation vector between the first measured point and the second measured point can be calculated using the following formula:

[0078] T=argmin T ∑ i ||p i -q i || 2 ;

[0079] Where T is the translation vector; p i is the first measured point; q i The second measured point.

[0080] After obtaining the translation vector, a modulo calculation can be performed on the translation vector to obtain the translation distance.

[0081] S205: Determine positioning alarm information based on the translation distance and a preset distance threshold.

[0082] The preset distance threshold is used to reflect the acceptable range of vehicle position deviation. Specifically, based on the relationship between the translation distance and the preset distance threshold, the positioning alarm information can be determined.

[0083] As an optional but non-limiting implementation, the distance threshold includes a first distance threshold and a second distance threshold, wherein the first distance threshold is less than the second distance threshold. Accordingly, a positioning alarm message is determined based on the translation distance and the preset distance threshold, including: if the translation distance is greater than the first distance threshold but not greater than the second distance threshold, an alarm mechanism is triggered by the positioning monitoring module, and a prompt message is displayed on the target vehicle's interactive interface; if the translation distance is greater than the second distance threshold, an emergency stop command is sent to the autonomous driving system through the positioning monitoring module, and the autonomous driving system is notified of a positioning anomaly.

[0084] Specifically, the distance threshold can be set based on actual needs. The distance threshold can be set to two distance thresholds, one large and one small, where the smaller value is used as the first distance threshold and the larger value is used as the second distance threshold. If the translation distance is greater than the first distance threshold but not greater than the second distance threshold, indicating that the position of the target vehicle has deviated, but the degree of deviation is not too serious, the positioning monitoring module can be used to trigger an alarm mechanism and display a prompt message on the interactive interface of the target vehicle; if the translation distance is greater than the second distance threshold, indicating that the position of the target vehicle has deviated significantly and the degree of deviation is relatively serious, the positioning monitoring module will send an emergency stop command to the automatic driving system and prompt the automatic driving system that there is a positioning abnormality.

[0085] S206. Determine the area ratio relationship between the perceived drivable area and the actual drivable area through the perception monitoring module; if the area ratio relationship exceeds a preset ratio threshold, detect the roadside through the perception monitoring module to obtain a roadside detection result, and determine the perception alarm information based on the roadside detection result; wherein, the perceived drivable area is the road area where the target vehicle is passable as determined by the automatic driving system; the actual drivable area is the road area where the target vehicle is actually passable that corresponds to the perceived drivable area in the high-precision map.

[0086] As an optional but non-limiting implementation, detecting the roadside using the perception monitoring module to obtain the roadside detection result may include the following steps B1-B5:

[0087] Step B1: The perception monitoring module is used to fit the perception results output by the autonomous driving system to obtain a first roadside line.

[0088] The first roadside line is a roadside line in the perception result of the autonomous driving system. Specifically, the first roadside line is obtained by fitting the perception result output by the autonomous driving system through the perception monitoring module.

[0089] Step B2: Determine a second route corresponding to the first route based on the high-precision map.

[0090] The second road edge line corresponds to the first road edge line in the high-definition map. Specifically, the position information obtained based on the positioning of the vehicle and the high-definition map can be used to determine the second road edge line corresponding to the first road edge line.

[0091] Step B3: interval sampling is performed on the first road edge line to determine each fitting sampling point, and the actual sampling point corresponding to each fitting sampling point is determined on the second road edge line.

[0092] In the interval sampling, the sampling interval can be set differently according to actual needs. For example, the sampling interval can be 2 m. Specifically, interval sampling can be performed on the first road edge line according to the sampling interval, so as to determine each fitting sampling point. Then, each actual sampling point corresponding to each fitting sampling point is determined on the second road edge line.

[0093] Step B4: determine the distance deviation value between each fitting sampling point and the corresponding actual sampling point.

