A big data-based automatic driving vehicle collision warning method and system

Abstract

The application is suitable for the field of automatic driving technology, and provides a collision warning method and system for an automatic driving vehicle based on big data, which comprises the following steps: obtaining driving information of a target vehicle, obtaining driving information of surrounding vehicles in a preset space-time range with the target vehicle, and all the driving information forming a warning data set; the driving information at least comprises identification information, position, speed and acceleration data of the vehicle; calculating the difference between the acceleration of all the surrounding vehicles in the warning data set and a reference acceleration; when the value of one or more of the differences is greater than a first acceleration threshold, obtaining the position, speed and acceleration data of the vehicle corresponding to the one or more differences, and combining the position, speed and acceleration data of the vehicle with the position, speed and acceleration data of the target vehicle to form a first data set; the application increases a pre-judgment mechanism, further deeply processes corresponding data according to requirements based on big data, realizes collision warning, and takes into account the cruising range.

Description

Technical Field

[0001] This invention belongs to the field of autonomous driving technology, and in particular relates to a collision warning method and system for autonomous vehicles based on big data. Background Technology

[0002] According to official data from the China Association of Automobile Manufacturers, as of the end of March 2022, the total number of motor vehicles in China reached 402 million, with 307 million being automobiles, accounting for 76.37% of the total. The number of new energy vehicles reached 8.915 million, representing 2.90% of the total number of automobiles. Autonomous vehicles on the market are primarily new energy vehicles, and as a replacement for traditional motor vehicles, new energy vehicles have broad market prospects and enormous development value.

[0003] Autonomous vehicles, according to industry standards, can be broadly classified into Level 1 to Level 5. This classification is widely accepted by car manufacturers in the industry. Therefore, continuously improving the automation and intelligence level of autonomous vehicles is a goal that people in this field are constantly pursuing. Combining vehicle control with new-generation information technologies such as big data and AI will greatly help advance autonomous driving technology and realize the market promotion of high-level autonomous driving.

[0004] More specifically, most car companies are currently developing and manufacturing vehicles with Level 2 autonomous driving capabilities. Level 2 autonomous driving can automatically perform steering and acceleration / deceleration during driving. Of course, during this process, the driver needs to remain focused, monitor the surrounding environment in real time, and be prepared to take over the vehicle at any time. A few car companies are mass-producing vehicles that meet Level 3 autonomous driving capabilities (conditional automation). The upgrade from Level 2 to Level 3 autonomous driving is considered a major leap in automotive intelligence, because from Level 3 onwards, the car itself can monitor the surrounding environment in real time using sensors such as lidar. However, this level still requires driver control, and the driver must remain focused at all times and take over the vehicle when the system cannot perform its tasks. As for Level 4 autonomous driving (high automation): Compared to Level 3, Level 4 autonomous driving can complete all tasks and does not require human control in most situations. Of course, the driver can still choose to operate the vehicle manually. Level 5 autonomous driving (full automation): Full automation means that the car does not require human intervention. The system can control all key personnel, monitor the surrounding environment, and recognize unique driving conditions. It is still in the experimental stage.

[0005] As mentioned above, the integration of vehicle control with next-generation information technologies such as big data and AI will bring about innovation in data processing, but it will also place an additional burden on the vehicle's battery range. Currently, vehicles with Level 3 and Level 4 driving automation can only achieve closed driving on specific road surfaces and do not yet have the mass production capabilities for regular roads. In order to realize the possibility of mass production or marketability of high-level driving automation vehicles, while also addressing the range anxiety brought about by big data technology processing, it is necessary to develop a collision warning method and system for autonomous vehicles based on big data. Summary of the Invention

[0006] The purpose of this invention is to provide a collision warning method and system for autonomous vehicles based on big data, aiming to increase the possibility of mass production or marketization of existing high-level autonomous vehicles.

[0007] This invention is implemented as follows: a collision warning method for autonomous vehicles based on big data, the method comprising:

[0008] The system acquires the driving information of the target vehicle and the driving information of surrounding vehicles within a preset time and space range. All driving information is used to form a warning dataset. The driving information includes at least the vehicle's identification information, location, speed, and acceleration data.

[0009] Calculate the difference between the acceleration of all surrounding vehicles in the warning dataset and the reference acceleration;

[0010] When one or more of the differences are greater than the first acceleration threshold, the position, speed, and acceleration data of the vehicle corresponding to the one or more differences are obtained and combined with the position, speed, and acceleration data of the target vehicle to form the first dataset;

[0011] Based on the first dataset, the trend of the distance between the target vehicle and the corresponding vehicle within a certain period of time after the current time node is inferred, and the target vehicle is determined to enter the collision warning mode based on the trend of the distance change, and a collision warning signal is issued in a timely manner.

[0012] When all the differences are less than or equal to the first acceleration threshold, the target vehicle remains in autonomous driving mode.

