Intelligent updating method and system for automatic driving navigation map

Abstract

The invention relates to the technical field of automobile navigation, in particular to an intelligent updating method and system for an automatic driving navigation map, and aims to obtain the space influence of each congestion area on a target area based on the change of navigation path planning and obtain the time difference of congestion of each congestion area and the target area. According to the method, the correlation between the congestion phenomenon of each congestion area and the target area is obtained, the time influence is obtained through the correlation, the time influence of each congestion area on the target area is determined, the initial traffic flow of the target area is corrected according to the space influence and the time influence, and the accurate and real traffic flow of the target area can be obtained. The accuracy of vehicle flow acquisition is improved, and the accuracy of map updating priority acquisition is improved.

Description

Technical Field

[0001] The present invention relates to the field of automobile navigation technology, and in particular to an intelligent updating method and system for an automatic driving navigation map. Background Art

[0002] Autonomous driving technology is highly dependent on high-precision maps to achieve functions such as accurate positioning, path planning, and environmental perception. Map updates are crucial for autonomous driving because the environment is constantly changing, such as road construction, traffic sign adjustments, etc. If these changes are not reflected in the map in a timely manner, they may affect the safety and reliability of autonomous driving vehicles. Therefore, continuously updating map data is the key to ensuring the efficient operation of the autonomous driving system.

[0003] The existing method of updating autonomous driving maps is to update the maps of each area according to the map update priority of each area on the map, where the higher the map update priority, the higher the priority of the update. At present, the map update priority of each area is obtained directly from the traffic volume of each area. The larger the traffic volume, the higher the map update priority. However, the traffic volume of a single area is usually affected by other related areas around it, resulting in the traffic volume not being its actual traffic volume, thereby affecting the accuracy of obtaining the map update priority. Summary of the invention

[0004] In order to solve the technical problem of inaccurate acquisition of vehicle flow in the existing area, the purpose of the present invention is to provide an intelligent updating method and system for autonomous driving navigation maps. The technical solutions adopted are as follows: In a first aspect of the present invention, a method for intelligently updating an autonomous driving navigation map is provided, comprising: Based on the change of the navigation path planning, the spatial impact of each congested area on the target area is obtained, and the congested area is an area with vehicle congestion within a preset neighborhood of the target area; Combined with the time difference between the occurrence of congestion in each congested area and the target area, the correlation between the congestion phenomenon in each congested area and the target area is obtained; determining a time impact of each congested area on the target area, the time impact being obtained from the correlation; The initial vehicle flow in the target area is modified according to the spatial influence and the temporal influence.

[0005] In an exemplary embodiment, the process of acquiring the congested area includes: Acquire a related area, where the related area is an area within a preset neighborhood of the target area; Obtain the average speed of the autonomous driving vehicles in each associated area during the target period and compare it with the preset average speed threshold; If the average speed is less than a preset average speed threshold, the corresponding associated area is determined to be a congested area.

[0006] In an exemplary embodiment, obtaining the spatial impact of each congested area on the target area based on the change of navigation path planning includes: Acquire a first number of autonomous driving vehicles whose original navigation paths pass through a first congested area and do not pass through the target area during a target time period; the first congested area is any congested area; Acquire a second number of autonomous driving vehicles that change an original navigation path into an actual navigation path during the target time period; the actual navigation path is a navigation path that passes through the target area and does not pass through the first congested area; A ratio of the second number to the first number is obtained as a spatial impact factor of the first congested area on the target area.

[0007] In an exemplary embodiment, the process of obtaining the time difference between the occurrence of congestion in each congested area and the target area includes: Obtaining historical traffic volume of a first congestion area and a target area; the first congestion area is any congestion area; According to the relationship between the historical traffic volume and time in the first congestion area and the target area, a linear regression model is constructed; The linear regression model is solved to obtain the time difference between the occurrence of congestion in the first congested area and the target area.

[0008] In an exemplary embodiment, the linear regression model is: ; in, Indicates that the target area is Historical traffic volume over time, Indicates that the oth congested area is Historical traffic volume over time, It represents the time difference between the congestion in the oth congested area and the target area. represents the regression coefficient, represents the error term of the linear regression model.

[0009] In an exemplary embodiment, the process of obtaining the correlation between the congestion phenomena of each congested area and the target area includes: Acquire a first vehicle flow sequence of a first congestion time period in a first congestion area in a target time period; the first congestion area is any congestion area, and the first congestion time period is a congestion time period in the first congestion area in the target time period; According to the first congestion time period and the time difference, a second congestion time period is obtained, where the second congestion time period is a congestion time period of a target area; Acquire a second traffic flow sequence of the target area in the second congestion time period; Obtain a correlation coefficient between the first vehicle flow sequence and the second vehicle flow sequence.

