A method and device for processing geographic information data

Abstract

The present invention relates to the field of intelligent transportation, and provides a method and device for processing geographic information data, wherein the method includes: obtaining geographic information data of a target area within a predetermined time range; calculating multiple data sets of a traffic feature set and target feature values ​​of each data set based on the geographic information data; using a predetermined physical knowledge base, once screening the features in the traffic feature set to obtain a spare traffic feature set, the physical knowledge base includes multiple feature sets with physical association relationships; based on the strong correlation between the features in the spare traffic feature set and the strong correlation between the features in the spare traffic feature set and the target features, performing a secondary screening on the features in the spare traffic feature set to obtain an optimal traffic feature set. The present invention can comprehensively and quickly determine the optimal traffic feature set that affects the target features, and is suitable for scenarios where computing resources on a satellite are limited.

Description

Technical Field

[0001] The present invention relates to the field of intelligent transportation, and in particular to a method and device for processing geographic information data. Background Art

[0002] The commonly used feature selection method is the filtering method. Specifically, the filtering method is to screen the features of the data before model training and operation, and there is no interaction between the feature screening and the modeling process. This method can avoid the contamination of the model data on the original data. The screening criteria of the filtering method mainly include distance criterion, information criterion, correlation criterion and consistency criterion. The distance criterion uses the distance criterion to measure the similarity between the samples in the data set, and then characterizes the contribution or effect of each feature in the feature set to the target feature. The information criterion uses attributes such as information entropy to measure the degree of uncertainty of each feature in the data set for the target feature. The higher the degree of uncertainty of the feature, the greater the help to the target variable. The correlation criterion measures the importance of each feature by measuring the degree of correlation between the feature and the target feature. The consistency criterion uses the inconsistency rate of the data sample to measure the resolution ability of the candidate feature subset. The inconsistency rate is the ratio of the number of inconsistent samples to the total number of samples.

[0003] The filtering feature selection method has the problem of high time and space complexity, which is not conducive to scenarios with limited resources such as on-board satellite computing. Summary of the invention

[0004] From the perspective of data processing, existing technologies have the disadvantages of high time and space complexity, which is not conducive to scenarios with limited resources such as on-board satellite computing.

[0005] In order to solve the above technical problems, the present invention provides a method for processing geographic information data in a first aspect, comprising:

[0006] Obtain geographic information data of the target area within a predetermined time range;

[0007] Calculating multiple data sets of traffic feature sets and target feature values ​​of each data set based on the geographic information data;

[0008] Using a predetermined physical knowledge base, screening the features in the traffic feature set to obtain a spare traffic feature set, wherein the physical knowledge base includes a plurality of feature sets having a physical association relationship;

[0009] According to the strong correlation between the features in the backup traffic feature set and the strong correlation between the features in the backup traffic feature set and the target features, the features in the backup traffic feature set are secondary screened to obtain the optimal traffic feature set, and a target feature prediction model is established based on the optimal traffic feature set and the target features.

[0010] As a further embodiment of the present invention, based on the strong correlation between the features in the spare traffic feature set and the strong correlation between the features in the spare traffic feature set and the target features, performing secondary screening on the features in the spare traffic feature set includes:

[0011] For each first traffic feature in the spare traffic feature set, determining the correlation between the first traffic feature and the target feature according to the data set of the first traffic feature and the target feature value of the data set; when the first traffic feature is not correlated with the target feature, deleting the first traffic feature from the spare traffic feature set;

[0012] For every two traffic features in the spare traffic feature set, the correlation between the two traffic features is determined according to the data sets of the two traffic features; when the two traffic features are correlated, one of the traffic features is deleted from the two traffic features.

[0013] As a further embodiment of the present invention, when the first traffic feature is related to the target feature, traffic features are traversed from the backup traffic feature set, and the traversed traffic features are used as the second traffic feature;

[0014] The first traffic feature and the second traffic feature are used as the two traffic features, the correlation between the two traffic features is calculated, and when the two traffic features are correlated, the first traffic feature is deleted from the spare traffic feature set.

[0015] As a further embodiment of the present invention, when the two traffic characteristics are unrelated, the steps of traversing the next second traffic characteristic from the backup traffic characteristic set and subsequent steps are repeated until the second traffic characteristic cannot be traversed, and then the next first traffic characteristic is traversed from the backup traffic characteristic set and the correlation between the first traffic characteristic and the target characteristic is repeatedly determined.

[0016] As a further embodiment of the present invention, the traffic features in the traffic feature set include at least: traffic operation features corresponding to the road section i at time t, traffic design features corresponding to the road section i at time t, and traffic facility features corresponding to the road section i at time t;

[0017] The target features include: the travel time of section i at time t+1 or traffic conditions.

[0018] As a further embodiment of the present invention, the physical knowledge base establishment process includes:

[0019] Acquiring physical knowledge information of the traffic characteristics and physical knowledge information of the target characteristics;

[0020] De-duplicate the acquired physical knowledge information;

[0021] Filtering target physical knowledge information from the deduplicated physical knowledge information, wherein the target physical knowledge information includes at least two traffic characteristics, and / or one traffic characteristic and a target characteristic;

[0022] Extracting the correlation relationship between physical quantities from the target physical knowledge information;

[0023] Traffic features with physical quantity correlation, and traffic features and target features with physical quantity correlation constitute a physical knowledge set.

[0024] As a further embodiment of the present invention, a predetermined physical knowledge base is used to screen the features in the traffic feature set, including:

[0025] Matching the traffic features in the traffic feature set with the features in the feature set of the physical knowledge base one by one, and recording the number of traffic feature matches;

[0026] For each feature set in the physical knowledge base, if the feature set has at least two matching traffic features, the traffic feature with the smallest number of matches is retained in the traffic feature set.

[0027] As a further embodiment of the present invention, before performing a screening, the method further comprises:

[0028] Randomly selecting a predetermined number of traffic features from the traffic feature set to form a candidate feature subset, and deleting the traffic features in the candidate feature subset from the traffic feature set;

[0029] The primary screening and secondary screening processes are replaced by:

[0030] Performing a primary screening and a secondary screening on the features in the candidate feature subset, and recording the number of deleted traffic features;

[0031] When all features in the candidate feature subset have been analyzed and the traffic feature set is not empty, traffic features of the number of deleted traffic features are randomly selected from the traffic feature set and moved to the candidate feature subset, and the number of deleted traffic features is reset to zero, and the process of primary and secondary screening of the candidate feature subset is repeated.

[0032] As a further embodiment of the present invention, when all features in the candidate feature subset have been analyzed and the number of deleted traffic features is zero and the traffic feature set is not empty, at least one traffic feature is selected from the traffic feature set and moved to the candidate feature subset, and the selected traffic feature is deleted from the traffic feature set, and the process of primary and secondary screening of the candidate feature subset is repeated.

[0033] As a further embodiment of the present invention, a predetermined number of traffic features are randomly selected from the traffic feature set to form a candidate feature subset, including:

[0034] Get performance data of distributed servers;

[0035] Determine the characteristic processing capacity of each server based on the performance data of each server;

[0036] According to the feature processing capacity of each server, features are obtained from the traffic feature set to obtain a candidate feature subset of each server;

[0037] The candidate feature subsets of each server are sent to each server for analysis, so that each server can screen the candidate feature subsets, and count the number of traffic features deleted from the candidate feature subsets, and obtain traffic features from the traffic feature set based on the number of traffic features deleted by each server and the performance of each server.