[0094] The distance deviation value is the distance value of the fitting sampling point and the corresponding actual sampling point in the bird's eye view, that is, the distance value of the two after projection on the same plane. In the embodiment of the application, the distance deviation value can be calculated based on the following formula:

[0095]

[0096] wherein x i is the horizontal coordinate of the fitting sampling point, y i is the vertical coordinate of the fitting sampling point; is the horizontal coordinate of the actual sampling point, is the vertical coordinate of the actual sampling point; d i is the distance deviation value between the fitting sampling point (x i , y i ) and the actual sampling point .

[0097] Similarly, the distance deviation value between each fitting sampling point and the corresponding actual sampling point can be determined based on the above formula.

[0098] Step B5: determine the road edge line detection result based on the numerical distribution of each distance deviation value.

[0099] The numerical distribution is used to reflect the dispersion of each distance deviation value calculated.

[0100] Specifically, since the sampling points follow a specific order, the resulting distance deviation values ​​also follow a specific order. If only one distance deviation value fails to meet the requirements, this could be due to inaccurate data. However, if several consecutive distance deviation values ​​fail to meet the requirements, this could indicate an anomaly in the autonomous driving system's roadside detection. Therefore, the roadside detection results can be determined based on the numerical distribution of the distance deviation values.

[0101] As an optional but non-limiting implementation method, the roadside detection result is determined based on the numerical distribution between each distance deviation value, including: if a preset number of consecutive distance deviation values ​​in each distance deviation value are all smaller than the distance deviation threshold, then the roadside detection result is that there is an abnormality in the roadside detection.

[0102] Specifically, if a preset number of consecutive distance deviation values ​​among the distance deviation values ​​are all smaller than the distance deviation threshold, it means that the first roadside line and the second roadside line obtained by the perception fitting of the autonomous driving system are already quite different. In this case, the roadside line detection result is that there is an abnormality in the roadside line detection.

[0103] S207. Determine, through the planning monitoring module, the number of failed trajectory update operations of the autonomous driving system within a preset time period, and determine a planning alarm message based on the relationship between the number of failures and a preset threshold; the trajectory update operation is used to update the trajectory points of the target vehicle during driving.

[0104] As an optional but non-limiting implementation, determining, by the planning monitoring module, the number of failed trajectory update operations of the autonomous driving system within a preset time period may include the following steps C1-C3:

[0105] Step C1: Determine each trajectory point planned by the autonomous driving system within a preset time period through the planning monitoring module.

[0106] The track points indicate the locations that the target vehicle needs to reach at various points in time during its journey. The automated driving system performs track updates to update the target vehicle's track points during its journey, allowing the target vehicle to use the track points as aiming points at various points in time and move toward the corresponding track points.

[0107] Specifically, the planning monitoring module can monitor the topic messages of the autonomous driving system to determine the various trajectory points planned by the autonomous driving system within a preset time period.

[0108] Step C2: determining the moving distance between adjacent track points in each track point; the moving distance is used to reflect the movement status between two adjacent track points.

[0109] Specifically, each trajectory point is planned for the autonomous driving system, and the moving distance between all two adjacent trajectory points in each trajectory point can be determined.

[0110] Step C3: Count the number of movement distances that are less than the preset distance value to obtain a statistical number; and use the statistical number as the number of failures.

[0111] The preset distance value can be set based on actual needs, or it can be set to 0 by default. Specifically, if the moving distance between two adjacent trajectory points is less than the preset distance value, it means that no update has been achieved between the two adjacent trajectory points. Therefore, the number of movement distances less than the preset distance value can be counted to obtain a statistical number, and the statistical number can be used as the number of failures.

[0112] As an optional but non-limiting implementation, the planning monitoring module calculates the collision time between the target vehicle and a reference obstacle in real time. The reference obstacle is an obstacle surrounding the target vehicle, and the collision time is determined based on a first distance and a first speed. The first distance is the distance between the target vehicle and the reference obstacle, and the first speed is the relative speed between the target vehicle and the reference obstacle. If the collision time falls below a safety time threshold, the planning monitoring module issues an alarm. If the number of alarms issued by the planning monitoring module exceeds a set threshold, the planning monitoring module sends an emergency stop command to the autonomous driving system and notifies the planning monitoring module of an abnormality.