[0010] In an exemplary embodiment, determining the time impact of each congested area on the target area includes: A result of normalizing the correlation coefficient corresponding to the first congested area is used as a time influence factor of the first congested area on the target area.

[0011] In an exemplary embodiment, the modifying the initial vehicle flow of the target area according to the spatial impact and the temporal impact includes: According to the spatial impact and the temporal impact, a comprehensive impact factor of each congested area on the target area is obtained; The comprehensive impact factors of all congested areas on the target area are integrated to obtain the final impact factor; Determining a correction factor, wherein the correction factor is negatively correlated with the final impact factor; The correction factor is multiplied by the initial vehicle flow of the target area to obtain the corrected vehicle flow of the target area.

[0012] In an exemplary embodiment, the spatial impact is specifically a spatial impact factor, and the temporal impact is specifically a temporal impact factor; The comprehensive impact factor of each congested area on the target area is obtained according to the spatial impact and the temporal impact, including: Obtaining a third number of autonomous driving vehicles whose driving paths pass through a first congestion area during a target period, and a fourth number of autonomous driving vehicles whose driving paths pass through a target area; the first congestion area being any congestion area; Obtaining a ratio of the third number to the total number as a weight of the time impact corresponding to the first congested area; the total number is the sum of the third number and the fourth number; According to the weights, the spatial impact factor and the temporal impact factor are weighted and summed to obtain a comprehensive impact factor of the first congested area on the target area.

[0013] In a second aspect of the present invention, there is provided an intelligent update system for an autonomous driving navigation map, comprising: a memory and a processor; the memory is connected to the processor; the memory is used to store program instructions; the processor is used to implement the above-mentioned intelligent update method for the autonomous driving navigation map when the program instructions are executed.

[0014] The present invention has the following beneficial effects: since areas with vehicle congestion within a preset neighborhood of a target area will affect the traffic flow of the target area, based on changes in navigation path planning, the spatial influence of each congested area on the target area is obtained, and according to the time difference between the occurrence of congestion in each congested area and the target area, and the correlation between the congestion phenomena of each congested area and the target area, the temporal influence of each congested area on the target area is determined, and the influence of each congested area on the traffic flow of the target area is obtained from the two dimensions of spatial influence and temporal influence, and finally the initial traffic flow of the target area is corrected according to the spatial influence and the temporal influence, so that the accurate and real traffic flow of the target area can be obtained, and the accuracy of traffic flow acquisition is improved, thereby improving the accuracy of map update priority acquisition. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a flow chart of an intelligent updating method of an autonomous driving navigation map provided by an embodiment of the present invention; Figure 2 is a flow chart for obtaining a congested area provided by an embodiment of the present invention; Figure 3 is a flow chart of obtaining spatial influence provided by an embodiment of the present invention; Figure 4 is a flow chart of obtaining a time difference provided by an embodiment of the present invention; Figure 5 is a flowchart of obtaining correlation provided by an embodiment of the present invention; Figure 6 is a correction flow chart for correcting the initial vehicle flow in the target area provided by an embodiment of the present invention; Figure 7 This is a flowchart for obtaining a comprehensive impact factor provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0016] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0017] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs. The data and information collected in this application have been obtained with full consent and authorization, and the collection, use and processing of relevant information must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0018] This embodiment provides an intelligent updating method for an autonomous driving navigation map. Figure 1 As shown, the following steps are included: Step 1: Based on the change of navigation path planning, the spatial impact of each congested area on the target area is obtained. The congested area is an area with vehicle congestion within a preset neighborhood of the target area. Step 2: Combine the time difference between the congestion in each congested area and the target area to obtain the correlation between the congestion phenomenon in each congested area and the target area; Step 3: Determine the time impact of each congested area on the target area. The time impact is obtained by correlation. Step 4: Modify the initial traffic flow in the target area based on spatial and temporal impacts.

[0019] Each step is described in detail below with reference to the accompanying drawings.

[0020] Step 1: Based on the change of navigation path planning, the spatial impact of each congested area on the target area is obtained. The congested area is an area with vehicle congestion within a preset neighborhood of the target area.

[0021] Since this application is to correct the initial traffic flow of the target area, it is necessary to divide the original map into regions in advance to obtain multiple regions. The division method of each region is set according to the actual situation, such as dividing according to administrative regions, or dividing the map into multiple regions.