[0013] Preferably, the step of acquiring the target vehicle's driving information and acquiring the driving information of surrounding vehicles within a preset spatiotemporal range, with all driving information forming a warning dataset, specifically includes the following steps:

[0014] The time and space zoning information of the spatiotemporal range is set with the target vehicle as the origin, and iteratively updated with the real-time location of the target vehicle;

[0015] Obtain the navigation information of the target vehicle, obtain the driving information of the target vehicle, and correct the driving information of the target vehicle using the navigation information of the target vehicle;

[0016] Based on the set time and space range, obtain the driving information of surrounding vehicles within the preset time and space range of the target vehicle;

[0017] Classify the driving information of the target vehicle and the driving information of surrounding vehicles;

[0018] The classified driving information is normalized to form an early warning dataset.

[0019] Preferably, the method further includes:

[0020] Based on the navigation information of the target vehicle, the driving route of the target vehicle is divided into a driving link consisting of several nodes; wherein several nodes correspond to auxiliary facilities on the current road in the navigation map;

[0021] Based on the aforementioned driving link, the frequency of obtaining driving information at node locations is increased.

[0022] Preferably, after dividing the target vehicle's travel route into a travel link consisting of several nodes based on the target vehicle's navigation information, the travel link is optimized, specifically including:

[0023] Based on big data comparison of changes in auxiliary facilities, the corresponding nodes are adjusted accordingly.

[0024] Based on the number of nodes within two adjacent distances in a specified distance segment, increase transition points proportionally along the direction of increasing nodes on the driving link;

[0025] Adjust the frequency of acquiring driving information based on the aforementioned transition point location.

[0026] Preferably, the step of calculating the difference between the acceleration of all surrounding vehicles in the early warning dataset and the reference acceleration specifically includes:

[0027] Select the initial reference acceleration based on the target vehicle's historical driving data;

[0028] Based on the target vehicle's current speed data, the initial reference acceleration is increased to determine the optimal reference acceleration;

[0029] Calculate the difference between the acceleration of vehicles in front and behind in the warning data and the reference acceleration;

[0030] Calculate the difference between the acceleration of vehicles on the side of the warning data and the reference acceleration.

[0031] Preferably, the step of inferring the trend of distance changes between the target vehicle and the corresponding vehicle within a certain period after the current time node based on the first dataset, determining whether the target vehicle enters the collision warning mode based on the distance change trend, and issuing a collision warning signal in a timely manner specifically includes:

[0032] Based on the first dataset, the acceleration, velocity, and position of the target vehicle are recursively calculated according to the time sequence to infer the position data of the target vehicle at the next time node;

[0033] The acceleration, velocity, and position of the corresponding vehicle are recursively calculated according to the time sequence to infer the position data of the vehicle at the next time point.

[0034] Infer the trend of vehicle distance changes based on the current location data of the target vehicle and the corresponding vehicle, and the inferred location data.

[0035] If the lowest point of the vehicle distance change trend is less than zero, the collision warning mode will be activated and a collision warning signal will be issued in a timely manner; otherwise, no action will be taken.

[0036] Preferably, the method further includes: sending collision information to an emergency contact list when a collision occurs, the collision information including vehicle identification information, current location information, and alarm information.

[0037] Another objective of this invention is to provide a collision warning system for autonomous vehicles based on big data, the system comprising: a warning dataset construction unit, a warning dataset calculation unit, a secondary dataset calculation unit, and a collision warning unit;

[0038] The early warning dataset construction unit is used to acquire the driving information of the target vehicle and the driving information of surrounding vehicles within a preset time and space range. All driving information constitutes the early warning dataset. The driving information includes at least: vehicle identification information, location, speed, and acceleration data.

[0039] The warning dataset calculation unit is used to calculate the difference between the acceleration of all surrounding vehicles in the warning dataset and the reference acceleration.

[0040] When one or more of the differences are greater than the first acceleration threshold, the secondary dataset calculation unit acquires the position, speed, and acceleration data of the vehicle corresponding to the one or more differences, and combines them with the position, speed, and acceleration data of the target vehicle to form the first dataset.

[0041] The collision warning unit is used to infer the trend of the distance between the target vehicle and the corresponding vehicle within a certain period of time after the current time node based on the first dataset, and to determine whether the target vehicle enters the collision warning mode based on the trend of the distance change, and to issue a collision warning signal in a timely manner.

[0042] The secondary dataset calculation unit is also used to keep the target vehicle in autonomous driving mode when all the difference values ​​are less than or equal to the first acceleration threshold.

[0043] Preferably, the early warning dataset construction unit includes: a spatiotemporal range setting module, a first driving information acquisition module, a second driving information acquisition module, and a driving information classification and processing module;

[0044] The spatiotemporal range setting module is used to set the time and spatial zoning information of the spatiotemporal range with the target vehicle as the origin, and to iteratively update it based on the real-time location of the target vehicle.