[0113] The reference obstacle is an obstacle surrounding the target vehicle, such as a road bollard or guardrail. The time-to-collision (TCR) can be used to reflect the potential collision risk of the target vehicle and quickly assess the safety of the vehicle's trajectory. The TCR can be determined by dividing a first distance (the distance between the target vehicle and the reference obstacle) by a first speed (the relative speed between the target vehicle and the reference obstacle).

[0114] Specifically, after calculating the collision time, if the collision time is lower than the safety time threshold, an alarm will be issued through the planning monitoring module; if the number of alarm prompts issued by the planning monitoring module exceeds the set threshold, an emergency stop command will be sent to the automatic driving system through the planning monitoring module, and the planning monitoring module will be prompted to have an abnormality.

[0115] S208, calculating, by the control monitoring module, a lateral and longitudinal error between the actual control position and the target control position of the target vehicle, and determining the control alarm information based on a relationship between the lateral and longitudinal error and an error threshold; wherein the actual control position is a position actually reached by the target vehicle after being controlled by the automatic driving system according to the vehicle control parameter; and the target control position is a position expected to be reached by the target vehicle after being controlled by the automatic driving system according to the vehicle control parameter.

[0116] As an optional but non-limiting implementation manner, the lateral and longitudinal error includes a lateral error and a longitudinal error; correspondingly, the lateral and longitudinal error between the actual control position and the target control position of the target vehicle calculated by the control monitoring module can be calculated by the following formula:

[0117] e x =|T e,x -T a,x |;

[0118] e y =|T e,y -T a,y |;

[0119] wherein e x is the lateral error; e y is the longitudinal error; T e,x is the horizontal coordinate corresponding to the target control position; T e,y is the vertical coordinate corresponding to the target control position; T a,x is the horizontal coordinate corresponding to the actual control position; and T a,y is the vertical coordinate corresponding to the actual control position.

[0120] S209, recording, by the task monitoring module, execution state information of the automatic driving system in each execution stage when processing the target task, and determining the task alarm information based on the execution state information;

[0121] S210, detecting, by the interaction monitoring module, communication state information between the front-end interaction interface and the back-end of the automatic driving system, and determining the interaction alarm information based on the communication state information.

[0122] The technical solution provided by the present invention comprises: a positioning monitoring module comparing a first position and a second position of a target vehicle at a next moment to obtain a position deviation result, and determining a positioning alarm message based on the position deviation result; a perception monitoring module determining an area ratio between a perceived drivable area and an actual drivable area; if the area ratio exceeds a preset ratio threshold, the perception monitoring module detecting a roadside line to obtain a roadside line detection result, and determining a perception alarm message based on the roadside line detection result; a planning monitoring module determining the number of failed trajectory update operations by the autonomous driving system within a preset time period, and determining a planning alarm message based on the relationship between the number of failures and a preset number threshold; a control monitoring module calculating the lateral and longitudinal errors between the actual control position and the target control position of the target vehicle, and determining a control alarm message based on the relationship between the lateral and longitudinal errors and the error threshold; a task monitoring module recording execution status information at each execution stage when the autonomous driving system processes a target task, and determining a task alarm message based on the execution status information; and an interaction monitoring module detecting communication status information between a front-end interactive interface and a back-end of the autonomous driving system, and determining an interaction alarm message based on the communication status information. The solution of the present invention can monitor the positioning, perception, planning, control, tasks and interaction links of the autonomous driving system separately. Through multi-level, multi-dimensional and all-round monitoring, it can promptly detect anomalies in the autonomous driving system and quickly and accurately locate the link where the anomaly occurs, thereby improving the safety and stability of the autonomous driving system.

