Vehicle control method and system, electronic equipment, storage medium and product

Abstract

The invention discloses a vehicle control method and system, electronic equipment, a storage medium and a product. The method is applied to a cloud server, and comprises the following steps: in response to a passing request of a target vehicle for a target area, obtaining vehicle driving data when a plurality of historical vehicles pass in the target area; determining a road surface obstacle existing in the target area based on the vehicle driving data; generating a vehicle control instruction based on the road obstacle and the vehicle model parameter of the target vehicle; and issuing the vehicle control instruction to the target vehicle, so that the target vehicle executes passage control based on the vehicle control instruction. According to the scheme, the driving data of the plurality of historical vehicles in the target area are gathered, the cloud server analyzes and determines the road obstacle, and then the personalized vehicle control instruction is generated in combination with the specific model parameters of the target vehicle, so that the robustness and accuracy of road obstacle recognition are remarkably improved, and the user experience is improved. And cloud cooperative vehicle control with high reliability and high adaptability is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic driving, and in particular to a vehicle control method and system, an electronic device, a storage medium and a product. BACKGROUND

[0002] In the existing autonomous driving technology, the perception of road conditions mainly relies on the direct scanning and identification of the road surface by the vehicle-mounted vision or laser radar sensors. However, this direct perception method is prone to cause perception blind spots or misjudgments when facing adverse weather, road debris or front vehicle shielding, etc. environmental conditions, which leads to the vehicle control system being unable to obtain reliable road obstacle information, thereby affecting its high-precision and high-safety collaborative control decision-making. SUMMARY

[0003] The present application provides a vehicle control method and system, an electronic device, a storage medium and a product to solve the problem of affecting the safety and accuracy of vehicle collaborative control decision-making due to the inability to obtain reliable road obstacle information in the prior art.

[0004] According to an aspect of the present application, a vehicle control method is provided, applied to a cloud server, the method comprising:

[0005] In response to a target vehicle's request for passing through a target area, obtaining vehicle driving data of multiple historical vehicles when passing through the target area;

[0006] Based on the vehicle driving data, determining the road obstacles existing in the target area;

[0007] Based on the road obstacles and the vehicle model parameters of the target vehicle, generating a vehicle control instruction;

[0008] Issuing the vehicle control instruction to the target vehicle, so that the target vehicle executes the passing control based on the vehicle control instruction.

[0009] According to another aspect of the present application, a vehicle control system is provided, configured in a cloud server, the system comprising:

[0010] A data acquisition module for obtaining vehicle driving data of multiple historical vehicles when passing through a target area in response to a target vehicle's request for passing through the target area;

[0011] An obstacle determination module for determining the road obstacles existing in the target area based on the vehicle driving data;

[0012] An instruction generation module for generating a vehicle control instruction based on the road obstacles and the vehicle model parameters of the target vehicle;

[0013] The instruction issuing module is configured to issue the vehicle control instruction to the target vehicle, so that the target vehicle performs the passage control based on the vehicle control instruction.

[0014] According to another aspect of the present application, there is provided an electronic device comprising:

[0015] at least one processor; and

[0016] a memory communicatively connected to the at least one processor; wherein

[0017] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to perform the vehicle control method according to any one of the embodiments of the present application.

[0018] According to another aspect of the present application, there is provided a computer readable storage medium storing computer instructions for enabling a processor to perform the vehicle control method according to any one of the embodiments of the present application when executed by the processor.

[0019] According to another aspect of the present application, there is provided a computer program product comprising a computer program for enabling a processor to perform the vehicle control method according to any one of the embodiments of the present application when executed by the processor.

[0020] The technical solution of the embodiments of the present application is applied to a cloud server, and the cloud server acquires vehicle driving data of a plurality of historical vehicles when passing through a target area in response to a passage request of a target vehicle to the target area, determines a road obstacle existing in the target area based on the vehicle driving data, generates a vehicle control instruction based on the road obstacle and a vehicle model parameter of the target vehicle, and issues the vehicle control instruction to the target vehicle, so that the target vehicle performs passage control based on the vehicle control instruction. The technical solution collects driving data of a plurality of historical vehicles in the target area, analyzes and determines a road obstacle by the cloud server, and then generates a personalized vehicle control instruction in combination with a specific model parameter of the target vehicle, effectively solves the defect that reliable road condition information cannot be obtained due to a blind area or misjudgment of single vehicle perception in the prior art, significantly improves the robustness and accuracy of road obstacle identification, and realizes cloud collaborative vehicle control with high reliability and high adaptability.

[0021] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present application, nor to limit the scope of the present application. Other features of the present application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed in the embodiment description. Obviously, the drawings described below only show some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without any creative effort.

[0023] Figure 1 is a flow chart of a vehicle control method according to an embodiment of the present application;

[0024] Figure 2 is a structural schematic diagram of a vehicle control system according to an embodiment of the present application;

[0025] Figure 3 is a structural schematic diagram of an electronic device implementing the vehicle control method of the present application. DETAILED DESCRIPTION

[0026] In order to make the technical personnel in the art better understand the present application, the following will combine the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without any creative effort should be within the scope of protection of the present application.

[0027] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not necessarily limit to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0028] Embodiment one

[0029] Figure 1 The flow chart of a vehicle control method according to an embodiment of the present application is provided, and the present embodiment can be applied to the case of indirectly sensing road obstacles based on historical vehicle driving data and providing cooperative control strategy for the target vehicle currently requesting to pass. The method can be executed by a vehicle control system, which can be realized in the form of hardware and / or software, and the vehicle control system can be configured in an electronic device. As shown inFigure 1 The vehicle control method provided in this embodiment is applied to a cloud server and specifically includes the following steps.

[0030] S110, in response to a request for passing through the target area by the target vehicle, obtaining vehicle driving data of a plurality of historical vehicles passing through the target area.

[0031] The cloud server can be a server processing device or a server cluster deployed in the cloud, capable of receiving vehicle requests and performing data processing and instruction issuing. In this scheme, it is mainly responsible for indirectly sensing the road obstacle situation based on the historical vehicle driving data in the target area, and providing a cooperative control strategy for the target vehicle currently requesting to pass through.

[0032] The target vehicle can be any intelligent connected vehicle that plans to pass through the target area, actively sends a request for passing through to the cloud server, and finally receives a vehicle control instruction to perform a passing control. For example, it can include but is not limited to an unmanned vehicle (such as an L4 / L5 level autonomous vehicle), a conditional autonomous vehicle (L3 level), and an intelligent connected vehicle (L2+ level) with navigation driving assistance function (capable of automatically adjusting vehicle speed and path).

[0033] The request for passing through can be an interactive signal initiated by the target vehicle to the cloud server, containing core information of the request for passing through, used to trigger the cloud server to start the process of obtaining related historical data of the target area and generating subsequent control instructions. For example, the request for passing through can include the target vehicle identifier, the target area location information, the request service type (such as the request for road obstacle analysis and safe passing strategy acquisition), and the like.

[0034] The target area can be a specific geographic road section or area range that the target vehicle plans to drive into and pass through, and is also a driving area that needs to analyze the current road damage situation. Its boundary can be defined by various ways such as latitude and longitude coordinate interval, road section code, and geographic fence. It is the core spatial range for subsequent historical vehicle data screening and road obstacle identification.

[0035] The historical vehicle can be any vehicle that passed through the target area earlier than the target vehicle and uploaded its vehicle driving data to the cloud server. It can include all vehicle individuals (such as cars, trucks, motorcycles, electric bicycles, etc.) whose vehicle model parameters (such as suspension system parameters) can be covered and processed by the cloud server. These vehicles collectively constitute a passive and distributed road state sensing network.

[0036] The vehicle driving data can refer to various data sets collected by a historical vehicle during the driving process in the target area, which are related to the driving state, vehicle parameters and road surface interaction. The vehicle driving data can include, but are not limited to, vehicle model parameters, vehicle trajectory sequence, vehicle speed sequence, monitoring height sequence (used to reflect the vertical height of the monitoring point on the vehicle relative to the ground over time), etc.