[0045] The first driving information acquisition module is used to acquire the navigation information of the target vehicle, acquire the driving information of the target vehicle, and correct the driving information of the target vehicle through the navigation information of the target vehicle.

[0046] The second driving information acquisition module is used to acquire driving information of surrounding vehicles within a preset time and space range relative to the target vehicle, based on a set time and space range.

[0047] The driving information classification and processing module is used to classify the driving information of the target vehicle and the driving information of surrounding vehicles; and to normalize the classified driving information to form an early warning dataset.

[0048] Preferably, the early warning dataset calculation unit includes: a reference acceleration selection module, a reference acceleration gain module, a first acceleration difference calculation module, and a second acceleration difference calculation module;

[0049] The reference acceleration selection module is used to select an initial reference acceleration based on the historical driving data of the target vehicle.

[0050] The reference acceleration gain module is used to determine the optimal reference acceleration by increasing the initial reference acceleration based on the current speed data of the target vehicle.

[0051] The first acceleration difference calculation module is used to calculate the difference between the acceleration of vehicles in front and behind in the early warning data and the reference acceleration.

[0052] The second acceleration difference calculation module is used to calculate the difference between the acceleration of the vehicle on the side of the warning data and the reference acceleration.

[0053] This invention provides a collision warning method for autonomous vehicles based on big data. First, the target vehicle enters autonomous driving mode. During this process, the driving information of the target vehicle and surrounding vehicles within a preset spatiotemporal range are acquired in real time, forming a warning dataset. Then, the difference between the acceleration of all surrounding vehicles and a reference acceleration in the warning dataset is calculated, without processing all data in the dataset. This reduces the data processing load and allows for faster and more accurate processing of specific data, such as the difference between acceleration and reference acceleration, under the same processing performance conditions. Finally, a judgment is made based on the ratio of this difference to a first acceleration threshold. This adds a pre-judgment mechanism to determine the priority of data processing, prioritizing scenarios where a collision is likely. The process involves acquiring the position, speed, and acceleration data of the vehicles corresponding to one or more differences, and combining this data with the position, speed, and acceleration data of the target vehicle to form a first dataset. Based on this first dataset, the position change routes of the target vehicle and its corresponding vehicle are statistically analyzed or plotted. This allows for the prediction of the distance change trend between the target vehicle and its corresponding vehicle within a certain timeframe after the current time point. Based on this distance change trend, it is determined whether the target vehicle should enter collision warning mode, and a collision warning signal is issued in a timely manner. Otherwise, the target vehicle remains in autonomous driving mode. Because not all data is processed, the processing performance requirements are reduced, achieving a better balance between performance and power consumption. Furthermore, the timeliness of collision warnings is improved, thereby increasing the likelihood of mass production or marketability of advanced autonomous vehicles. Attached Figure Description

[0054] Figure 1 A flowchart illustrating a collision warning method for autonomous vehicles based on big data, provided as an embodiment of the present invention;

[0055] Figure 2 This is a flowchart illustrating the specific composition of the early warning dataset in this embodiment of the invention;

[0056] Figure 3 A flowchart illustrating another collision warning method for autonomous vehicles based on big data, provided as an embodiment of the present invention;

[0057] Figure 4 This is a flowchart illustrating the adjustment of the acquisition frequency of driving information in an embodiment of the present invention.

[0058] Figure 5 A flowchart illustrating the specific process of calculating the difference between the acceleration of all surrounding vehicles in the early warning dataset and the reference acceleration in this embodiment of the invention;

[0059] Figure 6 This is a flowchart illustrating the process of determining whether a target vehicle enters a collision warning mode based on the first dataset in this embodiment of the invention.

[0060] Figure 7 A structural block diagram of a collision warning system for autonomous vehicles based on big data, provided for an embodiment of the present invention;

[0061] Figure 8 This is a structural block diagram of an early warning dataset construction unit provided in an embodiment of the present invention;

[0062] Figure 9 This is a structural block diagram of an early warning dataset calculation unit provided in an embodiment of the present invention;

[0063] Figure 10 This is a block diagram of the internal structure of a computer device in one embodiment. Detailed Implementation

[0064] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0065] It is understood that the terms “first,” “second,” etc., used in this application may be used herein to describe various elements, but unless otherwise stated, these elements are not limited by these terms. These terms are used only to distinguish one element from another.

[0066] Figure 1 A flowchart of a collision warning method for autonomous vehicles based on big data, provided for embodiments of the present invention, may specifically include the following steps:

[0067] Step S101: Obtain the driving information of the target vehicle and the driving information of surrounding vehicles within a preset time and space range. All driving information constitutes a warning dataset. The driving information includes at least: vehicle identification information, position, speed, and acceleration data. Specifically, the driving information of the target vehicle can be directly retrieved from the data stored in the vehicle's computer or shared with the target vehicle's sensors. The driving information of surrounding vehicles can be obtained through a set of sensors, such as cameras, lidar, and acoustic radar, or by establishing communication with surrounding vehicles via a wireless network. Alternatively, it can be shared through a shared navigation map to obtain the data.