[0038] As a further embodiment of the present invention, after obtaining the geographic information data, the method further includes:

[0039] Preprocessing geographic information data, including:

[0040] Smoothing the non-binary data in geographic information data;

[0041] Carry out coordinate system transformation on spatial data in geographic information data;

[0042] Clean the time-varying data in geographic information data for spatiotemporal redundancy;

[0043] Complete the default data.

[0044] A second aspect of the present invention provides a device for processing geographic information data, comprising:

[0045] A data acquisition unit, used to acquire geographic information data of a target area within a predetermined time range;

[0046] A calculation unit, used for calculating a plurality of data sets of traffic characteristic sets and a target characteristic value of each data set according to the geographic information data;

[0047] A first screening unit, configured to screen the features in the traffic feature set using a predetermined physical knowledge base to obtain a backup traffic feature set, wherein the physical knowledge base includes a plurality of feature sets having a physical association relationship;

[0048] The second screening unit is used to perform secondary screening on the features in the backup traffic feature set according to the strong correlation between the features in the backup traffic feature set and the strong correlation between the features in the backup traffic feature set and the target features, so as to obtain an optimal traffic feature set, and establish a target feature prediction model according to the optimal traffic feature set and the target features.

[0049] A third aspect of the present invention provides a navigation system, comprising: an unmanned aerial device, a satellite server and a client;

[0050] The unmanned aerial device and the satellite server are installed with a travel time prediction model, the target feature of the travel time prediction model is the travel time of the road section, and the traffic feature set of the travel time prediction model is determined by the method described in any of the above embodiments;

[0051] The client is used for allowing the user to input a destination and detect satellite signals. When a satellite signal is detected, the destination information is sent to a satellite server, so that the satellite server uses the travel time prediction model to calculate the travel time to the destination and returns it to the client. When no satellite signal is detected, the destination information is sent to a nearby unmanned aerial device, so that the unmanned aerial device uses the travel time prediction model to calculate the travel time to the destination and returns it to the client.

[0052] The method and device for processing geographic information data provided by the present invention use a physical knowledge base to perform a primary screening of features in a traffic feature set to obtain a backup traffic feature set, perform a secondary screening of features in the backup traffic feature set based on strong correlations between features in the backup traffic feature set and strong correlations between features in the backup traffic feature set and target features to obtain an optimal traffic feature set, and establish a target feature prediction model based on the optimal traffic feature set and the target features, thereby reducing the complexity of the traffic feature selection algorithm for geographic information data, increasing its applicable scenarios, ensuring calculation accuracy under limited resources, and reducing feature redundancy.

[0053] In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0055] Figure 1 A first flow chart showing a method for processing geographic information data according to an embodiment of the present invention;

[0056] Figure 2 A flow chart showing a secondary screening process of traffic features according to an embodiment of the present invention is shown;

[0057] Figure 3 A flow chart showing a process of establishing a physical knowledge base according to an embodiment of the present invention is shown;

[0058] Figure 4 A flow chart showing a process of screening traffic features in an embodiment of the present invention;

[0059] Figure 5 A second flow chart showing the method for processing geographic information data according to an embodiment of the present invention;

[0060] Figure 6 A flowchart of a process for determining a candidate feature subset according to an embodiment of the present invention is shown;

[0061] Figure 7 A first structural diagram of a device for processing geographic information data according to an embodiment of the present invention is shown;

[0062] Figure 8 A second structural diagram of a device for processing geographic information data according to an embodiment of the present invention is shown;

[0063] Fig. 9 A structural diagram of a navigation system according to an embodiment of the present invention is shown;

[0064] Fig.10 A schematic diagram of the Pearson correlation coefficient results between binary features and other non-binary features in a specific embodiment of the present invention is shown;

[0065] Fig.11 A schematic diagram of the Pearson correlation coefficient results between non-binary features and target features in a specific embodiment of the present invention is shown;

[0066] Fig.12 A schematic diagram of a neural network architecture of a specific embodiment of the present invention is shown;

[0067] Fig.13 A schematic diagram of a dynamic graph generator according to a specific embodiment of the present invention is shown;

[0068] Fig.14 A schematic diagram of a dynamic graph convolution module according to a specific embodiment of the present invention is shown;

[0069] Fig.15 A schematic diagram of an error distribution diagram of a specific embodiment of the present invention is shown;

[0070] Fig.16A A schematic diagram of a local road network according to a specific embodiment of the present invention is shown;

[0071] Fig. 16B A traffic state prediction diagram of a local road network according to a specific embodiment of the present invention is shown;

[0072] Fig.17 A structural diagram of a computer device according to an embodiment of the present invention is shown.

[0073] Description of the accompanying symbols:

[0074] 701, data acquisition unit;

[0075] 702, computing unit;

[0076] 703, first screening unit;

[0077] 704, second screening unit;

[0078] 705, split unit;

[0079] 706. Update control unit;

[0080] 901. Unmanned aerial vehicle;

[0081] 902, satellite server;

[0082] 903, client;

[0083] 1702. Computer equipment;

[0084] 1704, processor;

[0085] 1706. Memory;

[0086] 1708, driving mechanism;

[0087] 1710, input / output module;

[0088] 1712. Input devices;

[0089] 1714. Output devices;

[0090] 1716. Presentation equipment;

[0091] 1718. Graphical user interface;

[0092] 1720, network interface;

[0093] 1722, communication link;

[0094] 1724. Communication bus. DETAILED DESCRIPTION

[0095] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0096] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, device, product or equipment that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0097] The term " cycle"," "Moment" and the like are used to refer generally to objects that meet specific conditions, not to a specific object. It should be understood that the data used by such objects may be changed in order and other aspects under appropriate circumstances, so that the embodiments of the present invention described herein can be implemented in similar scenarios other than those shown or described.

[0098] This specification provides method operation steps as described in the embodiments or flow charts, but more or fewer operation steps may be included based on routine or non-creative work. The order of steps listed in the embodiments is only one way of executing the order of many steps and does not represent the only execution order. When the system or device product is executed in practice, it can be executed in the order of the method shown in the embodiments or the drawings or in parallel.

[0099] It should be noted that the geographic information involved in this application (including but not limited to data used for analysis, stored data, displayed data, etc.) are all information and data authorized by the user or fully authorized by all parties, and the acquisition, transmission, storage, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions.

[0100] It should be noted that in the embodiments of the present application, certain software, components, models and other existing solutions in the industry may be mentioned, and they should be regarded as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the solution.

[0101] In the prior art, the feature selection for modeling usually adopts the filtering method, which directly screens the initial features by distance criteria, information criteria, correlation criteria and consistency criteria. It has the problem of high time and space complexity and is not conducive to scenarios with limited resources such as satellite on-board computing. In addition, this method does not consider the physical characteristics between features, and there are redundant features, which leads to poor modeling effect.

[0102] In order to solve the above technical problems, the present invention provides a method for processing geographic information data. It should be noted that the method for processing geographic information data in the attached drawings can be executed in a computer system including executable instructions, in a simulation system using computer simulation, in an embedded device, in a set of interconnected intelligent devices, in a distributed computing system, and in other executable carriers. Figure 1 As shown, the method for processing geographic information data includes:

[0103] Step 101, obtaining geographic information data of a target area within a predetermined time range.

[0104] The current area can be set by the user according to the actual application area, and the predetermined time range can be set by the user according to actual needs. The present invention does not specifically limit the target area and the predetermined time range.