[0037] In the embodiment of the present application, the cloud server can continuously monitor the communication link from the Internet of Vehicles. When the target vehicle plans to drive into a certain target area (such as a section of highway or urban intersection) and needs to analyze the current road loss of the area, the vehicle-mounted system will automatically generate a pass request (which can also be generated based on user manual selection of the target area, and the present embodiment does not limit this), and upload it to the cloud server through communication methods such as 5G, V2X, etc. After the cloud server receives the pass request, it will parse the target vehicle identifier, target area location information and other information from it, and then retrieve and subscribe to the original data packets uploaded by all historical vehicles that have passed through the target area within a predetermined historical period (such as the past 24 hours) according to the target area location information in the data storage node (such as cloud database, edge node cache, etc.). Then, from the subscribed data packets, the key vehicle driving data of each historical vehicle is extracted, and these data are preprocessed such as time stamp alignment, filtering and denoising, so as to form structured information that can be used for analysis, such as vehicle model parameters, vehicle trajectory data, vehicle speed, vehicle height monitoring sequence, etc.

[0038] S120, based on the vehicle driving data, determining the road obstacles existing in the target area.

[0039] The road obstacle can refer to the abnormal form of the road surface existing in the target area that affects the normal and safe driving of the vehicle, which can include, but is not limited to, potholes, bumps, cracks, loose manhole covers, road subsidence and other structures or forms that destroy the flatness of the road surface.

[0040] In the embodiment of the present application, the cloud server can determine the vehicle body height condition that each historical vehicle should theoretically maintain based on the vehicle model parameters and vehicle trajectory data of the historical vehicle, in combination with a high-precision road reference model (such as an ideal three-dimensional surface corresponding to the target area), that is, obtain a reference height sequence, which is used to establish a reference for the normal posture of the vehicle when the road surface is perfect; then, the actual monitored vehicle body height sequence (i.e., the monitoring height sequence) of each historical vehicle is compared with the corresponding reference height sequence point by point, and when there is a significant and continuous deviation between the actual height and the theoretical height, it is marked as a height difference point. By analyzing the distribution of the difference points of a single vehicle in the entire target area, the potential abnormal trajectory position perceived by the vehicle can be extracted. Since the abnormal signal of a single vehicle may be caused by occasional factors (such as avoiding debris), the cloud server can perform multi-vehicle cooperative analysis on the abnormal positions reported by all historical vehicles, for example, by counting the frequency of a specific geographic position being repeatedly reported as an abnormal point by different vehicles, and setting a reasonable threshold to filter out accidental interference, so as to reliably locate the stable road obstacles that are verified by multiple vehicles.

[0041] S130, generating a vehicle control instruction based on the road obstacle and the vehicle model parameters of the target vehicle.

[0042] The vehicle control instruction can be an instruction signal generated by the cloud server for guiding the target vehicle to safely pass through the target area, and its types can include but are not limited to driving avoidance instructions, speed threshold limit instructions, passing path adjustment instructions, etc.

[0043] In the embodiment of the present application, the cloud server can first determine the complete information of the road obstacle in the target area, including the type (concave / convex, etc.), spatial position, geometric characteristics (size, shape, etc.), and distribution range of the obstacle, and at the same time, call the vehicle model parameters of the target vehicle to ensure that the parameters completely cover the key dimensions related to the vehicle structure and driving performance; then, based on the information of the road obstacle and the model parameters of the target vehicle, the passing influence is evaluated, and whether the obstacle exceeds the safe passing ability of the target vehicle is analyzed, which can include, for example, the spatial interference risk evaluation of the obstacle and the vehicle chassis, the posture stability evaluation of the vehicle when passing through the obstacle, the adaptability evaluation of the braking / steering system, etc.; then, according to the evaluation result, the type of the vehicle control instruction is determined, for example, if the severity of the obstacle exceeds the safe passing ability of the target vehicle (such as the obstacle size is too large, the shape is complex, and the vehicle cannot safely pass through), a driving avoidance type instruction is determined to be generated; if the obstacle can be safely passed through but the driving state needs to be limited, a driving parameter limit type instruction (such as speed threshold, steering angle limit, etc.) is determined to be generated; finally, a protocol format is preset, and the information such as the instruction type, the obstacle position, and the instruction parameters (such as the speed limit value) is packaged into a structured vehicle control instruction (such as but not limited to JSON, Protobuf, etc.).

[0044] S140, issuing the vehicle control instruction to the target vehicle, so that the target vehicle performs the passage control based on the vehicle control instruction.

[0045] Wherein, the passage control can refer to the operation process of the target vehicle adjusting the driving state (such as switching the driving path, adjusting the vehicle speed) through the self-control system to avoid the road obstacle or safely pass through the obstacle area after receiving the vehicle control instruction.

[0046] In the embodiment of the application, the cloud server can issue the packaged vehicle control instruction to the target vehicle through a low-delay and high-reliability Internet of Vehicles communication link (such as 5G, C-V2X, etc.), and at the same time, in order to ensure the reliable delivery of key instructions, a multi-channel redundant transmission mechanism can also be used. After the vehicle terminal or domain controller of the target vehicle receives the instruction, it will perform security verification and analysis, identify the instruction type and specific content, and then adjust the corresponding execution components (such as the steering system, power system, brake system, etc.) combined with the current driving state of the vehicle, so as to perform the corresponding passage control operation, such as re-planning the driving path to avoid the obstacle area, reducing the vehicle speed to below the safety threshold, etc.

[0047] The technical scheme of the embodiment of the application is applied to a cloud server, and the cloud server obtains vehicle driving data of a plurality of historical vehicles when the vehicles pass through a target area in response to a passage request of a target vehicle to the target area; determines a road obstacle existing in the target area based on the vehicle driving data; generates a vehicle control instruction based on the road obstacle and a vehicle model parameter of the target vehicle; and issues the vehicle control instruction to the target vehicle, so that the target vehicle performs passage control based on the vehicle control instruction. The technical scheme collects the driving data of a plurality of historical vehicles in the target area, analyzes and determines the road obstacle by the cloud server, and then generates a personalized vehicle control instruction combined with the specific model parameter of the target vehicle, effectively solving the defect that reliable road condition information cannot be obtained due to the blind area or misjudgment of single vehicle perception in the prior art, significantly improving the robustness and accuracy of road obstacle identification, and realizing cloud collaborative vehicle control with high reliability and high adaptability.

[0048] Further, on the basis of the above-mentioned embodiment of the application, in response to a passage request of a target vehicle to a target area, the vehicle driving data of a plurality of historical vehicles when the vehicles pass through the target area is obtained, comprising:

[0049] According to the position information of the target area carried in the passage request, the vehicle driving data of a plurality of historical vehicles passing through the target area in a preset historical period is obtained; wherein the vehicle driving data at least includes the vehicle model parameter, the vehicle trajectory sequence, the vehicle speed sequence and the monitoring height sequence of the vehicle monitoring point relative to the road surface of each historical vehicle.

[0050] The position information can be data used to define the spatial range of the target area, for example, can include but is not limited to latitude and longitude coordinate interval, road name and road section start and end point, geographic area code, etc.

[0051] The preset historical period can be a time range (such as the past 7 days, the past 30 days, etc.) for filtering historical vehicles, which is set by the cloud server in advance or dynamically adjusted, and the setting needs to balance the data timeliness and data accumulation amount.

[0052] The vehicle model parameter can be a set of physical parameters used to uniquely identify or define the vehicle dynamics characteristics, for example, can include but is not limited to vehicle brand model code, chassis height, suspension system parameters, stiffness coefficient, tire radius, vehicle body size, vehicle mass, etc.

[0053] The vehicle trajectory sequence can be a set of continuous position coordinates recorded in time sequence when the historical vehicle passes through the target area, which can reflect the driving path and position change of the vehicle in the target area.

[0054] The vehicle speed sequence can be a set of instantaneous speed data recorded in time sequence synchronously when the historical vehicle passes through the target area, which can reflect the driving speed state of the vehicle at different positions and different times in the target area.