[0123] Figure 3 This is a schematic diagram of the structure of a monitoring system for an autonomous driving system provided by an embodiment of the present invention. The embodiment of the present invention is applicable to monitoring the autonomous driving system configured on a vehicle. The monitoring system for the autonomous driving system can be implemented in the form of software and / or hardware. The monitoring system for the autonomous driving system can be deployed on the vehicle and run based on an independent thread. Figure 3 As shown, the monitoring system for the autonomous driving system according to the embodiment of the present invention may include a positioning monitoring module 310, a perception monitoring module 320, a planning monitoring module 330, a control monitoring module 340, a task monitoring module 350, and an interaction monitoring module 360. Among them:

[0124] A positioning monitoring module 310 is configured to compare a first position and a second position of a target vehicle at a next moment to obtain a position deviation result, and to determine a positioning alarm message based on the position deviation result; wherein the first position is a predicted position of the target vehicle at the next moment based on historical position information of the target vehicle; and the second position is an actual position of the target vehicle at the next moment based on positioning by an autonomous driving system configured on the target vehicle;

[0125] The perception monitoring module 320 is configured to determine the area ratio between the perceived drivable area and the actual drivable area; if the area ratio exceeds a preset ratio threshold, the module detects the roadside to obtain a roadside detection result, and determines a perception alarm message based on the roadside detection result; wherein the perceived drivable area is the road area where the target vehicle is permitted to pass as determined by the autonomous driving system; and the actual drivable area is the road area where the target vehicle is permitted to pass as determined by the autonomous driving system in the high-precision map.

[0126] A planning monitoring module 330 is configured to determine the number of failed trajectory update operations performed by the autonomous driving system within a preset time period, and to determine a planning alarm message based on a relationship between the number of failed operations and a preset threshold number; the trajectory update operation is configured to update the trajectory points of the target vehicle during driving;

[0127] The control monitoring module 340 is configured to calculate the lateral and longitudinal errors between the actual control position of the target vehicle and the target control position, and to determine a control alarm message based on the relationship between the lateral and longitudinal errors and the error thresholds; wherein the actual control position is the position actually reached by the autonomous driving system after controlling the target vehicle according to the vehicle control parameters; and the target control position is the position that the autonomous driving system expects the target vehicle to reach after controlling the target vehicle according to the vehicle control parameters;

[0128] a task monitoring module 350 for recording execution status information of the autonomous driving system at various execution stages when processing a target task, and determining task alarm information based on the execution status information;

[0129] The interaction monitoring module 360 ​​is used to detect the communication status information between the front-end interactive interface and the back-end of the autonomous driving system, and determine the interaction alarm information based on the communication status information.

[0130] As an optional but non-limiting implementation manner, the historical position information comprises position information of the target vehicle at a current time and position information of the target vehicle at a previous time; correspondingly, the first position determination process comprises: obtaining, by the positioning monitoring module, position information of the target vehicle at the current time and position information of the target vehicle at the previous time obtained by the automatic driving system positioning; inputting, by the positioning monitoring module, the position information at the current time and the position information at the previous time into the trajectory prediction model, and outputting the first position; wherein the trajectory prediction model can predict the position information to be reached by the target vehicle at a next time based on the position information of the target vehicle at the current time.

[0131] As an optional but non-limiting implementation manner, the positioning monitoring module 310 comprises a point cloud data acquisition unit, a matching point group determination unit, a translation distance determination unit, and a positioning alarm information determination unit. Wherein: the point cloud data acquisition unit is configured to, if the position deviation result exceeds the preset deviation threshold, acquire point cloud data of the target vehicle at a current time and point cloud data of the target vehicle at a next time; the matching point group determination unit is configured to start a point cloud matching algorithm to determine a corresponding matching point group in the point cloud data at the current time and the point cloud data at the next time; wherein the matching point group comprises a first measured point and a second measured point, the first measured point belongs to the point cloud data at the current time, and the second measured point belongs to the point cloud data at the next time; the translation distance determination unit is configured to determine a translation vector between the first measured point and the second measured point in the matching point group, and convert the translation vector into a translation distance; and the positioning alarm information determination unit is configured to determine the positioning alarm information based on the translation distance and a preset distance threshold.