[0105] Geographic information data refers to geographic information data obtained by independent means or mixed means such as satellite remote sensing, geographic surveying and mapping, and aerial surveying and mapping. Geographic information data includes spatial information that does not change over time in the target area and traffic information that changes over time. Among them, spatial information includes spatial boundaries, regional shapes, etc. Traffic information includes information related to vehicles on the road. Specifically, geographic information data includes specific items such as longitude, latitude, and object attributes, and the data structure forms include vector data structure, raster data structure, and vector-raster integrated data structure.

[0106] Step 102: Calculate multiple data sets of traffic feature sets and target feature values ​​of each data set based on geographic information data.

[0107] When this step is implemented, traffic features are first extracted based on geographic information data, and the extracted traffic features form a traffic feature set. The traffic features in the present invention are all features involved in actual traffic scenarios, including traffic operation features, traffic design features, and traffic facility features at each time and each road section, such as road length, road width, number of road lanes, road restrictions, road type, traffic density, traffic flow speed, travel time, traffic signal type, travel mode, traffic signal control method, intersection area, and other specific traffic variables and geographic information variables.

[0108] The traffic feature extraction process includes:

[0109] (1) Static feature extraction: Static features are variables whose values ​​do not change over time, such as intersection area, number of road lanes, road length, etc. The static feature data set calculation method is to use the spatial data in the geographic information data, according to the spatial boundaries and regional shapes, through the coordinates of the boundary intersections, to calculate or count the static variable values.

[0110] (2) Dynamic feature extraction: Dynamic features are variables whose values ​​change over time, such as traffic density, traffic flow, number of users, etc. The data set calculation methods for dynamic features are divided into instantaneous sampling and spatiotemporal statistics. The instantaneous sampling is to obtain the variable value according to the spatial range at a certain moment, and the spatiotemporal statistics is to calculate the spatiotemporal statistics within a certain time interval.

[0111] The target feature is a quantity preset by the user based on traffic characteristics, such as the travel time or traffic status of a road section at the predicted time. The traffic status includes: smooth traffic, slow traffic, and congestion. In a specific embodiment, the traffic status classification method is shown in Table 1 below:

[0112] Table 1

[0113]

[0114] In other specific embodiments, the traffic status may be further divided according to speed, road type and traffic flow.

[0115] The traffic feature set obtained in this step is the original feature set. The multiple data sets of the traffic feature set are traffic feature entry data sets arranged according to timestamps, as shown in Table 2 below:

[0116] Table 2

[0117]

[0118] Step 103, using a predetermined physical knowledge base, screen the features in the traffic feature set to obtain a spare traffic feature set. The physical knowledge base is shown in Table 3 below, including multiple feature sets with physical association relationships.

[0119] Table 3

[0120]

[0121] The purpose of this step is to delete features associated with physical knowledge, thereby reducing the number of features for subsequent secondary screening and reducing feature redundancy.

[0122] Step 104, based on the strong correlation between the features in the spare traffic feature set and the strong correlation between the features in the spare traffic feature set and the target features, the features in the spare traffic feature set are screened again to obtain the optimal traffic feature set. After obtaining the optimal traffic feature set, the target feature prediction model can be established using the optimal traffic feature set and the target features. The features in the optimal traffic feature set obtained in this embodiment are features that have an impact on the target features and have no similarity between the features.

[0123] In detail, the strong correlation calculation method includes the t-test method and the Pearson correlation coefficient analysis method. Before implementing this method, the lower bound of the strong correlation corresponding to the t-test method needs to be pre-configured. θ t , in the t-test method, the incoherence corresponds to the upper bound of the correlation coefficient, and the strong coherence corresponds to the lower bound of the correlation coefficient in the Pearson correlation coefficient analysis method θ p , the upper bound of the incoherent corresponding correlation coefficient in the Pearson correlation coefficient analysis method τ p The process of determining strong correlation using the t-test method and the Pearson correlation coefficient analysis method can be referred to in the subsequent embodiments and will not be described in detail here.

[0124] In the embodiment of this step, it is first determined whether the traffic feature is a binary feature. If it is a binary feature, the strong correlation is calculated using the t-test method. If it is not a binary feature, the strong correlation is calculated using the Pearson correlation coefficient analysis method.

[0125] This embodiment uses a physical knowledge base to perform a primary screening of features in a traffic feature set to obtain a backup traffic feature set, and based on the strong correlation between features in the backup traffic feature set and the strong correlation between features in the backup traffic feature set and target features, performs a secondary screening of features in the backup traffic feature set to obtain an optimal traffic feature set. This can reduce the complexity of the traffic feature selection algorithm for geographic information data, improve feature extraction accuracy, increase its applicable scenarios, ensure calculation accuracy under limited resources, and reduce feature redundancy.

[0126] In one embodiment of the present invention, after obtaining the geographic information data in step 101, the following steps are further included:

[0127] Preprocessing the geographic information data includes: smoothing the non-binary data in the geographic information data; performing coordinate system transformation on the spatial data in the geographic information data; cleaning the spatiotemporal redundancy of the time-varying data in the geographic information data; and completing the default data.

[0128] This embodiment can improve the efficiency and accuracy of subsequent data analysis by preprocessing geographic information data.

[0129] In one embodiment of the present invention, the above step 104, based on the strong correlation between the features in the spare traffic feature set and the strong correlation between the features in the spare traffic feature set and the target feature, performs secondary screening on the features in the spare traffic feature set, including:

[0130] For each first traffic feature in the spare traffic feature set, determining the correlation between the first traffic feature and the target feature according to the data set of the first traffic feature and the target feature value of the data set; when the first traffic feature is not correlated with the target feature, deleting the first traffic feature from the spare traffic feature set;

[0131] For every two traffic features in the spare traffic feature set, the correlation between the two traffic features is determined according to the data sets of the two traffic features; when the two traffic features are correlated, one of the traffic features is deleted from the two traffic features.

[0132] In a further embodiment, when the first traffic feature is associated with the target feature, traversing a second traffic feature from the set of backup traffic features;

[0133] The first traffic feature and the second traffic feature are used as the two traffic features, the correlation between the two traffic features is calculated, and when the two traffic features are correlated, the first traffic feature is deleted from the spare traffic feature set.

[0134] When the two traffic features are unrelated, repeat traversing the next second traffic feature from the backup traffic feature set and subsequent steps until the second traffic feature cannot be traversed, then traverse the next first traffic feature from the backup traffic feature set and repeatedly determine the correlation between the first traffic feature and the target feature.

[0135] In one embodiment of the present invention, Figure 2As shown, the above step 104 performs secondary screening of the features in the spare traffic feature set according to the strong correlation between the features in the spare traffic feature set and the strong correlation between the features in the spare traffic feature set and the target feature, including:

[0136] Step 201: randomly select a first traffic feature from a backup traffic feature set.

[0137] Step 202 : determining the correlation between the first traffic feature and the target feature according to the data set of the first traffic feature and the target feature value of the data set.

[0138] Step 203: If the first traffic feature is not relevant to the target feature, the first traffic feature is deleted from the spare traffic feature set, and the process returns to step 201 to continue to retrieve the next first traffic feature until all traffic features have been traversed.

[0139] Step 204: If the first traffic feature is related to the target feature, a second traffic feature is randomly selected from the backup traffic feature set.