[0068] like Figure 2 As shown, in one embodiment, step S101, which involves acquiring the driving information of the target vehicle and acquiring the driving information of surrounding vehicles within a preset spatiotemporal range, with all driving information forming a warning dataset, specifically includes the following steps:

[0069] Step S1011: Set the time and space zoning information of the spatiotemporal range with the target vehicle as the origin, and iteratively update it based on the real-time position of the target vehicle.

[0070] In this step, the spatiotemporal range can be decomposed into a spatial range and a temporal range. Spatial range: with the target vehicle as the origin and the target vehicle's driving direction as forward (or ahead), a rectangular area is defined, with each side 50m in front and behind and 3m in the left and right. Temporal range: with the target vehicle's current time point as the base point, the range extends to five seconds in the past and five seconds in the future.

[0071] Step S1012: Obtain the navigation information of the target vehicle, obtain the driving information of the target vehicle, and correct the driving information of the target vehicle using the navigation information of the target vehicle; generally, the navigation information of the target vehicle is existing technology and can be obtained directly through existing navigation software; or it can be obtained through a simple reverse calculation step.

[0072] Step S1013: Based on the set time and space range, obtain the driving information of surrounding vehicles within the preset time and space range of the target vehicle.

[0073] Step S1014: Classify the driving information of the target vehicle and the driving information of surrounding vehicles.

[0074] In this step, the driving information can generally be categorized according to the types of information it contains, such as vehicle position, speed, and acceleration. Vehicle position can be divided into two subcategories: position a certain time before the current time point and current position. Acceleration can be divided into two subcategories: vertical acceleration and lateral acceleration. Obviously, for vehicles located in front of or behind the target vehicle, the vertical acceleration data is more important, and the difference in lateral acceleration is calculated first. For vehicles located to the side of the target vehicle (left rear side, right rear side), the lateral acceleration data is more important, and the difference in vertical acceleration is calculated first.

[0075] Step S1015: Normalize the classified driving information to form a warning dataset; the normalization process uses a widely used normalization algorithm; it is only necessary to ensure that the driving information of the target vehicle and the driving information of surrounding vehicles have the same amount of data when compared.

[0076] In one example of an embodiment, when acquiring driving information, the vehicle's identification information, location, speed, and acceleration data can typically be marked or bound and encoded. This facilitates the creation of an index based on the vehicle's identification information to establish a directory for the warning dataset. It also facilitates subsequent tracing and differentiation based on the encoding. The vehicle's identification information can be the vehicle's license plate number, engine number, chassis number, or other displacement identification code. The vehicle's license plate number can be selected, but it is not limited to this.

[0077] In one embodiment, the driving information may further include tire pressure information. When the detected tire pressure is significantly lower than a set pressure limit, monitoring and warnings can be issued for surrounding vehicles corresponding to that tire pressure. For example: For small cars: normal tire pressure is 210-230 kPa; when the seats and trunk are fully loaded, it is 220-240 kPa; when the vehicle speed exceeds 140 km / h, it is 220-230 kPa. For compact cars: normal tire pressure is 220-240 kPa; when the seats and trunk are fully loaded, it is 230-250 kPa; when the vehicle speed exceeds 140 km / h, it is 230-250 kPa. 250 kPa; Medium-sized vehicles: Normal tire pressure is 230-250 kPa, 240-260 kPa when fully loaded with seats and trunk, and 240-260 kPa at speeds exceeding 140 km / h; Mid-to-large vehicles: Normal tire pressure is 250-270 kPa, 270-290 kPa when fully loaded with seats and trunk, and 270-290 kPa at speeds exceeding 140 km / h; Large vehicles: Normal tire pressure is 260-300 kPa, 280-310 kPa when fully loaded with seats and trunk, and 280-310 kPa at speeds exceeding 140 km / h. Therefore, the standard tire pressure limit can be simplified to a uniform 2.0 kPa, or even 1.8 kPa, or can be set individually; this can proactively avoid the possibility of collisions with the target vehicle caused by malfunctions of surrounding vehicles within a preset time and space range.

[0078] like Figure 1 As shown, in one embodiment, the method may further include: step S103, calculating the difference between the acceleration of all surrounding vehicles in the warning dataset and the reference acceleration;

[0079] like Figure 5 As shown, in one example of an embodiment, step S103, the step of calculating the difference between the acceleration of all surrounding vehicles in the warning dataset and the reference acceleration, specifically includes:

[0080] Step S1031: Select an initial reference acceleration based on the historical driving data of the target vehicle; at the same time, if there is no historical driving data for the target vehicle, an initial reference acceleration can be set based on experience, and then the reference acceleration can be optimized and iterated based on the obtained driving data.