[0055] The vehicle monitoring point can be a preset sensing point on the historical vehicle for collecting relative height data from the road surface, for example, can include but is not limited to the position near the wheel, the chassis support, etc.

[0056] The monitoring height sequence can be a set of real-time vertical distance data from the road surface collected in time sequence by the vehicle monitoring point during the process of the historical vehicle passing through the target area.

[0057] In the embodiment of the application, when the target vehicle needs to analyze the current road damage situation of a target area, a passing request can be sent to the cloud server, so that the cloud server will trigger the road obstacle analysis and vehicle cooperative control process in this embodiment after receiving the request. Specifically, the cloud server can extract the position information of the target area from the passing request to determine the spatial boundary of the target area (such as determining the latitude and longitude coordinate range, locking specific road sections); then, combined with the pre-configured preset historical period (such as the past 24 hours or a week), all historical vehicles (i.e. different vehicles passing through the target area in the period) and their corresponding original perception data records are retrieved in the cloud database or other data storage locations; then the vehicle model parameter, vehicle trajectory sequence, vehicle speed sequence, and monitoring height sequence of each historical vehicle are extracted from the original perception data record, and through time stamp alignment, filtering and denoising, etc. preprocessing, finally forming the vehicle driving data corresponding to each historical vehicle.

[0058] The embodiment realizes accurate and efficient acquisition of historical vehicle driving data by constructing a dynamic data acquisition mechanism driven by a target vehicle request, reduces data redundancy through "space + time" double limitation, guarantees the pertinence and reliability of the data through directional acquisition of core data and elimination of invalid data, and improves the adaptability of the scheme to different road scenes through the flexible adjustment of the preset historical period. The group data support of multiple historical vehicles avoids the contingency of single data, and provides high-quality data guarantee for the subsequent cloud server to identify the road obstacles in the target area and generate safe control instructions adapted to the target vehicle.

[0059] Further, on the basis of the above-mentioned embodiment of the application, based on the vehicle driving data, the road obstacles existing in the target area are determined, comprising:

[0060] Based on the vehicle model parameters and the vehicle trajectory sequence of each historical vehicle, and in combination with the preset road reference model corresponding to the target area, the reference height sequence corresponding to each historical vehicle is determined;

[0061] The reference height sequence of each historical vehicle is compared with the corresponding monitoring height sequence to obtain the height difference information of each historical vehicle;

[0062] Based on the height difference information of each historical vehicle, the road abnormal area in the target area is determined through frequency statistics;

[0063] Based on the vehicle driving data associated with the road abnormal area, the contour model of the road obstacle is determined.

[0064] The preset road reference model can be a high-precision three-dimensional digital surface model pre-constructed to represent the ideal state of the road in the target area, which can include basic data such as static geographic parameters of the road (such as slope, curvature, coordinate boundary), road design flatness, reference horizontal plane coordinates, and can be used as a reference for evaluating whether the road has been damaged and deformed. In an embodiment, the position information of the target area can be input into the road network big data to retrieve the corresponding road distribution position data, and then a pre-set horizontal plane height is combined as a reference horizontal plane to generate the preset road reference model.

[0065] The height difference information can be related information reflecting the deviation between the theoretical height and the actual height of the monitoring point of the historical vehicle obtained by comparing the reference height sequence and the monitoring height sequence, which can include, for example, difference point position, deviation value, abnormal probability and other information.

[0066] The frequency statistics can be a statistical method for screening out common abnormal positions by quantitatively analyzing the frequency and proportion of a specific position being marked as a difference position.

[0067] The road surface abnormal area can refer to a continuous geographical area with a universal height difference filtered by frequency statistics, and is a core area basis for judging the existence of road surface obstacles.

[0068] The profile model of the road surface obstacle can refer to a digital model for quantitatively describing the three-dimensional geometric shape of the obstacle in the road surface abnormal area, for example, a concave model defined by a parameter equation (such as a quadratic surface), or a convex model defined by a voxel grid and a convex hull algorithm, etc., which can include key geometric parameters such as size, depth, and volume.

[0069] In the embodiments of the present application, the process of identifying the road surface obstacle in the target area can include:

[0070] (1) For each historical vehicle, the cloud server can calculate the theoretical height change process of each monitoring point (such as the wheel) of the vehicle when it is driving on an ideal flat road according to its specific vehicle model parameters (such as wheelbase, track, and suspension characteristics) and its vehicle trajectory sequence in the target area, combined with a high-precision preset road reference model, and then arrange these theoretical height values in the time / space order of the trajectory sequence to form the reference height sequence corresponding to the historical vehicle.

[0071] (2) For each historical vehicle, the reference height sequence calculated above is compared with the actual recorded monitoring height sequence point by point and time by time, and when the difference between the two exceeds the tolerance range, the position at that time is marked as abnormal. Through the analysis of the whole course data of a vehicle, the suspected abnormal point set found by the vehicle alone can be obtained, that is, the height difference information of the vehicle.

[0072] (3) Since the abnormal signal of a single vehicle can be caused by sensor error or temporary obstacles, the cloud server can aggregate the height difference information of all historical vehicles, and perform frequency statistics on the same geographical location. If a place is frequently reported as an abnormal point by multiple different vehicles, the probability of it being a fixed road defect is extremely high. Then, these high-frequency abnormal points are spatially clustered to outline the profile of the road surface abnormal area, effectively filtering out accidental interference.

[0073] (4) After locking the abnormal area of the road, the cloud server can call the complete vehicle driving data of the representative historical vehicles passing through the target area (including vehicle model parameters, vehicle speed sequence, and monitoring height sequence, etc.). Based on the principle of vehicle dynamics, reverse simulation is carried out, that is, it is tried to back-propagate a road surface geometry (profile model) that can make the vehicle of this type pass at a known speed and produce a height change curve completely consistent with the actual record. By solving this problem, it is finally determined whether the road obstacle is a pit or a bump, and the geometric parameters such as depth, area, and volume are quantified, so as to generate an accurate profile model.

[0074] The embodiment greatly eliminates single measurement error and temporary obstacle interference through the double verification mechanism of single vehicle comparison and multi-vehicle frequency statistics, greatly improves the accuracy and reliability of obstacle positioning; through the physical-based reverse simulation modeling, the accurate three-dimensional geometric characteristics of the obstacle can be inversely calculated from the vehicle group behavior data, which provides important quantitative input for subsequent vehicle individualized control (such as accurate speed limit or avoidance), thereby improving the safety and cooperative control accuracy of intelligent networked vehicles passing through abnormal road sections.

[0075] Further, on the basis of the above-mentioned embodiment of the application, based on the vehicle model parameters and vehicle trajectory sequence of each historical vehicle, in combination with the preset road reference model corresponding to the target area, the reference height sequence corresponding to each historical vehicle is determined, including:

[0076] Based on the vehicle model parameters of each historical vehicle, a corresponding vehicle three-dimensional model is matched from a preset vehicle three-dimensional model library;

[0077] The matched vehicle three-dimensional model is simulated in motion on the preset road reference model according to the vehicle trajectory sequence of each historical vehicle, to obtain driving posture data of each historical vehicle;

[0078] Based on the driving posture data, the theoretical height value of the monitoring point of each historical vehicle relative to the preset road reference model is determined to form the reference height sequence.

[0079] The preset vehicle three-dimensional model library can be a database pre-established to store digital three-dimensional models of various vehicle types, which can include information such as the appearance profile, internal structure, monitoring point position, and suspension system simulation parameters of the vehicle, and can be accurately matched through the vehicle model parameters.

[0080] The vehicle three-dimensional model can be a digital three-dimensional simulation model constructed based on specific vehicle model parameters, which can truly restore the physical structure and mechanical properties of the corresponding vehicle, and is used for subsequent motion state simulation in road scenes.