[0140] Step 205 : determining the correlation between the first traffic feature and the second traffic feature according to the data set of the first traffic feature and the data set of the second traffic feature.

[0141] Step 206: If the first traffic feature is related to the second traffic feature, the first traffic feature is deleted from the spare traffic feature set, and the process returns to step 201 to continue to extract the first traffic feature until all traffic features have been traversed.

[0142] If the first traffic feature is not related to the second traffic feature, the process returns to step 204 to continue extracting a new second traffic feature until all traffic features have been traversed, that is, all traffic features have been extracted as second traffic features.

[0143] This embodiment first calculates the correlation between the traffic characteristics and the target characteristics, and then calculates the traffic characteristics related to the traffic characteristics related to the target characteristics, so that the traffic characteristics related to the target characteristics can be quickly determined in a hierarchical manner and redundant characteristics can be reduced.

[0144] In one embodiment of the present invention, it can be applied to an emergency traffic dispatch scenario. Correspondingly, the traffic features in the traffic feature set in the scenario include: traffic operation features corresponding to section i at time t, traffic design features corresponding to section i at time t, and traffic facility features corresponding to section i at time t.

[0145] Among them, the traffic operation characteristics corresponding to the road section i at time t include: the traffic density corresponding to the road section i at time t , Traffic flow rate , average traffic speed , time occupancy , Space Occupancy , travel time wait.

[0146] Specifically, the traffic density of a road section: the number of vehicles on the road section / (road section length*number of lanes);

[0147] Traffic flow rate of a road section: number of vehicles at the road observation end / observation time, where vehicles can be observed through image acquisition equipment or remote sensing GIS;

[0148] Traffic speed of a road section: length of the road / time for vehicles to pass through the road;

[0149] Average traffic speed of a road section: the average of the vehicle traffic speeds of the road section;

[0150] Time occupancy rate of a road section: ratio of time when a road section has vehicles to observation time;

[0151] Spatial occupancy of a road section: area with cars / area without cars.

[0152] The traffic design characteristics corresponding to the road section i at time t include: the road congestion density corresponding to the road section i at time t ρ m , road critical density ρ c , road capacity q m , free flow speed on the road v f wait.

[0153] Specifically, we first fit the relationship diagram between traffic density and traffic flow rate based on the observed data, i.e., the basic traffic flow diagram, in which traffic density is the horizontal axis and traffic flow rate is the vertical axis. Then, we analyze the maximum value of the traffic flow rate in the relationship diagram to find the road capacity, the traffic density corresponding to the maximum traffic flow rate is the road critical density, the derivative when the traffic density is zero is the road free flow speed, and the intersection with the horizontal axis is non-zero as the road congestion density.

[0154] Traffic facility characteristics include: the road width corresponding to section i at time t η i , Road length l i , Road area α i , Traffic signal type s i , Area Type β i wait.

[0155] The target feature is the travel time of section i at time t+1 or traffic conditions.

[0156] In one embodiment of the present invention, Figure 3 As shown, the physical knowledge base establishment process in the above step 103 includes:

[0157] Step 301, obtaining physical knowledge information of traffic characteristics and physical knowledge information of target characteristics.

[0158] When implementing this step, the designer or relevant staff may pre-collect physical knowledge information related to traffic characteristics and multiple pieces of physical knowledge information related to target characteristics, and may also automatically capture physical knowledge information from an existing database based on traffic characteristics and target characteristics. The present invention does not specifically limit the process of obtaining physical knowledge information.

[0159] Step 302: De-duplicate the acquired physical knowledge information.

[0160] Through deduplication processing, the processing workload of subsequent steps can be reduced and the efficiency of establishing the physical knowledge base can be improved.

[0161] Step 303: Filter out target physical knowledge information from the physical knowledge information after deduplication, where the target physical knowledge information contains at least two traffic characteristics and / or one traffic characteristic and a target characteristic.

[0162] Step 304: extracting the correlation between physical quantities from the target physical knowledge information.

[0163] Step 305: The traffic features having physical quantity association relationships, and the traffic features and target features having physical association relationships are combined into a physical knowledge set.

[0164] In specific implementation, after analyzing the physical knowledge set through the above method, experts can also review and adjust the analysis results to form a final physical knowledge base.

[0165] For the above emergency traffic dispatch scenario, the physical knowledge set formed is shown in Table 4 below:

[0166] Table 4

[0167]

[0168] In one embodiment of the present invention, Figure 4 As shown, the above step 103 uses a predetermined physical knowledge base to screen the features in the traffic feature set, including:

[0169] Step 401, matching the traffic features in the traffic feature set with the features in the feature set of the physical knowledge base one by one, and recording the number of times the traffic features are matched.

[0170] Step 402: For each feature set in the physical knowledge base, if the feature set has at least two matching traffic features, the traffic feature with the smallest number of matches is retained in the traffic feature set.

[0171] In one embodiment of the present invention, before performing a screening, the method further comprises:

[0172] Randomly selecting a predetermined number of traffic features from the traffic feature set to form a candidate feature subset, and deleting the traffic features in the candidate feature subset from the traffic feature set;

[0173] The primary screening and secondary screening processes are replaced by:

[0174] Performing a primary screening and a secondary screening on the features in the candidate feature subset, and recording the number of deleted traffic features;

[0175] When all features in the candidate feature subset have been analyzed and the traffic feature set is not empty, traffic features of the number of deleted traffic features are randomly selected from the traffic feature set and moved to the candidate feature subset, and the number of deleted traffic features is reset to zero, and the process of primary and secondary screening of the candidate feature subset is repeated.

[0176] Furthermore, when all features in the candidate feature subset have been analyzed and the number of deleted traffic features is zero and the traffic feature set is not empty, at least one traffic feature is selected from the traffic feature set and moved to the candidate feature subset, and the selected traffic feature is deleted from the traffic feature set, and the process of primary and secondary screening of the candidate feature subset is repeated.

[0177] In one embodiment of the present invention, another method for processing geographic information data is provided, such as Figure 5 As shown, including:

[0178] Step 501, obtaining geographic information data of a target area within a predetermined time range.

[0179] Step 502: Calculate multiple data sets of traffic feature sets and target feature values ​​of each data set based on geographic information data.

[0180] The real-time information of step 501 to step 502 may refer to step 101 to step 102, which will not be described in detail here.

[0181] Step 503: randomly select a predetermined number of traffic features from the traffic feature set to form a candidate feature subset, and delete the traffic features in the candidate feature subset from the traffic feature set.

[0182] The predetermined number may be pre-configured, for example, 4, and may be determined based on the total amount of concentrated data on traffic characteristics.

[0183] Step 504 , performing primary screening and secondary screening on the features in the candidate feature subset, and recording the number of deleted traffic features.

[0184] The real-time process of the primary screening and the secondary screening can be referred to in the previous embodiments and will not be described in detail here.

[0185] Step 505 , determining whether all features in the candidate feature subset have been analyzed and whether the traffic feature set is empty.

[0186] Step 506, when all features in the candidate feature subset have been analyzed and the traffic feature set is not empty, and the number of deleted traffic features is not zero, randomly select traffic features of the number of deleted traffic features from the traffic feature set to add to the candidate feature subset, and reset the number of deleted traffic features to zero;

[0187] When the number of traffic features to be deleted is zero, at least one traffic feature is selected from the traffic feature set to be included in the candidate feature subset, and the selected traffic feature is deleted from the traffic feature set, and the process returns to step 504 to continue primary and secondary screening of the candidate feature subset.