[0022] In an exemplary embodiment, first, a grid division method is used to convert a map into a grid data structure, and the map is divided into a plurality of grid cells of equal size. Then, according to the road topology information, the grid cells are merged or adjusted to form equal-sized areas, and the road coherence and integrity in the area are guaranteed as much as possible. Specifically: 1. The map data is converted into a grid format to determine the appropriate grid size. At the same time, the road topology information in the map is extracted, including the nodes and edges of the road and the connection relationship between them. 2. According to the set grid size, the map is evenly divided to obtain a series of grid cells. 3. Each grid cell is traversed to determine whether it contains a road. For the grid cell containing a road, its connection with the road in the adjacent grid cell is analyzed. 4. According to the coherence of the road, the adjacent grid cells with close road connection are merged into one area. For some isolated grid cells with few road connections to other areas, they are merged with the adjacent cells. Finally, the original map can be divided into regions. It should be understood that this embodiment is not limited to the regional division method.

[0023] The target area is determined from the divided areas. The target area is the area for traffic flow correction. Then, the map update priority is determined according to the corrected traffic flow, so as to achieve map update of the target area. The target area can be an area with large traffic flow in each area. Areas with large traffic flow mean that users have a more urgent need for real-time road condition information. Timely updating of map information in these areas can provide more accurate navigation services.

[0024] The present invention is based on the vehicle-mounted terminal in the autonomous driving vehicle to obtain the real-time positioning information of the autonomous driving vehicle (i.e., coordinate information, specifically including longitude and latitude), such as the vehicle-mounted terminal is equipped with a GPS positioning system. In addition, the vehicle-mounted terminal can also record the driving trajectory, real-time driving speed, timestamp and other related information of the autonomous driving vehicle. It should be understood that the autonomous driving control and other technologies of the autonomous driving vehicle are conventional technologies, and are not the focus of the present invention, and will not be repeated.

[0025] Each autonomous driving vehicle sends its own positioning information and other relevant data information to the background system. The background system processes the data based on the positioning information of each autonomous driving vehicle and the various areas in the divided map to obtain the corrected traffic flow in the target area.

[0026] This embodiment analyzes the target area to obtain the corrected traffic flow of the target period in the target area. The length of the target period and the position in the day are set according to actual needs. In an exemplary embodiment, the target period selects a representative morning peak period of the day, and the length of the target period is 2 hours, such as 7 am to 9 am.

[0027] For the current day, this embodiment also presets a historical time period, which includes several consecutive days in the past, and the last day in the historical time period is the day before the current day. The target time period in each of the past days and the target time period of the current day are both the same time period in the day.

[0028] For any area, the number of autonomous driving vehicles entering the area during the target period is counted. The number of autonomous driving vehicles entering the area can be determined through the positioning information of each autonomous driving vehicle and the boundary coordinates of each area, which is used as the traffic flow of the area. The traffic flow of the target period of each day in the historical time period is used as the historical traffic flow of the area every day in the historical time period. The traffic flow of the target period of the day is used as the initial traffic flow of the area.

[0029] The traffic volume in urban traffic has the characteristic of dynamic propagation, that is, the change of traffic volume in a certain area will affect the traffic volume in adjacent or nearby areas. For example, an accident occurred on the main road of a region, which caused a surge in traffic volume on this road, causing congestion in the region, and may also cause an increase in traffic volume in surrounding areas, and eventually expand to more roads and regions. Therefore, when analyzing the traffic volume between a single area, the mutual influence between areas must also be considered. Therefore, for the target area, it is necessary to obtain the congested area of ​​the target area. The congested area is the area with vehicle congestion within the preset neighborhood of the target area. In an exemplary embodiment, Figure 2 As shown, a specific acquisition process of the congestion area is given as follows: Step 1-1: Obtain the associated region, which is the region within the preset neighborhood of the target region.

[0030] With the target area as the center, obtain other areas within the preset neighborhood of the target area. The preset neighborhood is specifically a circular area within a preset radius with the center point of the target area as the center of the circle, wherein the size of the preset radius is set according to the actual situation, such as the size of each area. If the preset radius is set too small, there will be too few associated areas, so that no effective associated areas can be obtained. If the preset radius is set too large, areas with weaker associations will be included, thereby affecting the accuracy and efficiency of data processing. In an exemplary embodiment, the setting of the preset radius needs to at least ensure that all other areas in contact with the target area are included as associated areas of the target area.

[0031] Step 1-2: Obtain the average speed of the autonomous driving vehicles in each associated area during the target period and compare it with the preset average speed threshold.

[0032] When congestion occurs in a certain area, the driving speed of the autonomous driving vehicles in this area is relatively low during the target time period. Therefore, the congested area can be determined by the driving speed of the autonomous driving vehicles in this area.