[0132] As an optional but non-limiting implementation manner, the distance threshold comprises a first distance threshold and a second distance threshold, wherein the first distance threshold is smaller than the second distance threshold; correspondingly, the positioning alarm information determination unit is specifically configured to: if the translation distance is greater than the first distance threshold but not greater than the second distance threshold, trigger an alarm mechanism and display prompt information on an interactive interface of the target vehicle; and if the translation distance is greater than the second distance threshold, send an emergency stop instruction to the automatic driving system and prompt the automatic driving system to have a positioning abnormality.

[0133] As an optional but non-limiting implementation, the perception monitoring module 320 includes a first roadside determination unit, a second roadside determination unit, a sampling point determination unit, a distance deviation value determination unit, and a roadside detection result determination unit. The first roadside determination unit is configured to fit the perception result output by the autonomous driving system to obtain a first roadside line; the second roadside determination unit is configured to determine a second roadside line corresponding to the first roadside line based on the high-precision map; the sampling point determination unit is configured to perform interval sampling on the first roadside line to determine each fitting sampling point, and to determine the actual sampling point corresponding to each fitting sampling point on the second roadside line; the distance deviation value determination unit is configured to determine the distance deviation value between each fitting sampling point and the corresponding actual sampling point; and the roadside detection result determination unit is configured to determine the roadside detection result based on the numerical distribution between each distance deviation value.

[0134] As an optional but non-limiting implementation method, the roadside line detection result determination unit is specifically used to: if a preset number of consecutive distance deviation values ​​in each of the distance deviation values ​​are all smaller than the distance deviation threshold, then the roadside line detection result is that there is an abnormality in the roadside line detection.

[0135] As an optional but non-limiting implementation, the planning monitoring module 330 includes a trajectory point determination unit, a movement distance determination unit, and a failure count unit. The trajectory point determination unit is configured to determine each trajectory point planned by the autonomous driving system within a preset time period; the movement distance determination unit is configured to determine the movement distance between adjacent trajectory points within each trajectory point; the movement distance is configured to reflect the movement between two adjacent trajectory points; and the failure count unit is configured to calculate a statistical number of each movement distance that is less than a preset distance value, and use the statistical number as the failure count.

[0136] As an optional but non-limiting implementation, the planning monitoring module 330 is further used to: calculate in real time the collision time between the target vehicle and a reference obstacle; the reference obstacle is an obstacle around the target vehicle, and the collision time is determined based on a first distance and a first speed; wherein the first distance is the distance between the target vehicle and the reference obstacle; the first speed is the relative speed between the target vehicle and the reference obstacle; if the collision time is lower than a safety time threshold, an alarm is issued; if the number of alarm prompts exceeds a set number threshold, an emergency stop command is sent to the autonomous driving system, and the autonomous driving system is prompted that a planning abnormality has occurred.

[0137] As an optional but non-limiting implementation, the lateral and longitudinal errors include lateral errors and longitudinal errors; accordingly, the control monitoring module 340 is used to calculate the lateral and longitudinal errors between the actual control position and the target control position of the target vehicle, specifically using the following formula:

[0138] e x =|T e,x -T a,x |;

[0139] e y =|T e,y -T a,y |;

[0140] Among them, e x is the lateral error; e y is the longitudinal error; T e,x is the horizontal coordinate corresponding to the target control position; T e,y is the vertical coordinate corresponding to the target control position; T a,x is the horizontal coordinate corresponding to the actual control position; T a,y is the vertical coordinate corresponding to the actual control position.