[0081] Step S1032: Based on the current speed data of the target vehicle, the initial reference acceleration is increased to determine the optimal reference acceleration; in this step, the increase can be generated or output by big data through a specified deep learning model.

[0082] In one example of an embodiment, different initial reference accelerations can be set for village roads, rural roads, national highways, or expressways, and optimized and iterated using big data; similarly, different initial reference accelerations can also be set for urban roads, and optimized and iterated accordingly.

[0083] Step S1033: Calculate the difference between the acceleration of the vehicles in front and behind in the warning data and the reference acceleration.

[0084] Step S1034: Calculate the difference between the acceleration of the vehicle on the side of the warning data and the reference acceleration.

[0085] In one embodiment, when there are no vehicles in front of or behind the target vehicle, only the vehicles behind or in front can be calculated; when the target vehicle is in the left or right lane, only the side with the lane is calculated; in this embodiment, steps S1033 and S1034 can be performed independently or simultaneously, and the order of the steps can be adjusted according to real-time traffic conditions.

[0086] This invention provides a collision warning method for autonomous vehicles based on big data. First, the target vehicle enters autonomous driving mode. During this process, the driving information of the target vehicle and surrounding vehicles within a preset spatiotemporal range are acquired in real time, forming a warning dataset. Then, the difference between the acceleration of all surrounding vehicles and a reference acceleration in the warning dataset is calculated, without processing all data in the dataset. This reduces the data processing load and allows for faster and more accurate processing of specific data, such as the difference between acceleration and reference acceleration, under the same processing performance conditions. Finally, a pre-judgment mechanism is added to determine the priority of data processing based on the ratio of this difference to a first acceleration threshold, prioritizing scenarios where a collision is likely. The process involves acquiring the position, speed, and acceleration data of the vehicles corresponding to one or more differences, and combining this data with the position, speed, and acceleration data of the target vehicle to form a first dataset. Based on this first dataset, the path of position change between the target vehicle and its corresponding vehicle is statistically analyzed or plotted. This allows for the prediction of the distance change trend between the target vehicle and its corresponding vehicle within a certain timeframe after the current time point. Based on this distance change trend, it is determined whether the target vehicle should enter collision warning mode, and a collision warning signal is issued in a timely manner. Otherwise, the target vehicle remains in autonomous driving mode. Because not all data is processed, the processing performance requirements are reduced, achieving a better balance between performance and power consumption. Furthermore, the timeliness of collision warnings is improved, thereby increasing the likelihood of mass production or marketability of advanced autonomous vehicles.

[0087] In one embodiment, the method may further include: step S105, when the value of one or more of the differences is greater than a first acceleration threshold, the first acceleration threshold may be a fixed value; acquiring the position, speed, and acceleration data of the vehicle corresponding to the one or more differences, and combining them with the position, speed, and acceleration data of the target vehicle to form a first dataset.

[0088] like Figure 1 As shown, in one embodiment, the method may further include: step S107, based on the first dataset, inferring the trend of the distance between the target vehicle and the corresponding vehicle within a certain period of time after the current time node, and determining whether the target vehicle enters the collision warning mode based on the distance change trend, and issuing a collision warning signal in a timely manner.

[0089] In one embodiment, the method may further include: step S109, when the values ​​of all differences are less than or equal to a first acceleration threshold, then the target vehicle is kept in autonomous driving mode.

[0090] like Figure 6 As shown in one embodiment, step S107 specifically includes:

[0091] Step S1071: Based on the first dataset, recursively calculate the acceleration, velocity, and position of the target vehicle according to the time sequence to infer the position data of the target vehicle at the next time node;

[0092] Step S1072: Recursively calculate the acceleration, velocity and position of the corresponding vehicle according to the time sequence to infer the position data of the vehicle at the next time node;

[0093] Step S1073: Infer the trend of vehicle distance change based on the current position data of the target vehicle and the corresponding vehicle and the inferred position data;

[0094] Step S1074: If the lowest point of the vehicle distance change trend is less than zero, then enter the collision warning mode and issue a collision warning signal in a timely manner; otherwise, do not process it.

[0095] In the above example, the collision warning signal can be divided into three levels: green warning, yellow warning, and red warning, and is indicated by sound and display indicators; or it can be divided into two levels: yellow warning and red warning.

[0096] like Figure 2 As shown, in one embodiment, the method further includes:

[0097] Step S111: Based on the navigation information of the target vehicle, the driving route of the target vehicle is divided into a driving link consisting of several nodes; wherein several nodes correspond to auxiliary facilities on the current road in the navigation map; based on the driving link, the frequency of obtaining driving information at the node location is increased.