[0081] The theoretical height value can refer to a vertical distance of a monitoring point of a vehicle relative to a preset road reference model surface in an ideal scenario without road surface obstacles, which is calculated through motion simulation, and is a reference benchmark for judging whether a subsequent actual monitoring height is abnormal.

[0082] In the embodiment of the application, the process of determining the reference height sequence corresponding to each historical vehicle can include:

[0083] (1) Based on the vehicle model parameters (such as vehicle brand and model code) of each historical vehicle, a corresponding vehicle three-dimensional model is retrieved and matched in a preset vehicle three-dimensional model library.

[0084] (2) The preset road reference model corresponding to the target area is imported into the simulation system, and the vehicle three-dimensional models matched by the historical vehicles are loaded into the simulation system respectively; then, according to the vehicle trajectory sequence of the historical vehicles, the vehicle three-dimensional models in the simulation system are controlled to travel along the corresponding path of the preset road reference model, and the driving posture data of the vehicle at each trajectory coordinate point, such as the body posture angle, the suspension state, and the monitoring point space posture, are recorded in real time during the simulation process.

[0085] (3) Based on the obtained driving posture data, the vertical distances of each monitoring point (such as the sensor installation points at the four wheels) on the vehicle relative to the surface of the preset road reference model at each time are calculated, and these distance values are the theoretical height values; finally, the theoretical height values of all monitoring points are integrated in time sequence, thereby generating the reference height sequence corresponding to each historical vehicle.

[0086] The embodiment accurately matches the vehicle three-dimensional model, builds a motion simulation environment that fits the real scene, combines the time-aligned trajectory sequence and the preset road reference model, and generates the reference height sequence that accurately represents the ideal height state of the monitoring points of the historical vehicles in the scenario without road surface obstacles, and is accurately time-aligned with the trajectory sequence. This not only ensures the reference reliability of the reference height, but also improves the generality of the scheme by adapting to multiple vehicle types and complex road scenes, provides a high-quality reference benchmark for subsequent comparison with the actual monitoring height sequence, and thereby lays a core technical foundation for accurately identifying road surface obstacles in the target area.

[0087] Further, on the basis of the above-mentioned embodiment of the application, the reference height sequence of each historical vehicle is compared with the corresponding monitoring height sequence to obtain the height difference information of each historical vehicle, including:

[0088] For each historical vehicle, each trajectory position of the historical vehicle in the target area is traversed;

[0089] For the traversed trajectory position, the reference height and the monitoring height of each monitoring point at the trajectory position are extracted from the reference height sequence and the monitoring height sequence associated with the historical vehicle;

[0090] According to the height difference between the reference height and the monitoring height, the monitoring points are divided into abnormal point positions or normal point positions in combination with a preset height difference threshold value;

[0091] Based on the number of abnormal point positions and normal point positions, the abnormal probability of the trajectory position is determined;

[0092] If the abnormal probability is greater than or equal to a preset abnormal probability threshold value, the trajectory position is determined as a height difference trajectory position of the historical vehicle;

[0093] All height difference trajectory positions of the historical vehicle are summarized to form the height difference information.

[0094] In the embodiments of the present application, the determination process of the height difference information can include:

[0095] (1) For each historical vehicle passing through the target area, the complete driving trajectory thereof is discretized into a series of continuous trajectory positions, and each trajectory position is traversed; for the currently traversed trajectory position, the theoretical reference height of all monitoring points at the position is extracted from the reference height sequence, and the actual monitoring height of the corresponding monitoring points is extracted from the monitoring height sequence.

[0096] (2) The height difference (the difference between the reference height value and the monitoring height value) of each monitoring point is calculated, the height difference is compared with a preset height difference threshold value, the abnormal point positions (height difference ≥ threshold value) and the normal point positions (height difference < threshold value) are divided, and the number of the two types of point positions is recorded.

[0097] (3) Based on the number of abnormal point positions and the number of normal point positions of the current trajectory position, the proportion of abnormal point positions on the trajectory position is counted, and the abnormal probability is calculated, wherein the abnormal probability = abnormal point position number / (abnormal point position number + normal point position number).

[0098] (4) The calculated abnormal probability is compared with a preset abnormal probability threshold value, if it is greater than or equal to the preset abnormal probability threshold value, the current trajectory position is determined as a height difference trajectory position, and the associated information such as the spatial coordinates and the abnormal probability is recorded.

[0099] (5) The above process is repeated to complete the traversal of all trajectory positions of the historical vehicle in the target area, and all screened height difference trajectory positions are summarized to finally form the height difference information of the historical vehicle.

[0100] The embodiment effectively filters the instantaneous fluctuations caused by the slight bumps of the road surface or sensor noise by using the height difference threshold, reduces false positives, and through the abnormal probability index, the states of multiple monitoring points at a single trajectory position are integrated, false judgments caused by a single sensor failure are avoided, and the robustness of single vehicle anomaly detection is improved. Meanwhile, the height difference information output is no longer the original data stream, but refined semantic information representing "where the anomaly occurs", which greatly compresses the data volume, provides high-quality and high-credibility input for subsequent multi-vehicle data fusion and obstacle accurate positioning, and is a crucial data cleaning and feature extraction link in the entire cooperative perception process.

[0101] Further, on the basis of the above-mentioned embodiment of the application, based on the height difference information of each historical vehicle, the road surface abnormal area in the target area is determined by frequency statistics, including:

[0102] Summarize all height difference trajectory positions contained in the height difference information of each historical vehicle to obtain a set of de-duplicated height difference trajectory positions;

[0103] Statistical frequency of each geographical position being marked as a height difference trajectory position by different historical vehicles in the set of height difference trajectory positions;

[0104] Determine the geographical position with a frequency greater than or equal to a preset frequency threshold as a candidate abnormal position point.

[0105] Perform spatial clustering analysis on all candidate abnormal position points to obtain a road surface abnormal area.

[0106] The set of height difference trajectory positions can refer to a set formed by only retaining unique coordinate points after removing duplicate geographical position points after summarizing the height difference trajectory positions of all historical vehicles.

[0107] The candidate abnormal position point can refer to a geographical position point in the set of height difference trajectory positions, which is marked as a height difference trajectory position by different historical vehicles with a frequency greater than or equal to a preset frequency threshold.

[0108] The spatial clustering analysis can refer to a data analysis method for identifying continuous abnormal areas by grouping spatially adjacent candidate abnormal position points based on the spatial distance relationship of geographical position points.

[0109] In the embodiment of the application, the process of determining the road surface abnormal area can include:

[0110] (1) Collect the height difference information corresponding to all participating historical vehicles, extract all height difference trajectory positions of each historical vehicle in the target area from the height difference information, and form an initial difference position summary list.

[0111] (2) The highly different trajectory positions in the initial aggregated list are de-duplicated by comparing geographical position coordinates (such as longitude and latitude matching), and repeated coordinate points are removed, so that a highly different trajectory position set containing only unique geographical position points is finally obtained, and data redundancy is reduced.

[0112] (3) For each geographical position in the highly different trajectory position set, the number of times (i.e., the difference trigger number) that the position is marked as a highly different trajectory position by different historical vehicles is counted, and the difference trigger frequency (i.e., the frequency at which the geographical position is marked as a highly different trajectory position by different historical vehicles) of each position point is calculated according to the formula: abnormal trigger frequency = difference trigger number / total number of historical vehicles.

[0113] (4) The difference trigger frequency of each geographical position is compared with a preset frequency threshold, and geographical position points with a frequency greater than or equal to the threshold are selected and determined as candidate abnormal position points, so that accidental and single-vehicle or few-vehicle marked difference positions are excluded, and the screening reliability is ensured.

[0114] (5) Spatial clustering analysis (such as using clustering algorithms such as DBSCAN and K-means) is performed on all the selected candidate abnormal position points, and geographical positions that are close to each other are aggregated, so that one or more road surface abnormal regions in the target region are obtained.

[0115] The embodiment realizes high-robustness and high-precision collaborative detection of road obstacle regions by introducing the dual criteria of multi-vehicle high-frequency triggering and spatial clustering, and provides a solid and accurate data foundation for subsequent fine three-dimensional modeling and generation of safe vehicle control instructions.