[0033] Specifically, for any associated area, the average speed of each autonomous driving vehicle in the associated area during the target period of the day is obtained, and then the average of the average speeds of all autonomous driving vehicles in the associated area during the target period is calculated as the average speed corresponding to the associated area. Thus, the average speed corresponding to each associated area is obtained.

[0034] An average speed threshold is preset. The purpose of setting the preset average speed threshold is to compare with the average speed corresponding to each associated area. If the average speed is less than the preset average speed threshold, it means that the average speed is low and there is congestion in the corresponding associated area. Therefore, the preset average speed threshold can be obtained by the overall level of the driving speed of the autonomous driving vehicles in the area when there is congestion in a large number of areas in history.

[0035] Step 1-3: If the average speed is less than a preset average speed threshold, the corresponding associated area is determined to be a congested area.

[0036] The average speed corresponding to each associated area is compared with a preset average speed threshold, and the associated area with an average speed less than the preset average speed threshold is obtained, and the associated area with an average speed less than the preset average speed threshold is determined as a congestion area, thereby obtaining the congestion area of ​​the target area.

[0037] After obtaining the congested area of ​​the target area, the impact of each congested area on the target area is obtained from the spatial and temporal dimensions.

[0038] When there is congestion in the congested area during the target time period, the autonomous driving vehicle may choose other routes. If the route selected by the autonomous driving vehicle passes through the target area, the traffic volume in the target area may increase, but the traffic volume in the target area on that day is more affected by the congested area, rather than the actual traffic volume in the target area. Assuming that area A is congested, the autonomous driving vehicle changes its route to pass through area B, resulting in an increase in the traffic volume in area B. However, the increase in traffic volume in area B is due to the congestion in area A, not the normal traffic volume in area B itself. Therefore, the redirection behavior of the autonomous driving vehicle, that is, the change in the navigation path planning of the autonomous driving vehicle, will cause changes in the traffic volume in the target area, specifically, the traffic volume is artificially high. Therefore, based on the changes in the navigation path planning, the spatial impact of each congested area on the target area is obtained.

[0039] The autonomous driving vehicle rerouting behavior, that is, the change in the autonomous driving vehicle's navigation path planning, is used to indicate the impact of the congested area on the target area. When an autonomous driving vehicle that originally planned to pass through area A is rerouted to area B, and the number of autonomous driving vehicles that originally planned to pass through area A actually passes through area B is greater, it means that area A has a greater impact on area B.

[0040] In an exemplary embodiment, Figure 3 As shown, a specific implementation process of obtaining the spatial impact of each congested area on the target area based on the change of navigation path planning is given as follows: Step 1-4: Obtain a first number of autonomous driving vehicles whose original navigation paths pass through the first congested area and do not pass through the target area during the target time period.

[0041] For the sake of convenience, the first congested area is defined as any congested area in the target area.

[0042] The original navigation path of each autonomous driving vehicle in the target time period of the day is obtained, where the original navigation path is a navigation path that is planned in advance and has not been adjusted, and the original navigation path is specifically a navigation path that passes through the first congested area and does not pass through the target area. Thus, the number of autonomous driving vehicles that set such an original navigation path in the target time period of the day is obtained as the first number.

[0043] Step 1-5: Obtain a second number of autonomous driving vehicles that change the original navigation path to the actual navigation path during the target period.

[0044] Due to the existence of congestion, some autonomous driving vehicles will change from passing through the first congested area to passing through the target area, that is, the autonomous driving vehicles do not drive according to the original navigation path, and the actual driving path has changed, thus becoming the actual navigation path, thus achieving a diversion. The actual navigation path is the actual driving path of the autonomous driving vehicle, and the actual navigation path is a navigation path that passes through the target area and does not pass through the first congested area. Therefore, the number of autonomous driving vehicles whose original navigation paths are changed to actual navigation paths during the target period of the day is obtained as the second number.

[0045] Therefore, the second quantity is determined from the first quantity, the first quantity contains the second quantity, and the first quantity is greater than the second quantity.

[0046] Step 1-6: Obtain a ratio of the second number to the first number as a spatial impact factor of the first congested area on the target area.

[0047] The spatial impact of the first congested area on the target area can be represented by the diversion rate. The diversion rate refers to the proportion of autonomous driving vehicles that should have passed through the first congested area that diverted to the target area. Therefore, the ratio of the second number to the first number is calculated, and this ratio is used as the diversion rate to indicate the degree of spatial impact of the congestion in the first congested area on the increase in traffic volume in the target area. The larger the ratio, the more significant the spatial impact of the congestion in the first congested area on the target area. Therefore, this ratio is used as the spatial impact factor of the first congested area on the target area. Since the first number is greater than the second number, the numerical range of the ratio is 0-1. Thus, the spatial impact factor of each congested area on the target area is obtained.