[0098] In this step, the navigation information of the target vehicle is set before the target vehicle enters autonomous driving. This can be set manually by the driver or assisted by the target vehicle itself; for example, it can be conveniently selected based on historical navigation records or the driver's commuting location. The driving link can replace the navigation route in conventional navigation information, i.e., the target vehicle's driving route. Auxiliary facilities on the current road refer to school gates, bus stops, tunnels, bridges, etc.; while nodes can be used to represent intersections, traffic lights, pedestrian crossings, etc. In this embodiment, increasing the frequency of acquiring driving information at the corresponding locations of auxiliary facilities can improve the accuracy and timeliness of collision warnings and the ability to respond to complex road conditions during autonomous driving; this contributes to the concrete realization of high-level driving automation. Maintaining a normal frequency of acquiring driving information between nodes without or lacking auxiliary facilities can improve the driving range during autonomous driving.

[0099] like Figure 3As shown, in one embodiment, after dividing the target vehicle's travel route into a travel link consisting of several nodes based on the target vehicle's navigation information, the travel link is optimized, specifically including:

[0100] Step S401: Based on big data comparison of the changes in auxiliary facilities, adjust the corresponding nodes according to the changes;

[0101] In this step, the auxiliary facilities may include temporary roadblocks, falling objects, etc. After the temporary roadblocks or falling objects are added or removed, the auxiliary facilities can be updated into the driving link, thereby improving the ability to provide collision warnings.

[0102] Step S402: Based on the number of nodes in two adjacent distance segments within a specified distance segment, increase transition points proportionally along the direction of increasing nodes on the driving link; for example: if the specified distance segment is 500m, the more nodes there are in two adjacent 500m distance segments, the more complex the road conditions are, therefore, the number of nodes needs to be increased accordingly.

[0103] Step S403: Adjust the frequency of acquiring driving information with reference to the transition point position; the transition point is a dot on the driving link, but it represents the location of the auxiliary facility between two nodes. Adjusting the frequency of acquiring driving information when approaching the auxiliary facility can improve the accuracy of collision warning.

[0104] In a preferred embodiment, the method further includes: sending collision information to an emergency contact list when a collision occurs, the collision information including vehicle identification information, current location information, and alarm information.

[0105] In this embodiment, the emergency contact list may include roadside emergency phone numbers, emergency contact phone numbers, insurance company phone numbers, etc.

[0106] like Figure 7 As shown, in another embodiment, an autonomous vehicle collision warning system based on big data is provided, the system comprising: a warning dataset construction unit 100, a warning dataset calculation unit 200, a secondary dataset calculation unit 300, and a collision warning unit 400.

[0107] The early warning dataset construction unit 100 is used to acquire the driving information of the target vehicle and the driving information of surrounding vehicles within a preset time and space range. All driving information constitutes the early warning dataset. The driving information includes at least: vehicle identification information, location, speed, and acceleration data.

[0108] The early warning dataset calculation unit 200 is used to calculate the difference between the acceleration of all surrounding vehicles in the early warning dataset and the reference acceleration.

[0109] When one or more of the differences are greater than the first acceleration threshold, the secondary dataset calculation unit 300 acquires the position, speed, and acceleration data of the vehicle corresponding to the one or more differences, and combines them with the position, speed, and acceleration data of the target vehicle to form the first dataset.

[0110] The collision warning unit 400 is used to infer the trend of the distance between the target vehicle and the corresponding vehicle within a certain period of time after the current time node based on the first dataset, and to determine whether the target vehicle enters the collision warning mode based on the trend of the distance change, and to issue a collision warning signal in a timely manner.

[0111] The secondary dataset calculation unit 300 is also used to keep the target vehicle in autonomous driving mode when all the difference values ​​are less than or equal to the first acceleration threshold.

[0112] In this embodiment, when obtaining driving information through the early warning dataset construction unit 100, the vehicle's identification information, location, speed, and acceleration data can be marked or bound and encoded. This facilitates the creation of an index based on the vehicle's identification information to establish a directory for the early warning dataset. It also facilitates subsequent tracing and differentiation based on the encoding. The vehicle's identification information can be the vehicle's license plate number, engine number, chassis number, or other displacement identification code. The vehicle's license plate number can be selected, but it is not limited to this.

[0113] like Figure 8 As shown, in one embodiment, the early warning dataset construction unit 100 includes: a spatiotemporal range setting module 110, a first driving information acquisition module 120, a second driving information acquisition module 130, and a driving information classification and processing module 140;

[0114] The spatiotemporal range setting module 110 is used to set the time and space division information of the spatiotemporal range with the target vehicle as the origin, and to iteratively update it based on the real-time position of the target vehicle.

[0115] The first driving information acquisition module 120 is used to acquire the navigation information of the target vehicle, acquire the driving information of the target vehicle, and correct the driving information of the target vehicle through the navigation information of the target vehicle.