[0141] The technical solution provided by the present invention includes a positioning monitoring module for comparing the first and second positions of the target vehicle at the next moment to obtain a position deviation result, and determining a positioning alarm message based on the position deviation result; a perception monitoring module for determining the area ratio relationship between the perceived drivable area and the actual drivable area; if the area ratio relationship exceeds a preset ratio threshold, detecting the roadside to obtain a roadside detection result, and determining a perception alarm message based on the roadside detection result; a planning monitoring module for determining the number of failed trajectory update operations of the autonomous driving system within a preset time period, and determining a planning alarm message based on the relationship between the number of failures and a preset number threshold; a control monitoring module for calculating the lateral and longitudinal errors between the actual control position and the target control position of the target vehicle, and determining a control alarm message based on the relationship between the lateral and longitudinal errors and the error threshold; a task monitoring module for recording execution status information at each execution stage when the autonomous driving system processes the target task, and determining a task alarm message based on the execution status information; and an interaction monitoring module for detecting communication status information between the front-end interactive interface and the back-end of the autonomous driving system, and determining an interaction alarm message based on the communication status information. By adopting the solution of the present invention, the positioning, perception, planning, control, tasks and interaction links of the autonomous driving system can be detected separately. Through multi-level, multi-dimensional and all-round monitoring, anomalies in the autonomous driving system can be discovered in a timely manner, and the link where the anomaly occurs can be quickly and accurately located, thereby improving the safety and stability of the autonomous driving system.

[0142] The monitoring system for an autonomous driving system provided in an embodiment of the present invention can be used to execute a monitoring method for an autonomous driving system, and has corresponding functional modules and beneficial effects for executing the monitoring method for an autonomous driving system.

[0143] It is worth noting that the various units and modules included in the above system are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the embodiments of the present invention.

[0144] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the module, program segment, or a part of code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order than that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of the boxes in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for executing the specified function or operation, or can be implemented by a combination of dedicated hardware and computer instructions.

[0145] The units involved in the embodiments of the present invention may be implemented in software or in hardware. The name of the unit does not, in some cases, limit the unit itself. The functions described above may be at least partially performed by one or more hardware logic components. For example, exemplary types of hardware logic components that may be used include, but are not limited to, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0146] The above description is merely a preferred embodiment of the present invention and an illustration of the technical principles employed. Those skilled in the art should understand that the scope of disclosure involved in the present invention is not limited to the technical solutions formed by the specific combination of the above-mentioned technical features, but also includes other technical solutions formed by any combination of the above-mentioned technical features or their equivalents without departing from the above-mentioned disclosed concepts. For example, a technical solution formed by replacing the above-mentioned features with (but not limited to) technical features with similar functions disclosed in the present invention.

[0147] In addition, although adopting specific order to describe each operation, this should not be interpreted as requiring these operations to be executed in the specific order shown or in sequential order.Under certain environment, multitasking and parallel processing may be advantageous.Similarly, although comprising some specific implementation details in the above discussion, these should not be interpreted as limiting the scope of the present invention.Some features described in the context of independent embodiment can also be implemented in single embodiment in combination.On the contrary, the various features described in the context of independent embodiment also can be implemented in multiple embodiments individually or in the mode of any suitable subcombination.

[0148] Although the subject matter has been described in language specific to structural features and / or methodological logical acts, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are merely example forms of implementing the claims.