[0048] Step 2: Based on the time difference between the occurrence of congestion in each congested area and the target area, the correlation between the congestion phenomena in each congested area and the target area is obtained.

[0049] When congestion occurs in each congestion area, the target area may have obvious congestion after a period of time. At this time, the traffic flow in the target area may be caused by the propagation effect of the congestion area, that is, when congestion occurs in a certain area, the traffic congestion spreads from one area to another. This propagation effect makes the traffic flow in the target area likely to be affected by the congestion area.

[0050] However, this inter-regional propagation effect may have a time lag, that is, congestion in the congested area will not immediately affect the target area, and the impact may take some time to be transmitted to the target area through the transportation network. Therefore, when analyzing the time impact between regions, it is necessary to first determine the lag time for the congested area to cause congestion in the target area, that is, to determine the time difference between the occurrence of congestion in each congested area and the target area.

[0051] In an exemplary embodiment, Figure 4 As shown, a process for obtaining the time difference between the congestion in each congested area and the target area is given as follows: Step 2-1: Obtain the historical traffic volume of the first congested area and the target area.

[0052] Taking the first congested area as an example, according to the above description, the historical traffic volume of the first congested area and the target area during the target period of each day in the historical time period is obtained.

[0053] Step 2-2: Construct a linear regression model based on the relationship between the historical traffic volume and time in the first congested area and the target area.

[0054] Remember the target area is in the target period The historical traffic volume at time is , the oth congested area is in the target period The historical traffic volume at time is .set up represents the time difference between the oth congestion area and the target area congestion, then, the oth congestion area in the target period The historical traffic volume at time is .

[0055] Cross-analyze the traffic flow changes in the oth congested area and the traffic flow changes in the target area to find out the time delay between the two. Accordingly, the linear regression model is: ; in, represents the regression coefficient, Represents the error term of the linear regression model, and the least squares method (OLS) is used to estimate the model parameters.

[0056] Step 2-3: Solve the linear regression model to obtain the time difference between the first congested area and the target area when congestion occurs.

[0057] The changes in traffic volume between the first congested area and the target area are cross-analyzed. According to the relationship between traffic volume and time between regions given by the linear regression model, the linear regression model is solved to obtain the time difference between the congestion in the oth congested area and the target area. It should be understood that for the oth congestion area, since the time difference is obtained by the historical traffic volume of the target period of each day in the historical time period, if the time difference corresponding to each day in the historical time period is obtained, the final time difference corresponding to the oth congestion area can be obtained by calculating the average of the time difference. By using the above process, the time difference between each congestion area and the target area when congestion occurs can be obtained.

[0058] As other implementation methods, in addition to the above-mentioned time difference acquisition method, the following time difference acquisition method can also be used: for example: obtain the traffic volume of the first congestion area and the target area on the same day, and use the above-mentioned linear regression model analysis process according to the traffic volume of the day to obtain the time difference corresponding to the first congestion area. Or: in the target period of each day in the historical time period, count when congestion begins to appear in the first congestion area, and then count when congestion begins to appear in the target area, calculate the time interval between the two moments, and use it as the time difference between the first congestion area and the target area every day in the historical time period, and then calculate the average of the obtained time differences as the final time difference between the first congestion area and the target area.

[0059] The mutual influence between regions can be determined by time series analysis. When congestion occurs in the congested region, the target region will have congestion characteristics similar to those of the congested region when the obtained time difference arrives. The more the congestion characteristics of the congested region are correlated with the characteristics of the target region over time, the greater the influence of the congested region on the target region. Therefore, according to the obtained time difference, the correlation between the congestion phenomenon of each congested region and the target region is obtained. In an exemplary embodiment, Figure 5 As shown, a specific process of obtaining the correlation is given as follows: Step 2-4: Obtain a first traffic flow sequence in a first congestion time period of a first congestion area in a target time period.

[0060] Determine the time when the first congested area starts to be congested and the time when the congestion in the first congested area is relieved during the target period of the day. The time period between these two times is the congested time period, defined as the first congested time period. Then, the first congested time period is the congested time period of the first congested area during the target period.

[0061] An acquisition cycle is preset, and the time length of the acquisition cycle is determined by the length of the first congestion time period, thereby ensuring that the first congestion time period includes a plurality of acquisition cycles.

[0062] The vehicle flow of each acquisition cycle in the first congested area in the first congested time period of the day is obtained, thereby obtaining multiple vehicle flows in the first congested time period. Then, the vehicle flows are arranged in the order of the acquisition cycles to obtain a vehicle flow sequence, which is defined as the first vehicle flow sequence of the first congested area.