[0188] When all features in the candidate feature subset have been analyzed and the traffic feature set is empty, the screening ends.

[0189] This embodiment can improve the efficiency of traffic feature analysis by introducing the candidate feature subset. In specific implementation, in order to further improve the processing efficiency, the candidate feature subset can be sent to a distributed server for parallel processing.

[0190] In one embodiment of the present invention, Figure 6 As shown, the above step 503 randomly selects a predetermined number of traffic features from the traffic feature set to form a candidate feature subset, including:

[0191] Step 601, obtaining performance data of a distributed server, wherein the performance data of the distributed server includes: CPU performance, memory performance, etc.

[0192] Step 602: Determine the characteristic processing capacity of each server based on the performance data of each server.

[0193] When this step is implemented, the feature processing capacity of the server can be determined based on the CPU performance and memory usage required for feature processing. Here, the feature processing capacity refers to the workload that the server can handle.

[0194] Step 603 , according to the feature processing amount of each server, features are acquired from the traffic feature set to obtain a candidate feature subset of each server.

[0195] When this step is implemented, after obtaining features from the traffic feature set, the original features in the traffic feature set will be deleted to avoid repeated acquisition of the same features in the subsequent process.

[0196] Step 604, sending the candidate feature subsets of each server to each server for analysis, so that each server can screen the candidate feature subsets and count the number of traffic features deleted from the candidate feature subsets; according to the number of traffic features deleted by each server and the performance of each server, traffic features are obtained from the traffic feature set.

[0197] This embodiment determines the feature processing amount according to the performance of the server, which can ensure the maximum working efficiency of the server and improve the feature processing efficiency. In addition, this embodiment can achieve parallel processing of feature processing and improve feature processing efficiency by distributing candidate feature subsets.

[0198] After the present invention screens out the above-mentioned optimal traffic feature set that affects the target variable, a prediction model of the target variable can be established based on the data set of the optimal traffic feature set. On the basis of establishing the prediction model, the prediction model can be used to predict the target variable value according to the real-time optimal traffic feature value.

[0199] Based on the same inventive concept, the present invention also provides a device for processing geographic information data, as described in the following embodiments. Since the principle of solving the problem by the device for processing geographic information data is similar to the method for processing geographic information data, the implementation of the device for processing geographic information data can refer to the method for processing geographic information data, and the repeated parts will not be repeated.

[0200] Specifically, Figure 7 As shown, the device for processing geographic information data includes:

[0201] The data acquisition unit 701 is used to acquire geographic information data of a target area within a predetermined time range.

[0202] The calculation unit 702 is used to calculate multiple data sets of traffic feature sets and target feature values ​​of each data set according to the geographic information data.

[0203] The first screening unit 703 is used to screen the features in the traffic feature set by using a predetermined physical knowledge base to obtain a backup traffic feature set, wherein the physical knowledge base includes a plurality of feature sets having a physical association relationship.

[0204] The second screening unit 704 is used to perform secondary screening on the features in the backup traffic feature set according to the strong correlation between the features in the backup traffic feature set and the strong correlation between the features in the backup traffic feature set and the target features, so as to obtain an optimal traffic feature set, and establish a target feature prediction model according to the optimal traffic feature set and the target features.

[0205] This embodiment uses a physical knowledge base to perform a primary screening of features in a traffic feature set to obtain a backup traffic feature set. Based on the strong correlation between features in the backup traffic feature set and the strong correlation between features in the backup traffic feature set and target features, a secondary screening is performed on features in the backup traffic feature set to obtain an optimal traffic feature set. This can reduce the complexity of the traffic feature selection algorithm for geographic information data, increase its applicable scenarios, ensure calculation accuracy under limited resources, and reduce feature redundancy.

[0206] In a further embodiment of the present invention, Figure 8 The geographic information data processing device further comprises:

[0207] The splitting unit 705 is used to randomly select a predetermined number of traffic features from the traffic feature set to form a candidate feature subset, and delete the traffic features in the candidate feature subset from the traffic feature set.

[0208] Furthermore, the first screening unit 703 and the second screening unit 704 are used to perform a primary screening and a secondary screening on the features in the candidate feature subset, and record the number of deleted traffic features.

[0209] The update control unit 706 is used to randomly select the number of traffic features to be deleted from the traffic feature set to the candidate feature subset when all features in the candidate feature subset have been analyzed and the number of deleted traffic features is not zero, and to reset the number of deleted traffic features to zero; when the number of deleted traffic features is zero, at least one traffic feature is selected from the traffic feature set to the candidate feature subset, and the selected traffic feature is deleted from the traffic feature set; the process of primary and secondary screening of the candidate feature subset is repeated using the first screening unit 703 and the second screening unit 704; when the traffic feature set is empty, the screening is terminated.

[0210] In one embodiment of the present invention, the target feature is the travel time to the destination, and the optimal traffic feature of the target feature can be determined according to the method described in the above embodiment. On this basis, a travel time prediction model is established using the data set of the optimal traffic feature and the corresponding target feature data set. When modeling, a neural network can be selected as the initial model. The present invention does not limit the specific implementation process of the modeling.

[0211] Based on the established travel time prediction model, navigation application can be performed. In one embodiment of the present invention, a navigation system is provided, such as Fig. 9 As shown, it includes: unmanned aerial device 901, satellite server 902 and client 903.

[0212] The travel time prediction model is installed on the unmanned aerial device 901 and the satellite server 902. The unmanned aerial device 901 may be a drone, an airship, a balloon, or the like.

[0213] The client 903 is used for the user to input the destination and detect satellite signals. When a satellite signal is detected, the destination information is sent to the satellite server 902, so that the travel time to the destination is calculated by the travel time prediction model in the satellite server 902 and returned to the client 903. When no satellite signal is detected, the destination information is sent to a nearby unmanned aerial device, so that the unmanned aerial device 901 uses the travel time prediction model to calculate the travel time to the destination and return it to the client 903.

[0214] In detail, the travel time prediction model is used to calculate the travel time from the current location to the destination, including:

[0215] First determine all roads between the current location and the destination and the traffic characteristics of the roads;

[0216] Input the traffic characteristic information of each road into the travel time prediction model to obtain the travel time of each road;

[0217] Calculate the sum of the travel time of the roads involved in the path from the current location to the destination, get the travel time of the path from the current location to the destination, and return the minimum travel time and path to the client.

[0218] This embodiment can ensure that navigation can still be achieved through the unmanned aerial device when there is no satellite signal, thereby improving user experience.

[0219] In order to more clearly illustrate the technical solution of the present invention, the following is an explanation using emergency traffic dispatch as the background scenario. The objects of the emergency traffic dispatch refer to emergency vehicles such as fire trucks, ambulances, road rescue vehicles, and police vehicles. The dispatch scenario has the characteristics of high timeliness, high stability, and dedicated communication, and needs to have the ability to quickly perceive traffic conditions and ensure information security. The algorithm of the present invention is used for traffic status perception. An application example is a traffic accident scenario. The location of the traffic accident and the location of the rescue vehicle are known. The prediction model provided by the present invention is used to determine the traffic conditions of each road between the location of the traffic accident and the rescue vehicle. A path is planned for the rescue vehicle based on the traffic conditions of the roads, and a rescue vehicle with the smallest planned path is selected from all rescue vehicles. The rescue vehicle with the smallest planned path is notified to go to the location of the traffic accident for rescue.