[0116] Further, on the basis of the above-mentioned embodiment of the application, a profile model of the road obstacle is determined based on vehicle driving data associated with the road abnormal region, comprising:

[0117] Selecting at least one of the vehicle model parameters, vehicle speed sequence, and monitoring height sequence of a historical vehicle from the vehicle driving data associated with the road abnormal region;

[0118] Determining the wheel motion curve of the historical vehicle passing through the road abnormal region based on the monitoring height sequence;

[0119] Obtaining road inclination data corresponding to the road abnormal region;

[0120] Simulating the obstacle for the historical vehicle based on the vehicle model parameters, vehicle speed sequence, wheel motion curve, and road inclination data to obtain geometric parameters of the road obstacle; wherein the geometric parameters at least include obstacle extreme point height, obstacle extreme width point width, and obstacle extreme length point length.

[0121] The contour model of the road obstacle is constructed based on the geometric parameters, wherein the contour model comprises a concave contour model or a convex contour model.

[0122] The wheel movement curve can be a curve reflecting the change of the height of the historical vehicle wheel with time or driving distance in the process of passing through the road abnormal area, which is obtained based on the monitoring height sequence and through data processing, and can indirectly represent the road undulating state.

[0123] The road inclination angle data can be the inclination angle data of the road longitudinal direction (driving direction) and the road transverse direction (perpendicular to the driving direction) corresponding to the road abnormal area, and is used to restore the real road posture, and can be obtained from a road basic database, a high-precision map and the like.

[0124] The obstacle simulation can be a process of simulating the real process of the historical vehicle passing through the road obstacle by using a simulation tool or an algorithm, inputting the vehicle model parameters, the vehicle speed sequence, the wheel movement curve and the road inclination angle data, and then backstepping the geometric parameters of the obstacle.

[0125] The geometric parameters of the road obstacle can be core parameters for describing the spatial form of the road obstacle, and are the basis for constructing the contour model, and can at least include an obstacle extreme point height (the vertical distance between the highest point or the lowest point and the reference road surface), an obstacle extreme width point width (the maximum transverse span of the obstacle) and an obstacle extreme length point length (the maximum longitudinal span of the obstacle).

[0126] The concave contour model can be a contour model constructed based on the geometric parameters of the concave obstacle and representing the spatial form of the concave area, which can be used to describe the road pit, the concave and the like.

[0127] The convex contour model can be a contour model constructed based on the geometric parameters of the convex obstacle and representing the spatial form of the convex area, which can be used to describe the road bump, the speed reduction zone and the like.

[0128] In the embodiment of the application, the process of constructing the contour model of the road obstacle can comprise:

[0129] (1) at least one historical vehicle which has passed through the target area can be selected as a simulation vehicle, and the associated vehicle model parameters, the vehicle speed sequence and the monitoring height sequence of the simulation vehicle can be obtained.

[0130] (2) the monitoring height sequence of the simulation vehicle when passing through the abnormal area is analyzed in depth, and the wheel movement curve which can represent the continuous change of the wheel relative to the reference height of the road surface in the whole process of passing through the obstacle is fitted or reconstructed by processing the discrete height sampling points in the sequence, and the curve can be used to reflect the bumping response of the vehicle to the unknown obstacle.

[0131] (3) From the high-precision map or road database, obtain the road inclination data associated with the road anomaly area, such as longitudinal slope, transverse inclination, etc.

[0132] (4) The selected historical vehicle model parameters (restore vehicle physical characteristics), vehicle speed sequence (restore driving dynamics), wheel motion curve (restore the interaction between the wheel and the road), and road inclination data (restore the real road posture) are input into the pre-configured simulation system to simulate the whole process of the vehicle passing through the road anomaly area. The shape and size of a hypothetical virtual obstacle are continuously adjusted through simulation until the virtual vehicle in the simulation system passes through at the same speed, and the motion curve generated by the wheel is maximally consistent with the real wheel motion curve. When the simulation wheel motion curve matching degree reaches the optimal, the key geometric parameters of the virtual obstacle are recorded, including but not limited to obstacle extreme point height, obstacle extreme width point width, and obstacle extreme length point length.

[0133] (5) Using the geometric parameters solved in the previous step as constraint conditions, a continuous and computable contour model is generated through three-dimensional geometric modeling methods such as parametric surface fitting (for concave surfaces) or voxel modeling (for protrusions), which can include concave contour models or convex contour models, thereby completing the digital definition of the road obstacle from data to visualized and quantifiable three-dimensional entity.

[0134] This embodiment realizes reliable inversion from indirect and noisy vehicle motion signals to direct and accurate road obstacle three-dimensional geometric model, overcoming the technical limitations of traditional visual sensors being easily affected by shielding and environment; by fusing vehicle specific parameters and road environment information for physical simulation, the contour model constructed has high physical reality and accuracy, providing reliable input for subsequent safety evaluation; at the same time, the model represents the obstacle in a standardized three-dimensional digital form, making it possible for different vehicle types to pass through the simulation, risk quantification evaluation, and personalized control strategy generation, fundamentally improving the perception accuracy and safety upper limit of the cooperative control of the unmanned driving system for complex road conditions.

[0135] In a specific embodiment, in order to overcome the problem of slow optimization convergence and high computational cost that may exist in the inversion of obstacle shape in traditional motion simulation, the embodiment proposes a fast obstacle modeling method based on neural network. Specifically, a recommended obstacle curve component can be constructed and trained on a cloud server, the core of which is a long short-term memory neural network (LSTM). The network takes the wheel motion curve, vehicle speed time series information, road inclination data, and vehicle type parameters of a single historical vehicle passing through a suspected obstacle region as input features, and directly outputs the predicted obstacle height coordinate sequence along the vehicle driving direction through supervised learning. During training, a first loss function (mean deviation of predicted height sequence and true height sequence) and a second loss function (deviation variance of predicted height sequence) are designed to jointly constrain the network. When the two loss functions are both stable below the preset threshold in continuous multiple training, it is determined that the model has converged and can be used. In actual application, the historical vehicle data to be analyzed can be input into the trained component to quickly obtain a recommended obstacle curve that fits the vehicle response, greatly speeding up the subsequent analysis process. Finally, the system integrates the analysis results of multiple vehicles to extract the extreme point height, extreme width width, and extreme length length of the obstacle as three core geometric labels, and constructs a final concave or convex profile model accordingly, thereby efficiently and accurately completing the three-dimensional digital representation of the road obstacle.

[0136] Further, on the basis of the above-mentioned embodiment of the application, a vehicle control instruction is generated based on the road obstacle and the vehicle model parameter of the target vehicle, comprising:

[0137] determining whether the severity of the road obstacle is greater than or equal to a preset avoidance threshold;

[0138] if yes, generating a vehicle control instruction containing a driving avoidance identifier;

[0139] if no, performing a safe passage simulation based on the vehicle model parameter of the target vehicle and the profile model of the road obstacle, determining a safe passage speed, and generating a vehicle control instruction containing a driving speed threshold identifier corresponding to the safe passage speed.

[0140] The preset avoidance threshold can be a pre-set critical standard for determining whether the vehicle needs to take evasive driving.

[0141] The safe passage simulation can be a virtual test process of simulating the target vehicle passing through the digital road obstacle profile model at a specific speed using a computer vehicle dynamics model, aiming to predict the dynamic response of the vehicle and evaluate the passing safety.

[0142] The safe passage speed can be the maximum driving speed of the target vehicle that can safely pass through the road obstacle region verified in the safe passage simulation.

[0143] The driving speed threshold identification can be identification information embedded in the vehicle control instruction, which informs the target vehicle of a safe passing speed limit to be followed.

[0144] In the embodiments of the present application, the generation process of the vehicle control instruction can include:

[0145] (1) After obtaining the accurate road obstacle model (profile model) of the target road section, the cloud server can quantitatively evaluate the severity thereof, for example, by comparing the key geometric parameters of the obstacle (such as the area of a pit or the volume of a protrusion) with a pre-set safety threshold (avoidance threshold) to determine it.