[0063] Step 2-5: Obtain a second congestion time period according to the first congestion time period and the time difference.

[0064] The second congestion time period is set as the congestion time period of the target area. Since there is a time lag between the congestion phenomenon in the first congestion area and the target area, the lag time is the time difference of the first congestion area obtained above. Therefore, the time when the first congestion area starts to be congested plus the time difference corresponding to the first congestion area is taken as the time when the congestion in the target area starts, and the time when the congestion in the first congestion area is relieved plus the time difference corresponding to the first congestion area is taken as the time when the congestion in the target area is relieved. The time period between the two obtained moments is the congestion time period of the target area, which is defined as the second congestion time period.

[0065] Step 2-6: Obtain a second traffic flow sequence of the target area in the second congestion time period.

[0066] After obtaining the second congestion time period, the traffic flow of each acquisition cycle of the target area in the second congestion time period of the day is obtained in the same way as the first traffic flow sequence of the first congestion area, according to the same preset acquisition cycle, so as to obtain multiple traffic flows in the second congestion time period. Then, the acquisition cycles are arranged in the order of the acquisition cycles to obtain a traffic flow sequence, which is defined as the second traffic flow sequence of the target area.

[0067] Step 2-7: Obtain the correlation coefficient between the first vehicle flow sequence and the second vehicle flow sequence.

[0068] Obtain the correlation coefficient between the first vehicle flow sequence and the second vehicle flow sequence. The correlation coefficient may be cosine similarity, Pearson correlation coefficient, etc. In this embodiment, the Pearson correlation coefficient algorithm is used to obtain the Pearson correlation coefficient between the first vehicle flow sequence and the second vehicle flow sequence as the correlation between the congestion phenomenon of the first congestion area and the target area. The greater the correlation, the greater the time impact of the first congestion area on the target area.

[0069] By adopting the above process, the correlation between the congestion phenomena of each congested area and the target area is obtained.

[0070] Step 3: Determine the time impact of each congested area on the target area. The time impact is obtained by correlation.

[0071] According to the correlation between the congestion phenomena of each congested area and the target area, the time impact of each congested area on the target area is obtained.

[0072] Taking the first congested area as an example, the Pearson correlation coefficient corresponding to the first congested area is normalized, and the normalized result is used as the time impact factor of the first congested area on the target area. The normalization method here can be: obtain the maximum Pearson correlation coefficient and the minimum Pearson correlation coefficient from the Pearson correlation coefficients corresponding to each congested area, and use the maximum and minimum normalization method to normalize the Pearson correlation coefficient corresponding to the first congested area. Thereby, the time impact factor of each congested area on the target area is obtained. It should be understood that the above normalization method is applicable to the situation where there are multiple congested areas. If the number of congested areas is relatively small, the following normalization method can be used: , It represents the processing object, and exp represents the exponential function with the natural constant e as the base.

[0073] Step 4: Modify the initial traffic flow in the target area based on spatial and temporal impacts.

[0074] Spatial impact and temporal impact are the effects of various congested areas on the traffic flow of the target area from the two dimensions of space and time. Then, the initial traffic flow of the target area is corrected by combining the spatial impact and the temporal impact. Figure 6 As shown, a specific correction process is given: Step 4-1: Based on the spatial impact and temporal impact, obtain the comprehensive impact factor of each congested area on the target area.

[0075] The impact of the congested area on the target area is jointly manifested in space and time. According to the spatial impact factor and the temporal impact factor, the comprehensive impact factor of each congested area on the target area is obtained.

[0076] In an exemplary embodiment, different congested areas may be affected by space or time to different degrees. For example, if the congested area is a traffic hub area, the possibility of autonomous driving vehicles changing routes is low (i.e., the spatial impact is weak), but the congestion propagation is strong (i.e., the temporal impact is strong). If the congested area is a branch road area, the possibility of autonomous driving vehicles detouring is high (i.e., the spatial impact is strong), but the congestion propagation range is limited (i.e., the temporal impact is weak). Therefore, it is necessary to determine the weights of spatial impact and temporal impact according to the actual situation of each congested area, so as to determine the dominance of the temporal impact factor and the spatial impact factor through the weights.

[0077] In an exemplary embodiment, Figure 7 As shown, a specific process of obtaining the comprehensive impact factor is given as follows: Step 4-1-1: Obtain the third number of autonomous driving vehicles whose driving paths pass through the first congested area during the target time period, and the fourth number of autonomous driving vehicles whose driving paths pass through the target area.

[0078] When a congested area is a frequently selected area by users, vehicles on frequently selected roads (such as main roads) have strong path dependence and tend to stick to the original route even when congested, resulting in more significant congestion propagation, that is, a higher degree of time impact.