[0220] The implementation platform of this embodiment is an AMAX server with Ubuntu system installed. The server includes 80 2GHZ Intel Xeon Gold 6138 CPUs, 5 NVIDIA Quadro GV100 GPUs, 256G of operating memory and 4.2TB of storage hard disk. For the convenience of operation, the server is connected to an external monitor, mouse and keyboard.

[0221] This embodiment uses the prediction of urban street travel time as an example scenario, and its target feature is the travel time of road section i at time t+1. , the running speed is calculated according to the section travel time and section length, and the road state is determined based on the running speed. Its traffic feature set also includes: the traffic density corresponding to section i at time t , Traffic flow rate , average traffic speed , time occupancy , Space Occupancy , the road width corresponding to section i η i , the length of the road corresponding to section i l i , the road area corresponding to section i α i , road congestion density ρ m , road critical density ρ c , traffic capacity q m , free flow speed v f , the traffic signal type corresponding to road section i s i , the area type corresponding to road segment i β i , the travel time corresponding to section i at time t .

[0222] The public data set collection scope of this embodiment is the local road network of a certain area, which includes 35 road sections and 22 intersections, covering 3.04 square kilometers. The collection period is from May 21, 2019 to May 26, 2019 during the daytime, and a total of more than 10.8 million data items are generated.

[0223] The processing of the above geographic information data includes:

[0224] Step 1: Input and process the geographic information data of the above local network, wherein the geographic information data processing includes coordinate system conversion, smoothing, spatiotemporal redundancy cleaning and data completion.

[0225] Step 2: Extracting traffic features, the extraction method includes the following steps:

[0226] Step 2.1: Calculation of static features: The calculation method is to use the spatial data in the geographic information data, and calculate or count the static variables based on the coordinates of the boundary intersection points according to the spatial boundaries and regional shapes.

[0227] Step 2.2: Dynamic feature calculation: The calculation method is divided into instantaneous sampling and spatiotemporal statistics. The instantaneous sampling is to obtain the variable value according to the spatial range at a certain moment, and the spatiotemporal statistics is to calculate the spatiotemporal statistics within a certain time interval.

[0228] Step 3: Form the original traffic feature set, which is in the form of:

[0229] .

[0230] Step 4: Feature selection, the feature selection includes the following steps:

[0231] Step 4.1: Initialize parameters. In this embodiment, the initialization values ​​of the parameters are:

[0232] Establish the feature set to be selected Ω and convert the original traffic feature set Ω I Assign to the feature set to be selected, that is, Ω=Ω I ;

[0233] By randomly selecting three traffic features from the original traffic feature set Ω, a candidate feature subset Ω is established. C ,During specific implementation, other numbers of traffic features can be selected to form the ,candidate feature subset;

[0234] Establish and select feature subset Ω S And the best feature subset Ω O , and assign the following values: , ;

[0235] The lower bound of the correlation coefficient corresponding to strong coherence in the t-test method θ t =2.581, the upper bound of the correlation coefficient corresponding to incoherence in the t-test method τ t =0.005;

[0236] The lower bound of the correlation coefficient corresponding to strong coherence in Pearson's correlation coefficient analysis method θ p =0.850, the upper bound of the correlation coefficient of the irrelevant correspondence in the Pearson correlation coefficient analysis method τ p =0.300.

[0237] During specific implementation, other t-test methods and upper and lower bound parameters of Pearson correlation coefficient analysis may be selected according to actual needs.

[0238] The form of the traffic set in the physical knowledge base is shown in Table 5 below:

[0239] Table 5

[0240]

[0241] Step 4.2: Determine whether the feature set Ω to be selected is an empty set. If it is an empty set, go to step 4.9; otherwise, go to step 4.3.

[0242] Step 4.3: Determine the number of traffic features to be deleted N F Is it 0? If so, randomly select a traffic feature from the feature set Ω to be selected and add it to the candidate feature subset Ω C Otherwise, add and delete the number of traffic features from the feature set Ω to be selected N F Traffic characteristics.

[0243] Step 4.4: Delete the number of traffic features N F Return to zero.

[0244] Step 4.5: Substitute the candidate feature subset Ω C The traffic features in are deleted from the feature set Ω to be selected.

[0245] Step 4.6: Based on the traffic related sets in the physical knowledge base Update the candidate feature subset Ω C and the number of traffic features deleted N F The update method is to traverse the traffic related set , select the feature that can be completely replaced by other features, and remove the feature from the candidate feature subset Ω C Remove them, and calculate the number of removed features as the number of deleted traffic features N F The value of .

[0246] Step 4.7: Select the candidate feature subset Ω C Randomly select a traffic feature from F i .

[0247] Step 4.8: Calculate traffic characteristics F i The coherence with other features, the calculation method of the coherence comprises the following steps:

[0248] Step 4.8.1: Determine traffic characteristics F i Is it a binary feature? If it is a binary feature, execute step 4.8.2; otherwise, execute step 4.8.9.

[0249] In this embodiment, the binary feature is s i , β i , where s i =0, indicating that the signal control method is no signal control, s i =1, it means the signal control method is signal control, and β i =0, it means the road space is an urban section. β i =1, it indicates that the road space is an intersection.

[0250] Step 4.8.2: Calculate traffic characteristics F i t-test parameters with target feature Y t ( F i , Y ) and P i ( F i , Y ), the test parameters t ( F i , Y ) is calculated as follows:

[0251] (1)

[0252] in, Traffic characteristics F i The average of the difference between paired samples of Y under the action of binary features, μ 0 Traffic characteristics F i The expectation of the corresponding t distribution is n F Traffic characteristics F i The number of paired sample groups included, s d Traffic characteristics F i The standard deviation of the difference between paired samples of Y under the action of two values ​​is calculated as follows:

[0253] (2)

[0254] in, Traffic characteristics F i The difference between the jth pair of samples in Y under the action of binary features.

[0255] The P i ( F i , Y ) indicates traffic characteristics F i The probability of occurrence of binary features not being significant, the algorithm is based on the test parameters t ( F i , Y ) value calls the value table to obtain the corresponding real-time value; then in the binary feature s i , β i The test parameters for the target feature Y are shown in Table 6 below:

[0256] Table 6

[0257]

[0258] Step 4.8.3: Determine t-test parameters t ( F i , Y )and , P i ( F i , Y )and relationship, if or , then the traffic characteristics F i Not related to the target feature Y, the traffic feature F i From the candidate feature subset Ω C Remove from the list and delete the number of traffic features N F The value of is increased by 1, and then returns to step 4.7, otherwise, it indicates the traffic characteristics F i Related to the target feature Y, perform step 4.8.4.

[0259] In this embodiment, s i , β i They all have significant effects on the target features and are correlated.

[0260] Step 4.8.4: Determine the set {Ω C - F i} is an empty set, if so, return to step 4.2, otherwise, select C - F i Select traffic features in F k , perform step 4.8.5.

[0261] Step 4.8.5: Determine traffic characteristics F k Is it a binary feature? If so, calculate the t-test parameter t( F i , F k ) and P t ( F i , F k ), execute 4.8.6; otherwise, execute 4.8.7;

[0262] In this embodiment, s i , β i t( F i , F k ) and P t ( F i , F k The calculation results of ) are shown in Table 7 below:

[0263] Table 7

[0264]

[0265] Step 4.8.6: Determine the t-test parameter t( F i , F k )and θ t , P t ( F i , F k )and τ t If t( F i , F k )> θ t And P t ( F i , F k)< τ t , then the traffic characteristics F i and F k Related, the traffic characteristics F i From the candidate feature subset Ω C Remove from the list and delete the number of traffic features N F The value of is increased by 1, and then, return to step 4.7, otherwise, it indicates the traffic characteristics F i and F k Not relevant, return to step 4.8.4.