Claims

1. A monitoring method for an automatic driving system, characterized in that: The method comprises: A positioning monitoring module compares a first position and a second position of the target vehicle at a next moment to obtain a position deviation result, and determines a positioning alarm message based on the position deviation result; wherein the first position is a position of the target vehicle at the next moment predicted based on historical position information of the target vehicle; and the second position is an actual position of the target vehicle at the next moment obtained based on positioning by an autonomous driving system; and the autonomous driving system is configured on the target vehicle; The perception monitoring module determines the area ratio between the perceived drivable area and the actual drivable area; if the area ratio exceeds a preset ratio threshold, the perception monitoring module detects the roadside to obtain a roadside detection result, and determines a perception alarm message based on the roadside detection result; wherein the perceived drivable area is the road area where the target vehicle is navigable as determined by the autonomous driving system; and the actual drivable area is the road area where the target vehicle is actually navigable corresponding to the perceived drivable area in the high-precision map; Determining, by a planning monitoring module, the number of failed trajectory update operations of the autonomous driving system within a preset time period, and determining a planning alarm message based on a relationship between the number of failures and a preset threshold; the trajectory update operation is used to update the trajectory points of the target vehicle during driving; The control monitoring module calculates the lateral and longitudinal errors between the actual control position and the target control position of the target vehicle, and determines a control alarm message based on the relationship between the lateral and longitudinal errors and the error threshold; wherein the actual control position is the position actually reached by the automatic driving system after controlling the target vehicle according to the vehicle control parameters; and the target control position is the position expected to be reached by the automatic driving system after controlling the target vehicle according to the vehicle control parameters; Recording, by means of a task monitoring module, execution status information of the autonomous driving system at each execution stage when processing a target task, and determining task alarm information based on the execution status information; The interaction monitoring module detects the communication status information between the front-end interaction interface and the back-end of the autonomous driving system, and determines the interaction alarm information based on the communication status information.

2. The method according to claim 1, characterized in that The historical location information includes the location information of the target vehicle at the current moment and the location information at the previous moment; Accordingly, the process of determining the first position includes: Obtaining the current position information and the previous position information of the target vehicle obtained by the automatic driving system through the positioning monitoring module; The position information at the current moment and the position information at the previous moment are input into the trajectory prediction model through the positioning monitoring module, and the first position is output; wherein, the trajectory prediction model can predict the position information that the target vehicle will reach at the next moment based on the position information of the target vehicle at the current moment.

3. The method according to claim 2, characterized in that The determining of positioning alarm information based on the position deviation result includes: If the position deviation result exceeds the preset deviation threshold, the point cloud data of the target vehicle at the current moment and the point cloud data at the next moment are obtained through the positioning monitoring module; Starting a point cloud matching algorithm to determine a corresponding matching point group in the point cloud data at a current moment and the point cloud data at a next moment; wherein the matching point group includes a first measured point and a second measured point, the first measured point belonging to the point cloud data at the current moment, and the second measured point belonging to the point cloud data at the next moment; Determine a translation vector between the first measured point and the second measured point in the matching point group, and convert the translation vector into a translation distance; The positioning alarm information is determined based on the translation distance and a preset distance threshold.

4. The method according to claim 3, characterized in that The distance threshold includes a first distance threshold and a second distance threshold, wherein the first distance threshold is smaller than the second distance threshold; Accordingly, the determining of the positioning alarm information based on the translation distance and a preset distance threshold includes: If the translation distance is greater than the first distance threshold but not greater than the second distance threshold, an alarm mechanism is triggered by the positioning monitoring module, and a prompt message is displayed on the interactive interface of the target vehicle; If the translation distance is greater than a second distance threshold, an emergency stop instruction is sent to the automatic driving system through the positioning monitoring module, and the automatic driving system is prompted that a positioning abnormality occurs.

5. The method according to claim 1, wherein The detecting of the roadside line by the perception monitoring module to obtain the roadside line detection result includes: Fitting the perception results output by the autonomous driving system through the perception monitoring module to obtain a first roadside line; Determining a second route corresponding to the first route based on the high-precision map; Performing interval sampling on the first roadside to determine each fitting sampling point, and determining the actual sampling point corresponding to each fitting sampling point on the second roadside; Determine the distance deviation value between each fitting sampling point and the corresponding actual sampling point; The roadside detection result is determined based on the numerical distribution between the respective distance deviation values.