[0116] The second driving information acquisition module 130 is used to acquire driving information of surrounding vehicles within a preset time and space range relative to the target vehicle, based on a set time and space range.

[0117] The driving information classification and processing module 140 is used to classify the driving information of the target vehicle and the driving information of surrounding vehicles; and to normalize the classified driving information to form an early warning dataset.

[0118] In this embodiment, the first driving information acquisition module 120 can be software, hardware, or a combination thereof, and can be a computer program. When the computer program is executed by a computer, it can call or read the data of the in-vehicle computer of the target vehicle. The second driving information acquisition module 120 can be a program that is paired with a sensor group to acquire the driving information of surrounding vehicles through the sensor group. The sensor group can be a speed sensor, an angular velocity sensor, etc., as well as a part or all of a camera (such as a high-definition camera, an infrared camera, or an industrial high-speed camera), a sonic radar, a lidar, and a ultrasonic radar, to achieve speed measurement, displacement measurement, etc. It can also be a navigation module including a wireless network to acquire and update all driving information within a preset time and space range.

[0119] As Figure 9 shown, in one embodiment, the warning dataset calculation unit 200 includes: a reference acceleration selection module 210, a reference acceleration gain module 220, a first acceleration difference calculation module 230, and a second acceleration difference calculation module 240;

[0120] The reference acceleration selection module 210 is used to select an initial reference acceleration according to the historical driving data of the target vehicle;

[0121] The reference acceleration gain module 220 is used to determine the best reference acceleration by gaining the initial reference acceleration according to the current speed data of the target vehicle;

[0122] The first acceleration difference calculation module 230 is used to calculate the difference between the accelerations of the front and rear vehicles in the warning dataset and the reference acceleration;

[0123] The second acceleration difference calculation module 240 is used to calculate the difference between the acceleration of the side vehicle in the warning dataset and the reference acceleration.

[0124] Figure 10An internal structural diagram of a computer device according to one embodiment is shown. The computer device includes a processor, memory, network interface, input device, and display screen connected via a system bus. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and may also store a computer program that, when executed by the processor, causes the processor to perform the methods described above. The internal memory may also store a computer program that, when executed by the processor, causes the processor to perform the methods described above. The display screen of the computer device may be a liquid crystal display (LCD) or an e-ink display. The input device may be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the computer device casing, or an external keyboard, touchpad, or mouse, etc.

[0125] Those skilled in the art will understand that Figure 10 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0126] In one embodiment, the apparatus provided in this application can be implemented as a computer program, and the computer program can be implemented as follows: Figure 10 The system runs on the computer device shown. The computer device's memory can store the various program modules that make up this big data-based autonomous vehicle collision warning system, for example... Figure 7 The diagram shows a warning dataset construction unit, a warning dataset calculation unit, a secondary dataset calculation unit, and a collision warning unit. The computer program comprised of these program units causes the processor to execute the steps of the methods in the various embodiments of this application described in this specification.

[0127] For example, Figure 10 The computer device shown can be used as follows Figure 8 The spatiotemporal range setting module 110 in the warning dataset construction unit 100 shown executes step S1011. The computer device can execute step S1012 through the first driving information acquisition module 120. The computer device can execute step S1013 through the second driving information acquisition module 130. The computer device can execute step S1014 through the driving information classification and processing module 140.

[0128] It should be understood that although the steps in the flowcharts of the various embodiments of the present invention are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the various embodiments may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.

[0129] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0130] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A collision warning method for autonomous vehicles based on big data, characterized in that, The method includes: The system acquires the driving information of the target vehicle and the driving information of surrounding vehicles within a preset spatiotemporal range. All driving information constitutes a warning dataset. The driving information includes at least: vehicle identification information, location, speed, and acceleration data. Specifically, the system includes the following steps: The system defines a time and space zoning information within a spatiotemporal range, with the target vehicle as the origin, and iteratively updates this information based on the target vehicle's real-time location. It acquires the target vehicle's navigation information and driving information, and corrects the driving information using the navigation information. Based on the defined spatiotemporal range, it acquires the driving information of surrounding vehicles within the same range. The system then classifies the target vehicle's driving information and the driving information of surrounding vehicles. Finally, it normalizes the classified driving information to form a warning dataset. Calculate the difference between the acceleration of all surrounding vehicles in the warning dataset and the reference acceleration; When one or more of the differences are greater than the first acceleration threshold, the position, speed, and acceleration data of the vehicle corresponding to the one or more differences are obtained and combined with the position, speed, and acceleration data of the target vehicle to form the first dataset; Based on the first dataset, the trend of the distance between the target vehicle and the corresponding vehicle within a certain period after the current time node is inferred, and the target vehicle is determined to enter the collision warning mode based on the distance change trend, and a collision warning signal is issued in a timely manner.