[0146] (2) If the severity of the road obstacle is greater than or equal to the pre-set avoidance threshold, the system determines that the obstacle risk is too high, and direct passing at any speed will pose an unacceptable safety risk, therefore, the system will generate a vehicle control instruction containing a driving avoidance identification.

[0147] (3) If the severity of the road obstacle is less than the pre-set avoidance threshold, the system determines a maximum driving speed at which the target vehicle can safely pass through the road obstacle area as a safe passing speed according to the vehicle model parameters of the target vehicle and the profile model of the road obstacle, and generates a vehicle control instruction containing a driving speed threshold identification corresponding to the safe passing speed.

[0148] This embodiment combines the objective severity of the obstacle with the specific performance parameters of the vehicle, avoiding one-size-fits-all conservative control (such as detouring or passing at extremely low speed for all obstacles), thereby maximizing the passing efficiency of the road under the premise of absolute safety; for high-risk obstacles, the avoidance strategy is adopted decisively, fundamentally eliminating accidents; for passable obstacles, the highest safe speed that can be tolerated by different vehicle models is customized, ensuring that different vehicles such as sedans, SUVs, and trucks can pass safely, avoiding unnecessary passing delay, and improving the intelligentization and collaboration level of the overall traffic flow.

[0149] Further, on the basis of the above-mentioned embodiments of the present application, determining whether the severity of the road obstacle is greater than or equal to the pre-set avoidance threshold comprises:

[0150] If the profile model of the road obstacle is a concave profile model, the concave area of the concave profile model is compared with a pre-set concave area threshold;

[0151] If the profile model of the road obstacle is a convex profile model, the convex volume of the convex profile model is compared with a pre-set convex volume threshold;

[0152] When the notch area is greater than or equal to a preset notch area threshold, or the protrusion volume is greater than or equal to a preset protrusion volume threshold, it is determined that the severity of the road obstacle is greater than or equal to a preset avoidance threshold.

[0153] The notch area can be an area of a recessed region in the road surface horizontal projection plane in the concave profile model, and is a core parameter for quantifying the severity of the concave obstacle. The protrusion volume can be a space volume occupied by a protrusion region in the convex profile model, and is a core parameter for quantifying the severity of the convex obstacle.

[0154] In the embodiment of the application, the process of judging the severity of the road obstacle can include:

[0155] (1) First, identify whether the constructed road obstacle profile model is a concave profile model describing a road surface depression or a convex profile model describing a road surface protrusion; after determining the type of the profile model, extract the key parameter for evaluating the severity of the road obstacle from the corresponding profile model, such as the notch area for the concave profile model, and the protrusion volume for the convex profile model.

[0156] (2) Compare the notch area and the protrusion volume with the corresponding preset notch area threshold and the preset protrusion volume threshold, respectively; when the notch area is greater than or equal to the threshold, or the protrusion volume is greater than or equal to the threshold, it is determined that the severity of the road obstacle has reached or exceeded the preset avoidance threshold.

[0157] The embodiment realizes accurate and rapid classification of the road obstacle risk by automatically comparing the key geometric dimensions (notch area, protrusion volume) of the obstacle with the preset safety threshold, and ensures that the subsequent vehicle can adopt a reasonable passing strategy according to the severity of the obstacle.

[0158] Further, on the basis of the above-mentioned embodiment of the application, a safety passing simulation is performed based on the vehicle model parameter of the target vehicle and the profile model of the road obstacle to determine a safety passing speed, including:

[0159] Based on the vehicle model parameter of the target vehicle, a preset driving speed set containing a plurality of candidate speeds is determined;

[0160] For each candidate speed in the preset driving speed set, the target vehicle is simulated and tested based on the profile model;

[0161] Collect driving samples in the simulation test, and count the proportion of accident samples;

[0162] From all the candidate speeds with an accident sample proportion less than or equal to a preset accident sample proportion threshold, the maximum speed value is selected as the safety passing speed.

[0163] The preset driving speed set can be a set containing a plurality of different speed values, which is preset based on the driving performance of the target vehicle, road speed limit standards, common driving conditions and the like, and is used as a candidate range of safe passing speed. For example, for a car passing a concave obstacle, the corresponding preset driving speed set can be {30 km / h, 40 km / h, 50 km / h}; for a truck passing a convex obstacle, the corresponding preset driving speed set can be {20 km / h, 25 km / h, 30 km / h}.

[0164] The simulation test can be a test process of collecting vehicle driving state data (such as vehicle body posture, wheel force, braking state) in real time when the target vehicle passes the road obstacle at a certain candidate speed in the safe passing simulation environment, and judging whether a safety accident occurs.

[0165] The driving sample can be a complete data set related to the driving of the target vehicle recorded in a single simulation test process, including input parameters (candidate speed, vehicle model parameters), process data (driving trajectory, vehicle body posture), output results (whether an accident occurs) and the like.

[0166] The accident sample proportion can be a proportion of the number of driving samples in which a safety accident such as collision, rollover or bottoming occurs to the total number of test samples in a plurality of simulation tests carried out for the same candidate speed.

[0167] In the embodiment of the application, the determination process of the safe passing speed can include:

[0168] (1) A preset driving speed set containing a plurality of candidate speeds is automatically generated according to the vehicle model parameters (such as suspension stiffness, mass, wheelbase, minimum ground clearance and the like) of the target vehicle, which covers a reasonable range from low-speed cautious passing to high-speed passing, and provides a speed benchmark for subsequent simulation tests.

[0169] (2) Based on the profile model (precise concave or convex three-dimensional geometric model) of the road obstacle and the dynamics model of the target vehicle, the complete process of the vehicle passing the obstacle at the candidate speed is simulated, and the process is simulated a plurality of times (such as hundreds or thousands of times) to collect sufficient driving samples, which cover different dynamic responses that the vehicle can have.

[0170] (3) The safety of each driving sample generated by simulation is evaluated according to the preset accident judgment rule (for example, suspension compression to the limit, wheel separation from the road, chassis scratching and the like); the number of samples judged as “accident” in all simulation samples at a certain candidate speed is counted, and the accident sample proportion (i.e. accident rate) corresponding to the candidate speed is calculated.

[0171] (4) comparing the accident sample proportion of all candidate speeds with a preset accident sample proportion threshold, screening out safe candidate speeds with an accident sample proportion less than or equal to the threshold, and selecting a maximum speed value from the safe speeds to determine a safe passing speed recommended to the target vehicle, so as to maximize the passing efficiency under the premise of ensuring driving safety.

[0172] The embodiment can not only provide highly personalized and accurate speed limits for different types of vehicles when facing different shapes of obstacles, ensuring one vehicle one strategy safety, but more importantly, it actively excavates and recommends the highest passing efficiency under the current conditions by screening logic of taking the maximum value of safe speeds, thus essentially solving the core control problem of "how fast to drive when passing a risk section without being too conservative", and achieving intelligent balance between safety and passing efficiency.

[0173] Embodiment Two

[0174] Figure 2 A structural schematic diagram of a vehicle control system according to Embodiment Two of the present application is provided. As shown in the figure, the system is configured in a cloud server, comprising: Figure 2

[0175] A data acquisition module 21, configured to acquire vehicle driving data of multiple historical vehicles when passing a target region in response to a passing request of a target vehicle to the target region;

[0176] An obstacle determination module 22, configured to determine a road obstacle existing in the target region based on the vehicle driving data;

[0177] An instruction generation module 23, configured to generate a vehicle control instruction based on the road obstacle and a vehicle model parameter of the target vehicle;

[0178] An instruction issuing module 24, configured to issue the vehicle control instruction to the target vehicle, so that the target vehicle executes passing control based on the vehicle control instruction.