[0079] Taking the first congested area as an example, during the target period of each day in the historical time period, the driving path of each autonomous driving vehicle is obtained (the driving path is the path actually driven by the autonomous driving vehicle), and the number of autonomous driving vehicles whose driving paths pass through the first congested area is obtained as the third number. Then the sum of the third numbers corresponding to each day in the historical time period is calculated as the final third number. The larger the final third number, the more the first congested area is a user-frequently selected area.

[0080] Similarly, during the target period of each day in the historical time period, the number of autonomous driving vehicles whose driving paths pass through the target area is obtained as the fourth number. Then, the sum of the fourth numbers corresponding to each day in the historical time period is calculated as the final fourth number.

[0081] Step 4-1-2: Obtain a ratio of the third number to the total number as a weight of the time impact corresponding to the first congested area.

[0082] The sum of the final third number and the final fourth number is calculated as the total number. Then the ratio of the final third number to the total number is calculated, and the ratio is used as the weight of the time impact corresponding to the first congested area. The larger the ratio, the more likely the first congested area is to be a popular area. At this time, the spatial impact factor of the user's route change behavior is small, and the time impact factor is large, that is, the time impact of the first congested area is more dominant.

[0083] By adopting the above method, the weight of the time impact corresponding to each congested area is obtained.

[0084] Step 4-1-3: According to the weights, perform weighted summation of the spatial impact factor and the temporal impact factor to obtain the comprehensive impact factor of the first congested area on the target area.

[0085] The obtained weight is the weight of the time impact factor of the first congested area, and the weight of the space impact factor of the first congested area is obtained according to the weight. The weight of the space impact factor is the value 1 minus the weight.

[0086] In an exemplary embodiment, the calculation formula of the comprehensive impact factor is as follows: ; in, is the comprehensive impact factor of the oth congested area on the target area, is the spatial impact factor of the oth congested area on the target area, is the time impact factor of the oth congested area on the target area, is the weight of the time impact factor obtained in step 4-1-1 and step 4-1-2.

[0087] Through the above method, the comprehensive impact factor of each congested area on the target area is obtained.

[0088] As another implementation method, if the different degrees of spatial or temporal influence on different congested areas are not taken into consideration, the average value of the spatial influence factor and the temporal influence factor may be directly calculated as the comprehensive influence factor.

[0089] Step 4-2: Combine the comprehensive impact factors of all congested areas on the target area to obtain the final impact factor.

[0090] Based on the comprehensive impact factors of each congested area on the target area, the comprehensive impact factors of all congested areas on the target area are integrated to obtain a final impact factor. In an exemplary embodiment, the average of the comprehensive impact factors of each congested area on the target area is calculated as the final impact factor.

[0091] Step 4-3: Determine the correction factor, which is negatively correlated with the final impact factor.

[0092] The final impact factor characterizes the impact of all congested areas on the traffic volume of the target area. The larger the final impact factor, the more serious the impact on the traffic volume of the target area, that is, the more serious the false high degree of the traffic volume of the target area, the greater the difference between the actual traffic volume of the target area and the initial traffic volume, the more it is necessary to reduce the initial traffic volume of the target area. Then, according to the obtained final impact factor, a correction factor is obtained, and the correction factor is negatively correlated with the final impact factor. In an exemplary embodiment, the difference between the value 1 and the final impact factor is calculated, and the difference is the correction factor.

[0093] Step 4-4: Multiply the correction factor by the initial traffic flow of the target area to obtain the corrected traffic flow of the target area.

[0094] The product of the correction factor and the initial traffic flow of the target area is calculated to obtain the corrected traffic flow of the target area. The corrected traffic flow is the actual traffic flow of the target area on that day, thereby accurately obtaining the traffic flow of the target area.

[0095] According to the data processing process provided by this embodiment, the real traffic volume of each area on that day can be obtained. The greater the traffic volume, the higher the map update priority. Then, the map update priority of each area can be obtained according to the real traffic volume of each area, so as to realize the navigation map update according to the map update priority of each area.

[0096] This embodiment also provides an intelligent update system for an autonomous driving navigation map, including: a memory and a processor; the memory is connected to the processor, and the memory is used to store program instructions; the processor is used to implement the steps in the above-mentioned embodiment of the intelligent update method for the autonomous driving navigation map when the program instructions are executed.

[0097] In an exemplary embodiment, the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps in the above-mentioned embodiment of the intelligent update method of the autonomous driving navigation map.