6. The method according to claim 5, characterized in that The determining of the roadside detection result based on the numerical distribution of each of the distance deviation values ​​includes: If a preset number of consecutive distance deviation values ​​among the distance deviation values ​​are all smaller than the distance deviation threshold, the roadside line detection result indicates that an abnormality exists in the roadside line detection.

7. The method according to claim 1, characterized in that The determining, by the planning monitoring module, the number of failed trajectory update operations of the autonomous driving system within a preset time period includes: The planning monitoring module determines the trajectory points planned by the autonomous driving system within a preset time period; Determine the movement distance between adjacent trajectory points in each of the trajectory points; the movement distance is used to reflect the movement status between two adjacent trajectory points; The number of movement distances that are less than a preset distance value is counted to obtain a statistical number; and the statistical number is used as the number of failures.

8. The method according to claim 1, characterized in that The method further comprises: The planning and monitoring module calculates in real time a collision time between the target vehicle and a reference obstacle; the reference obstacle is an obstacle surrounding the target vehicle, and the collision time is determined based on a first distance and a first speed; wherein the first distance is the distance between the target vehicle and the reference obstacle; and the first speed is the relative speed between the target vehicle and the reference obstacle; If the collision time is lower than the safety time threshold, an alarm is issued through the planning monitoring module; If the number of alarm prompts issued by the planning monitoring module exceeds the set threshold, an emergency stop command will be sent to the autonomous driving system through the planning monitoring module, and the autonomous driving system will be prompted that a planning abnormality has occurred.

9. The method according to claim 1, characterized in that The transverse and longitudinal errors include transverse errors and longitudinal errors; Accordingly, the lateral and longitudinal errors between the actual control position and the target control position of the target vehicle are calculated by the control monitoring module using the following formula: e x =|T e,x -T a,x |; e y =|T e,y -T a,y |; Among them, e x is the lateral error; e y is the longitudinal error; T e,x is the horizontal coordinate corresponding to the target control position; T e,y is the vertical coordinate corresponding to the target control position; T a,x is the horizontal coordinate corresponding to the actual control position; T a,y is the vertical coordinate corresponding to the actual control position.

10. A monitoring system for an autonomous driving system, characterized in that: The system comprises: a positioning monitoring module, configured to compare a first position and a second position of a target vehicle at a next moment to obtain a position deviation result, and to determine a positioning alarm message based on the position deviation result; wherein the first position is a position of the target vehicle at the next moment predicted based on historical position information of the target vehicle; and the second position is an actual position of the target vehicle at the next moment obtained based on positioning by an autonomous driving system, wherein the autonomous driving system is configured on the target vehicle; A perception monitoring module is configured to determine a ratio between the perceived drivable area and the actual drivable area; if the ratio exceeds a preset ratio threshold, the module detects the roadside to obtain a roadside detection result, and determines a perception alarm message based on the roadside detection result; wherein the perceived drivable area is the road area perceived by the autonomous driving system as passable by the target vehicle; and the actual drivable area is the road area in the high-precision map corresponding to the perceived drivable area and actually passable by the target vehicle; a planning monitoring module, configured to determine a number of failed trajectory update operations by the autonomous driving system within a preset time period, and to determine a planning alarm message based on a relationship between the number of failures and a preset threshold; the trajectory update operation is configured to update trajectory points of the target vehicle during driving; a control monitoring module, configured to calculate the lateral and longitudinal errors between the actual control position of the target vehicle and the target control position, and to determine a control alarm message based on the relationship between the lateral and longitudinal errors and the error thresholds; wherein the actual control position is the position actually reached by the automatic driving system after controlling the target vehicle according to the vehicle control parameters; and the target control position is the position expected to be reached by the automatic driving system after controlling the target vehicle according to the vehicle control parameters; A task monitoring module is used to record execution status information of the autonomous driving system at each execution stage when processing the target task, and determine task alarm information based on the execution status information; The interaction monitoring module is used to detect the communication status information between the front-end interactive interface and the back-end of the autonomous driving system, and determine the interaction alarm information based on the communication status information.

Images