[0179] Further, on the basis of the above-mentioned embodiment of the application, the data acquisition module 21 comprises:

[0180] A data acquisition unit, configured to acquire vehicle driving data of multiple historical vehicles passing a target region within a preset historical period according to position information of the target region carried in a passing request; wherein the vehicle driving data at least includes a vehicle model parameter, a vehicle trajectory sequence, a vehicle speed sequence and a monitoring height sequence of a vehicle monitoring point relative to a road surface of each historical vehicle.

[0181] ​Furthermore, based on the above embodiments of the invention, the obstacle determination module 22 includes:

[0182] The reference height sequence determination unit is used to determine the reference height sequence corresponding to each historical vehicle based on the vehicle model parameters and vehicle trajectory sequence of each historical vehicle, combined with the preset road reference model corresponding to the target area.

[0183] The height difference information determination unit is used to compare the reference height sequence of each historical vehicle with the corresponding monitoring height sequence to obtain the height difference information of each historical vehicle.

[0184] The road surface anomaly area determination unit is used to determine the road surface anomaly area within the target area based on the height difference information of each historical vehicle and through frequency statistics.

[0185] The contour model determination unit is used to determine the contour model of road obstacles based on vehicle driving data associated with abnormal road areas.

[0186] Furthermore, based on the above embodiments of the invention, the reference height sequence determination unit is specifically used for:

[0187] Based on the vehicle model parameters of each historical vehicle, the corresponding vehicle 3D model is matched from the preset vehicle 3D model library.

[0188] The matched 3D vehicle models are used to perform motion simulation on a preset road reference model according to the vehicle trajectory sequence of each historical vehicle to obtain the driving posture data of each historical vehicle.

[0189] Based on driving posture data, the theoretical height values ​​of the monitoring points of each historical vehicle relative to the preset road reference model are determined to form a reference height sequence.

[0190] Furthermore, based on the above embodiments of the invention, the height difference information determination unit is specifically used for:

[0191] For each historical vehicle, iterate through every trajectory position of the historical vehicle within the target area;

[0192] For the traversed trajectory locations, extract the reference height and monitoring height of each monitoring point under the trajectory location from the reference height sequence and monitoring height sequence associated with historical vehicles;

[0193] Based on the height difference between the reference height and the monitoring height, and combined with the preset height difference threshold, the monitoring points are divided into abnormal points or normal points.

[0194] Based on the number of abnormal and normal points, the probability of anomalies in the trajectory location is determined.

[0195] determine the track position as the height difference track position of the historical vehicle if the abnormal probability is greater than or equal to a preset abnormal probability threshold;

[0196] aggregate all height difference track positions of the historical vehicles to form the height difference information.

[0197] Further, on the basis of the above-mentioned embodiments of the application, the road surface anomaly area determination unit is specifically configured to:

[0198] aggregate all height difference track positions contained in the height difference information of each historical vehicle to obtain a set of de-duplicated height difference track positions;

[0199] count a frequency at which each geographical position in the set of height difference track positions is marked as a height difference track position by different historical vehicles;

[0200] determine a geographical position whose frequency is greater than or equal to a preset frequency threshold as a candidate abnormal position point;

[0201] perform spatial clustering analysis on all candidate abnormal position points to obtain a road surface anomaly area.

[0202] Further, on the basis of the above-mentioned embodiments of the application, the contour model determination unit is specifically configured to:

[0203] select a vehicle model parameter, a vehicle speed sequence and a monitoring height sequence of at least one historical vehicle from vehicle driving data associated with the road surface anomaly area;

[0204] determine a wheel motion curve of the historical vehicle passing through the road surface anomaly area based on the monitoring height sequence;

[0205] obtain road inclination data corresponding to the road surface anomaly area;

[0206] simulate an obstacle for the historical vehicle based on the vehicle model parameter, the vehicle speed sequence, the wheel motion curve and the road inclination data to obtain geometric parameters of the road surface obstacle; wherein the geometric parameters at least include an obstacle extreme point height, an obstacle extreme width point width and an obstacle extreme length point length;

[0207] construct a contour model of the road surface obstacle based on the geometric parameters; wherein the contour model includes a concave contour model or a convex contour model.

[0208] Further, on the basis of the above-mentioned embodiments of the application, the instruction generation module 23 includes:

[0209] an obstacle judgment unit configured to judge whether a severity of the road surface obstacle is greater than or equal to a preset avoidance threshold;

[0210] a first instruction generation unit configured to generate a vehicle control instruction containing a driving avoidance identifier if the severity of the road surface obstacle is greater than or equal to the preset avoidance threshold.

[0211] The second instruction generation unit is configured to, if no, perform a safe passing simulation based on a vehicle model parameter of the target vehicle and a profile model of the road obstacle, determine a safe passing speed, and generate a vehicle control instruction containing a driving speed threshold corresponding to the safe passing speed.

[0212] Further, on the basis of the above-mentioned embodiments, the obstacle judgment unit is specifically configured to:

[0213] If the profile model of the road obstacle is a concave profile model, the concave area of the concave profile model is compared with a preset concave area threshold;

[0214] If the profile model of the road obstacle is a convex profile model, the convex volume of the convex profile model is compared with a preset convex volume threshold;

[0215] When the concave area is greater than or equal to the preset concave area threshold, or the convex volume is greater than or equal to the preset convex volume threshold, it is determined that the severity of the road obstacle is greater than or equal to a preset avoidance threshold.

[0216] Further, on the basis of the above-mentioned embodiments, the second instruction generation unit performs a safe passing simulation based on a vehicle model parameter of the target vehicle and a profile model of the road obstacle to determine a safe passing speed, which includes:

[0217] Based on the vehicle model parameter of the target vehicle, a preset driving speed set containing a plurality of candidate speeds is determined;

[0218] For each candidate speed in the preset driving speed set, the target vehicle is simulated and tested based on the profile model;

[0219] Driving samples in the simulation and testing are collected, and the proportion of accident samples is counted;

[0220] From all the candidate speeds with an accident sample proportion less than or equal to a preset accident sample proportion threshold, the maximum speed value is selected as the safe passing speed.

[0221] The vehicle control system provided in the embodiments can execute the vehicle control method provided in any of the embodiments, and has the corresponding functional modules and beneficial effects of the execution method.

[0222] Embodiment three

[0223] Figure 3A schematic diagram of an electronic device 30 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0224] like Figure 3 As shown, the electronic device 30 includes at least one processor 31 and a memory, such as a read-only memory (ROM) 32 or a random access memory (RAM) 33, communicatively connected to the at least one processor 31. The memory stores computer programs executable by the at least one processor. The processor 31 can perform various appropriate actions and processes based on the computer program stored in the ROM 32 or loaded from storage unit 38 into the RAM 33. The RAM 33 can also store various programs and data required for the operation of the electronic device 30. The processor 31, ROM 32, and RAM 33 are interconnected via a bus 34. An input / output (I / O) interface 35 is also connected to the bus 34.

[0225] Multiple components in electronic device 30 are connected to I / O interface 35, including: input unit 36, such as keyboard, mouse, etc.; output unit 37, such as various types of monitors, speakers, etc.; storage unit 38, such as disk, optical disk, etc.; and communication unit 39, such as network card, modem, wireless transceiver, etc. Communication unit 39 allows electronic device 30 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0226] Processor 31 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 31 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 31 performs the various methods and processes described above, such as vehicle control methods.

[0227] In some embodiments, the vehicle control method can be implemented as a computer program tangibly embodied in a computer readable storage medium, e.g., storage unit 38. In some embodiments, parts or all of the computer program can be loaded and / or installed onto electronic device 30 via, e.g., ROM 32 and / or communication unit 39. When the computer program is loaded onto RAM 33 and executed by processor 31, one or more steps of the vehicle control method described above can be performed. Alternatively, in other embodiments, processor 31 can be configured to perform the vehicle control method by way of other means (e.g., by way of firmware).