2. The method according to claim 1, characterized in that, The method further includes: Based on the navigation information of the target vehicle, the driving route of the target vehicle is divided into a driving link consisting of several nodes; wherein several nodes correspond to auxiliary facilities on the current road in the navigation map; Based on the aforementioned driving link, the frequency of obtaining driving information at node locations is increased.

3. The method according to claim 2, characterized in that, After dividing the target vehicle's travel route into a travel link consisting of several nodes based on the target vehicle's navigation information, the travel link is optimized, specifically including: Based on big data comparison of changes in auxiliary facilities, the corresponding nodes are adjusted accordingly. Based on the number of nodes within two adjacent distances in a specified distance segment, increase transition points proportionally along the direction of increasing nodes on the driving link; Adjust the frequency of acquiring driving information based on the aforementioned transition point location.

4. The method according to claim 1, characterized in that, The specific steps for calculating the difference between the acceleration of all surrounding vehicles in the early warning dataset and the reference acceleration include: Select the initial reference acceleration based on the target vehicle's historical driving data; Based on the target vehicle's current speed data, the initial reference acceleration is increased to determine the optimal reference acceleration; Calculate the difference between the acceleration of vehicles in front and behind in the warning data and the reference acceleration; Calculate the difference between the acceleration of vehicles on the side of the warning data and the reference acceleration.

5. The method according to claim 1, characterized in that, The steps of inferring the trend of distance changes between the target vehicle and the corresponding vehicle within a certain period after the current time node based on the first dataset, determining whether the target vehicle should enter the collision warning mode based on the distance change trend, and issuing a collision warning signal in a timely manner specifically include: Based on the first dataset, the acceleration, velocity, and position of the target vehicle are recursively calculated according to the time sequence to infer the position data of the target vehicle at the next time node; The acceleration, velocity, and position of the corresponding vehicle are recursively calculated according to the time sequence to infer the position data of the vehicle at the next time point. Infer the trend of vehicle distance changes based on the current location data of the target vehicle and the corresponding vehicle, and the inferred location data. If the lowest point of the vehicle distance change trend is less than zero, the collision warning mode will be activated and a collision warning signal will be issued in a timely manner; otherwise, no action will be taken.

6. The method according to claim 1, characterized in that, The method further includes sending collision information to an emergency contact list when a collision occurs, the collision information including vehicle identification information, current location information, and alarm information.

7. A collision warning system for autonomous vehicles based on big data, characterized in that, The system includes: an early warning dataset construction unit, an early warning dataset calculation unit, a secondary dataset calculation unit, and a collision early warning unit; The early warning dataset construction unit is used to acquire the driving information of the target vehicle and the driving information of surrounding vehicles within a preset time and space range. All driving information constitutes the early warning dataset. The driving information includes at least: vehicle identification information, location, speed, and acceleration data. The early warning dataset construction unit includes: a spatiotemporal range setting module, a first driving information acquisition module, a second driving information acquisition module, and a driving information classification and processing module; The spatiotemporal range setting module is used to set the time and spatial division information of the spatiotemporal range with the target vehicle as the origin, and iteratively update it based on the real-time position of the target vehicle; the first driving information acquisition module is used to acquire the navigation information of the target vehicle, acquire the driving information of the target vehicle, and correct the driving information of the target vehicle using the navigation information; the second driving information acquisition module is used to acquire the driving information of surrounding vehicles within the preset spatiotemporal range of the target vehicle according to the set spatiotemporal range; the driving information classification and processing module is used to classify the driving information of the target vehicle and the driving information of surrounding vehicles; and to normalize the classified driving information to form a warning dataset. The warning dataset calculation unit is used to calculate the difference between the acceleration of all surrounding vehicles in the warning dataset and the reference acceleration. When one or more of the differences are greater than the first acceleration threshold, the secondary dataset calculation unit acquires the position, speed, and acceleration data of the vehicle corresponding to the one or more differences, and combines them with the position, speed, and acceleration data of the target vehicle to form the first dataset. The collision warning unit is used to infer the trend of the distance between the target vehicle and the corresponding vehicle within a certain period of time after the current time node based on the first dataset, and to determine whether the target vehicle enters the collision warning mode based on the trend of the distance change, and to issue a collision warning signal in a timely manner.

8. The system according to claim 7, characterized in that, The early warning dataset calculation unit includes: a reference acceleration selection module, a reference acceleration gain module, a first acceleration difference calculation module, and a second acceleration difference calculation module; The reference acceleration selection module is used to select an initial reference acceleration based on the historical driving data of the target vehicle. The reference acceleration gain module is used to determine the optimal reference acceleration by increasing the initial reference acceleration based on the current speed data of the target vehicle. The first acceleration difference calculation module is used to calculate the difference between the acceleration of vehicles in front and behind in the early warning data and the reference acceleration. The second acceleration difference calculation module is used to calculate the difference between the acceleration of the vehicle on the side of the warning data and the reference acceleration.

Images