[0098] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0099] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for intelligently updating an autonomous driving navigation map, characterized in that: include: Based on the change of the navigation path planning, the spatial impact of each congested area on the target area is obtained, and the congested area is an area with vehicle congestion within a preset neighborhood of the target area; Combined with the time difference between the occurrence of congestion in each congested area and the target area, the correlation between the congestion phenomenon in each congested area and the target area is obtained; determining a time impact of each congested area on the target area, the time impact being obtained from the correlation; The initial vehicle flow in the target area is modified according to the spatial influence and the temporal influence.

2. The intelligent updating method of the autonomous driving navigation map according to claim 1, characterized in that: The process of obtaining the congested area includes: Acquire a related area, where the related area is an area within a preset neighborhood of the target area; Obtain the average speed of the autonomous driving vehicles in each associated area during the target period and compare it with the preset average speed threshold; If the average speed is less than a preset average speed threshold, the corresponding associated area is determined to be a congested area.

3. The intelligent updating method of the autonomous driving navigation map according to claim 1, characterized in that: The obtaining of the spatial impact of each congested area on the target area based on the change of the navigation path planning includes: Acquire a first number of autonomous driving vehicles whose original navigation paths pass through a first congested area and do not pass through the target area during a target time period; the first congested area is any congested area; Acquire a second number of autonomous driving vehicles that change an original navigation path into an actual navigation path during the target time period; the actual navigation path is a navigation path that passes through the target area and does not pass through the first congested area; A ratio of the second number to the first number is obtained as a spatial impact factor of the first congested area on the target area.

4. The intelligent updating method of an autonomous driving navigation map according to claim 1, characterized in that: The process of obtaining the time difference between the occurrence of congestion in each congested area and the target area includes: Obtaining historical vehicle flow in a first congestion area and a target area; the first congestion area is any congestion area; According to the relationship between the historical traffic volume and time in the first congestion area and the target area, a linear regression model is constructed; The linear regression model is solved to obtain the time difference between the occurrence of congestion in the first congested area and the target area.

5. The intelligent updating method of the autonomous driving navigation map according to claim 4, characterized in that: The linear regression model is: ； in, Indicates that the target area is Historical traffic volume over time, Indicates that the oth congested area is Historical traffic volume over time, represents the time difference between the congestion in the oth congested area and the target area, represents the regression coefficient, represents the error term of the linear regression model.

6. The intelligent updating method of an autonomous driving navigation map according to claim 1, characterized in that: The process of obtaining the correlation between the congestion phenomena of each congested area and the target area includes: Acquire a first vehicle flow sequence of a first congestion time period in a first congestion area in a target time period; the first congestion area is any congestion area, and the first congestion time period is a congestion time period in the first congestion area in the target time period; According to the first congestion time period and the time difference, a second congestion time period is obtained, where the second congestion time period is a congestion time period of a target area; Acquire a second traffic flow sequence of the target area in the second congestion time period; Obtain a correlation coefficient between the first vehicle flow sequence and the second vehicle flow sequence.

7. The intelligent updating method of the autonomous driving navigation map according to claim 6, characterized in that: The determining of the time impact of each congested area on the target area includes: A result of normalizing the correlation coefficient corresponding to the first congested area is used as a time influence factor of the first congested area on the target area.

8. The intelligent updating method of an autonomous driving navigation map according to claim 1, characterized in that: The modifying of the initial vehicle flow of the target area according to the spatial influence and the temporal influence includes: According to the spatial impact and the temporal impact, a comprehensive impact factor of each congested area on the target area is obtained; The comprehensive impact factors of all congested areas on the target area are integrated to obtain the final impact factor; Determining a correction factor, wherein the correction factor is negatively correlated with the final impact factor; The correction factor is multiplied by the initial vehicle flow of the target area to obtain the corrected vehicle flow of the target area.

9. The intelligent updating method of the autonomous driving navigation map according to claim 8, characterized in that: The spatial impact is specifically a spatial impact factor, and the temporal impact is specifically a temporal impact factor; The comprehensive impact factor of each congested area on the target area is obtained according to the spatial impact and the temporal impact, including: Obtaining a third number of autonomous driving vehicles whose driving paths pass through a first congestion area during a target period, and a fourth number of autonomous driving vehicles whose driving paths pass through a target area; the first congestion area being any congestion area; Obtaining a ratio of the third number to the total number as a weight of the time impact corresponding to the first congested area; the total number is the sum of the third number and the fourth number; According to the weights, the spatial impact factor and the temporal impact factor are weighted and summed to obtain a comprehensive impact factor of the first congested area on the target area.

10. An intelligent updating system for an autonomous driving navigation map, characterized in that it includes: Memory and processor; The memory is connected to the processor; The memory is used to store program instructions; The processor is used to implement the intelligent update method of the autonomous driving navigation map according to any one of claims 1 to 9 when the program instructions are executed.

Images