[0228] Various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, specially designed application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0229] In some embodiments, the vehicle control method can be implemented as a computer program tangibly embodied in a computer program product, the computer program implementing the vehicle control method of the present invention when executed by a processor, the computer program product can be understood as a software product that mainly realizes its solution by means of the computer program. The computer program for implementing the method of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the computer program, when executed by the processor, implements the functions / operations specified in the flow charts and / or the block diagrams. The computer program can be executed entirely on a machine, partially on a machine, partially on a machine as a standalone software package and partially on a remote machine, or entirely on a remote machine or server.

[0230] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. A computer-readable storage medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of a machine-readable storage medium will include one or more lines of a program of instructions in a transitory signal, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0231] To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the electronic device. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0232] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0233] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system, to solve the defects of large management difficulty and weak business scalability in traditional physical host and VPS service.

[0234] It should be understood that the various forms of flow shown above can be reordered, added to, or have steps deleted. For example, the steps described in the present application can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions of the present application can be achieved, which are not limited herein.

[0235] The above detailed description does not constitute a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A vehicle control method, characterized in that, Applied to a cloud server, the method includes: In response to a target vehicle's request to pass through a target area, vehicle driving data of multiple historical vehicles passing through the target area is obtained. Based on the vehicle driving data, road obstacles existing in the target area are identified; Based on the road obstacles and the vehicle model parameters of the target vehicle, generate vehicle control commands; The vehicle control command is sent to the target vehicle so that the target vehicle can perform traffic control based on the vehicle control command.

2. The method according to claim 1, characterized in that, In response to a target vehicle's request to pass through a target area, the method of acquiring vehicle driving data from multiple historical vehicles passing through the target area includes: According to the location information of the target area carried in the passage request, obtain the vehicle driving data of multiple historical vehicles that passed through the target area within a preset historical time period; wherein, the vehicle driving data includes at least the vehicle model parameters, vehicle trajectory sequence, vehicle speed sequence, and vehicle monitoring point monitoring height sequence relative to the road surface of each historical vehicle.

3. The method according to claim 2, characterized in that, The step of determining road obstacles existing in the target area based on the vehicle driving data includes: Based on the vehicle model parameters and vehicle trajectory sequence of each historical vehicle, and in conjunction with the preset road benchmark model corresponding to the target area, the benchmark height sequence corresponding to each historical vehicle is determined. The reference height sequence of each historical vehicle is compared with the corresponding monitoring height sequence to obtain the height difference information of each historical vehicle. Based on the height difference information of each historical vehicle, abnormal road surface areas within the target area are determined through frequency statistics; Based on the vehicle driving data associated with the abnormal road surface area, the contour model of the road obstacle is determined.

4. The method according to claim 3, characterized in that, The determination of the reference height sequence for each historical vehicle, based on the vehicle model parameters and vehicle trajectory sequence of each historical vehicle, and in conjunction with the preset road reference model corresponding to the target area, includes: Based on the vehicle model parameters of each of the historical vehicles, the corresponding vehicle 3D model is matched from the preset vehicle 3D model library; The matched 3D vehicle models are subjected to motion simulation on the preset road reference model according to the vehicle trajectory sequence of each historical vehicle to obtain the driving posture data of each historical vehicle. Based on the driving posture data, the theoretical height values ​​of the monitoring points of each historical vehicle relative to the preset road reference model are determined to form the reference height sequence.

5. The method according to claim 3, characterized in that, The step of comparing the reference height sequence of each historical vehicle with the corresponding monitoring height sequence to obtain the height difference information of each historical vehicle includes: For each of the historical vehicles, iterate through each trajectory position of the historical vehicle within the target area; For the traversed trajectory locations, extract the reference height and monitoring height of each monitoring point at the trajectory location from the reference height sequence and the monitoring height sequence associated with the historical vehicles; Based on the height difference between the reference height and the monitoring height, and combined with a preset height difference threshold, the monitoring points are divided into abnormal points or normal points. Based on the number of abnormal points and the number of normal points, the probability of anomaly in the trajectory position is determined; If the anomaly probability is greater than or equal to a preset anomaly probability threshold, then the trajectory position is determined as the height difference trajectory position of the historical vehicle. All the height difference trajectory locations of the historical vehicles are aggregated to form the height difference information.

6. The method according to claim 3, characterized in that, The step of determining abnormal road surface areas within the target area based on the height difference information of each historical vehicle through frequency statistics includes: By summing up all the height difference trajectory positions contained in the height difference information of each of the historical vehicles, a set of deduplicated height difference trajectory positions is obtained. Statistically determine the frequency with which each geographical location in the set of height difference trajectory locations is marked as a height difference trajectory location by different historical vehicles; Geographic locations with frequencies greater than or equal to a preset frequency threshold are identified as candidate anomaly locations. Spatial clustering analysis is performed on all the candidate anomaly locations to obtain the road surface anomaly areas.

7. The method according to claim 3, characterized in that, Determining the contour model of the road obstacle based on the vehicle driving data associated with the abnormal road surface area includes: From the vehicle driving data associated with the abnormal road surface area, select at least one of the historical vehicle's vehicle model parameters, vehicle speed sequence, and monitoring height sequence; The wheel motion curves of the historical vehicles passing through the abnormal road surface area are determined based on the monitoring height sequence. Obtain the road slope data corresponding to the abnormal road surface area; Based on the vehicle model parameters, the vehicle speed sequence, the wheel motion curve, and the road tilt angle data, obstacle simulation is performed on the historical vehicle to obtain the geometric parameters of the road obstacle; wherein, the geometric parameters include at least the height of the obstacle extreme point, the width of the obstacle extreme point, and the length of the obstacle extreme point; A contour model of the road surface obstacle is constructed based on the geometric parameters; wherein, the contour model includes a concave contour model or a convex contour model.

8. The method according to claim 1, characterized in that, The process of generating vehicle control commands based on the road obstacle and the vehicle model parameters of the target vehicle includes: Determine whether the severity of the road obstacle is greater than or equal to a preset avoidance threshold; If so, then generate the vehicle control command containing the driving avoidance indicator; If not, a safe passage simulation is performed based on the vehicle model parameters of the target vehicle and the contour model of the road obstacle to determine the safe passage speed and generate the vehicle control command containing the driving speed threshold identifier corresponding to the safe passage speed.

9. The method according to claim 8, characterized in that, The step of determining whether the severity of the road obstacle is greater than or equal to a preset avoidance threshold includes: If the contour model of the road obstacle is a concave contour model, then the area of ​​the concave contour model is compared with a preset concave area threshold. If the contour model of the road obstacle is a raised contour model, then the raised volume of the raised contour model is compared with a preset raised volume threshold. When the area of ​​the notch is greater than or equal to the preset notch area threshold, or the volume of the protrusion is greater than or equal to the preset protrusion volume threshold, the severity of the road obstacle is determined to be greater than or equal to the preset avoidance threshold.

10. The method according to claim 8, characterized in that, The step of performing a safe passage simulation based on the vehicle model parameters of the target vehicle and the contour model of the road obstacle to determine the safe passage speed includes: Based on the vehicle model parameters of the target vehicle, a preset driving speed set containing multiple candidate speeds is determined; For each candidate speed in the preset driving speed set, the target vehicle is simulated and tested based on the contour model; Collect driving samples from the simulation test and calculate the percentage of accident samples; The maximum speed value is selected as the safe passage speed from the candidate speeds where the proportion of all accident samples is less than or equal to a preset accident sample proportion threshold.

11. A vehicle control system, characterized in that, The system, configured on a cloud server, includes: The data acquisition module is used to acquire vehicle driving data of multiple historical vehicles passing through the target area in response to the target vehicle's request for passage through the target area; An obstacle determination module is used to determine road obstacles existing in the target area based on the vehicle driving data. The instruction generation module is used to generate vehicle control instructions based on the road obstacle and the vehicle model parameters of the target vehicle; The instruction issuing module is used to issue the vehicle control instruction to the target vehicle, so that the target vehicle can perform traffic control based on the vehicle control instruction.

12. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the vehicle control method according to any one of claims 1-10.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed by a processor, implement the vehicle control method according to any one of claims 1-10.

14. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the vehicle control method according to any one of claims 1-10.

Images