[0266] Step 4.8.7: Calculate the Pearson correlation coefficient P e ( F i , F k ), and execute step 4.8.8, the P e ( F i , F k ) is calculated as follows:

[0267] (3)

[0268] Among them, Cov( F i , F k ) are two random variables (i.e., traffic characteristics F i , F k ), D( F i ) is the traffic characteristics F i The variance of F k ) is the traffic characteristics F k The variance of .

[0269] In this embodiment, s i , β i The Pearson correlation coefficient calculation results with other non-binary features are as follows Fig.10 shown.

[0270] Step 4.8.8: Judgement If yes, then Fi , F k If relevant, the traffic characteristics F i From the candidate feature subset Ω C Remove from the list and delete the number of traffic features N F The value is increased by 1, otherwise, the following judgment is performed:

[0271] judge If yes, then F i , F k Not relevant, return to step 4.8.4.

[0272] Step 4.8.9 Calculate P e ( F i , Y ), execute step 4.8.10;

[0273] In this embodiment, the Pearson correlation coefficient between the non-binary feature and the target feature is as follows: Fig.11 shown.

[0274] Step 4.8.10: Judgement If yes, it means it is irrelevant and the traffic characteristics F i From the candidate feature subset Ω C Remove from the list and delete the number of traffic features N F The value of is increased by 1, then the process returns to step 4.7, otherwise, the process proceeds to step 4.8.11.

[0275] Step 4.8.11: Judgement , if yes, return to step 4.7, otherwise, return to step 4.8.4.

[0276] Step 4.9: Final candidate feature subset , let Ω 0 =Ω s =Ω C .

[0277] Step 5: Output candidate feature subset Ω C , as a result of the feature selection method and ends the algorithm.

[0278] Based on the optimal traffic feature set, the dynamic graph convolutional recurrent network (DGCRN) can be selected as the experimental neural network. The prediction models are established according to the traffic features before and after screening, and the prediction accuracy of the two prediction models is compared to verify the feasibility of the feature algorithm. The neural network architecture of DGCRN is as follows: Fig.12As shown, the architecture of the dynamic graph generator in the neural network of DGCRN is as follows Fig.13 As shown, the dynamic graph convolution module in the neural network of DGCRN is as follows Fig.14 shown.

[0279] The output of DGCRN in this embodiment is , the accuracy evaluation index is the error distribution, and the efficiency evaluation index is the average calculation time; the error distribution diagram is as follows Fig.15 shown.

[0280] Before and after feature selection, the prediction error of DGCRN has no obvious change, and the high error interval distribution is greater than 12s. The input sample size is lower after the feature selection algorithm is used. In terms of average calculation time, the corresponding calculation time of DGCRN without using the processing method of geographic information data of the present invention is 1.59s, and the corresponding calculation time of DGCRN using the processing method of geographic information data of the present invention is 1.16s. The implementation of the present invention reduces the calculation time of DGCRN. The above results show that the feature screening of the processing method of geographic information data of the present invention can maintain and improve the prediction accuracy of DGCRN and improve the calculation accuracy, which has practical value.

[0281] The traffic state prediction results of this embodiment are divided into smooth traffic, slow traffic and congestion. Please refer to Table 1 for the specific classification method.

[0282] Based on the predicted travel time of each road section, the traffic status prediction results can be calculated, such as Fig.16A and Fig. 16B As shown, Fig.16A This is a schematic diagram of the local road network. Fig. 16B It is the traffic status prediction map of the local road network.

[0283] On the basis of establishing the above-mentioned prediction model, the present invention can also be applied to on-board edge computing in specific urban scenarios such as emergency rescue, large-scale event evacuation, holiday travel distribution, etc., providing a new idea for low-cost and highly reliable solutions.

[0284] The present invention is used in the field of intelligent transportation systems and can screen traffic features obtained from geographic information data online or offline according to actual application scenarios.

[0285] Online traffic feature selection can be used for short-term traffic prediction and real-time traffic estimation needs. It can quickly train traffic models, agilely update traffic parameters, and intelligently mine traffic data. It can quickly obtain core traffic information from real-time large-scale data, ensure fast and accurate traffic modeling applications, and reduce the interference of redundant data on the accuracy of model results.

[0286] Offline traffic feature selection can be used for compressed storage of traffic big data, achieving miniaturization and high fidelity of traffic big data storage while minimizing information loss.

[0287] In addition, the present invention takes into account the inherent connection between traffic characteristic variables, conforms to the basic laws such as the macro traffic basic map and the optimal traffic system, takes into account both data-driven and model-driven architectures, and introduces traditional traffic flow models into traffic big data application scenarios.

[0288] The present invention is a prerequisite and basic technology for artificial intelligence, traffic modeling research and traffic estimation and prediction algorithms based on geographic information or traffic data, and can improve the reliability and generalization ability of geographic information data in related traffic fields.

[0289] In one embodiment of the present invention, a computer device is provided, such as Fig.17 As shown, the computer device 1702 may include one or more processors 1704, such as one or more central processing units (CPUs), each of which may implement one or more hardware threads. The computer device 1702 may also include any memory 1706, which is used to store any kind of information such as code, settings, data, etc. Non-limitingly, for example, the memory 1706 may include any one or more combinations of the following: any type of RAM, any type of ROM, flash memory device, hard disk, optical disk, etc. More generally, any memory may use any technology to store information. Further, any memory may provide volatile or non-volatile retention of information. Further, any memory may represent a fixed or removable component of the computer device 1702. In one case, when the processor 1704 executes an associated instruction stored in any memory or a combination of memories, the computer device 1702 may perform any operation of the associated instruction. The computer device 1702 also includes one or more drive mechanisms 1708 for interacting with any memory, such as a hard disk drive mechanism, an optical disk drive mechanism, etc.

[0290] The computer device 1702 may also include an input / output module 1710 (I / O) for receiving various inputs (via input device 1712) and for providing various outputs (via output device 1714). A specific output mechanism may include a presentation device 1716 and an associated graphical user interface 1718 (GUI). In other embodiments, the input / output module 1710 (I / O), input device 1712, and output device 1714 may not be included, and the computer device 1702 may be used as a computer device in a network. The computer device 1702 may also include one or more network interfaces 1720 for exchanging data with other devices via one or more communication links 1722. One or more communication buses 1724 couple the components described above together.

[0291] The communication link 1722 may be implemented in any manner, for example, through a local area network, a wide area network (e.g., the Internet), a point-to-point connection, etc., or any combination thereof. The communication link 1722 may include any combination of hardwired links, wireless links, routers, gateway functions, name servers, etc. governed by any protocol or combination of protocols.

[0292] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above method are executed.

[0293] An embodiment of the present invention further provides a computer-readable instruction, wherein when a processor executes the instruction, the program therein causes the processor to execute the method described in any of the aforementioned embodiments.

[0294] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0295] It should also be understood that in the embodiments of the present invention, the term "and / or" is only a description of the association relationship of the associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in the present invention generally indicates that the associated objects before and after are in an "or" relationship.

[0296] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented by electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0297] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0298] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, or it can be an electrical, mechanical or other form of connection.

[0299] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present invention.

[0300] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0301] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0302] The present invention uses specific embodiments to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A method for processing geographic information data, characterized in that: include: Obtain geographic information data of the target area within a predetermined time range; According to the geographic information data, multiple data sets of traffic feature sets and target feature values ​​of each data set are calculated, wherein the target feature is the travel time or traffic status of the road section; Randomly selecting a predetermined number of traffic features from the traffic feature set to form a candidate feature subset, and deleting the traffic features in the candidate feature subset from the traffic feature set; Using a predetermined physical knowledge base, the features in the candidate feature subset are screened to obtain a spare traffic feature set, and the number of deleted traffic features is recorded; the physical knowledge base includes a plurality of feature sets having a physical association relationship; Performing secondary screening on the features in the backup traffic feature set according to the strong correlation between the features in the backup traffic feature set and the strong correlation between the features in the backup traffic feature set and the target features; When all the features in the candidate feature subset have been analyzed and the traffic feature set is not empty, traffic features of the number of deleted traffic features are randomly selected from the traffic feature set and moved to the candidate feature subset, and the number of deleted traffic features is reset to zero, and the steps of screening the candidate feature subset and subsequent steps are repeated. When the traffic feature set is empty, an optimal traffic feature set is obtained, and a target feature prediction model is established based on the optimal traffic feature set and the target features.

2. The method according to claim 1, characterized in that According to the strong correlation between the features in the backup traffic feature set and the strong correlation between the features in the backup traffic feature set and the target features, performing secondary screening on the features in the backup traffic feature set includes: For each first traffic feature in the spare traffic feature set, determining the correlation between the first traffic feature and the target feature according to the data set of the first traffic feature and the target feature value of the data set; when the first traffic feature is not correlated with the target feature, deleting the first traffic feature from the spare traffic feature set; For every two traffic features in the spare traffic feature set, the correlation between the two traffic features is determined according to the data sets of the two traffic features; when the two traffic features are correlated, one of the traffic features is deleted from the two traffic features.

3. The method according to claim 2, characterized in that When the first traffic feature is related to the target feature, traversing a second traffic feature from the backup traffic feature set; The first traffic feature and the second traffic feature are used as the two traffic features, the correlation between the two traffic features is calculated, and when the two traffic features are correlated, the first traffic feature is deleted from the spare traffic feature set.

4. The method according to claim 3, characterized in that When the two traffic features are unrelated, repeat traversing the next second traffic feature from the backup traffic feature set and subsequent steps until the second traffic feature cannot be traversed, then traverse the next first traffic feature from the backup traffic feature set and repeatedly determine the correlation between the first traffic feature and the target feature.

5. The method according to claim 1, characterized in that The traffic features in the traffic feature set include at least: traffic operation features corresponding to the road section i at time t, traffic design features corresponding to the road section i at time t, and traffic facility features corresponding to the road section i at time t.

6. The method according to claim 1, characterized in that The physical knowledge base establishment process includes: Acquiring physical knowledge information of the traffic characteristics and physical knowledge information of the target characteristics; De-duplicate the acquired physical knowledge information; Filtering target physical knowledge information from the deduplicated physical knowledge information, wherein the target physical knowledge information includes at least two traffic characteristics, or the target physical knowledge information includes at least one traffic characteristic and a target characteristic; Extracting the correlation relationship between physical quantities from the target physical knowledge information; Traffic features with physical quantity correlation, and traffic features and target features with physical correlation constitute a physical knowledge set.

7. The method according to claim 1, characterized in that Using a predetermined physical knowledge base, the features in the traffic feature set are screened, including: Matching the traffic features in the traffic feature set with the features in the feature set of the physical knowledge base one by one, and recording the number of traffic feature matches; For each feature set in the physical knowledge base, if the feature set has at least two matching traffic features, the traffic feature with the smallest number of matches is retained in the traffic feature set.

8. The method according to claim 1, characterized in that When all features in the candidate feature subset have been analyzed and the number of deleted traffic features is zero and the traffic feature set is not empty, at least one traffic feature is selected from the traffic feature set and moved to the candidate feature subset, and the selected traffic feature is deleted from the traffic feature set, and the process of primary and secondary screening of the candidate feature subset is repeated.

9. The method according to claim 1, characterized in that A predetermined number of traffic features are randomly selected from the traffic feature set to form a candidate feature subset, including: Get performance data of distributed servers; Determine the characteristic processing capacity of each server based on the performance data of each server; According to the feature processing capacity of each server, features are obtained from the traffic feature set to obtain a candidate feature subset of each server; The candidate feature subsets of each server are sent to each server for analysis, so that each server can screen the candidate feature subsets, and count the number of traffic features deleted from the candidate feature subsets, and obtain traffic features from the traffic feature set based on the number of traffic features deleted by each server and the performance of each server.

10. The method according to claim 1, characterized in that After obtaining geographic information data, it also includes: Preprocessing geographic information data, including: Smoothing the non-binary data in geographic information data; Carry out coordinate system transformation on spatial data in geographic information data; Clean the time-varying data in geographic information data for spatiotemporal redundancy; Complete the default data.

11. A device for processing geographic information data, characterized in that: include: A data acquisition unit, used to acquire geographic information data of a target area within a predetermined time range; A calculation unit, used for calculating a plurality of data sets of traffic characteristic sets and a target characteristic value of each data set according to the geographic information data; a splitting unit, configured to randomly select a predetermined number of traffic features from the traffic feature set to form a candidate feature subset, and delete the traffic features in the candidate feature subset from the traffic feature set; A first screening unit is used to screen the features in the candidate feature subset using a predetermined physical knowledge base to obtain a spare traffic feature set, and record the number of deleted traffic features, wherein the physical knowledge base includes a plurality of feature sets having a physical association relationship; The second screening unit is used to perform secondary screening on the features in the backup traffic feature set according to the strong correlation between the features in the backup traffic feature set and the strong correlation between the features in the backup traffic feature set and the target features. When all the features in the candidate feature subset have been analyzed and the traffic feature set is not empty, traffic features of the number of deleted traffic features are randomly selected from the traffic feature set and moved to the candidate feature subset, and the number of deleted traffic features is reset to zero. The first screening unit is repeatedly used to screen the candidate feature subset once, and the second screening unit is used for screening. When the traffic feature set is empty, an optimal traffic feature set is obtained, and a target feature prediction model is established according to the optimal traffic feature set and the target features.

12. A navigation system, characterized in that: include: Unmanned aerial devices, satellite servers and clients; The unmanned aerial device and the satellite server are installed with a travel time prediction model, the target feature of the travel time prediction model is the travel time of the road section, and the traffic feature set of the travel time prediction model is determined by the method according to any one of claims 1 to 10; The client is used for allowing the user to input a destination and detect satellite signals. When a satellite signal is detected, the destination information is sent to a satellite server, so that the satellite server uses the travel time prediction model to calculate the travel time to the destination and returns it to the client. When no satellite signal is detected, the destination information is sent to a nearby unmanned aerial device, so that the unmanned aerial device uses the travel time prediction model to calculate the travel time to the destination and returns it to the client.

13. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 10 is implemented.

14. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor of a computer device, the method according to any one of claims 1 to 10 is implemented.

15. A computer program product, comprising a computer program, characterized in that: When the computer program is executed by a processor of a computer device, the method according to any one of claims 1 to 10 is implemented.

Images