Method and device for testing highway traffic meteorological disaster risk forecasting model

Abstract

The invention provides an inspection method and device for a highway traffic meteorological disaster risk forecasting model, and relates to the technical field of traffic meteorological disaster risk forecasting. The method comprises the following steps: acquiring traffic meteorological disaster event information of a selected highway section within a preset time range; based on the space-time deviation data, correcting the first time range and the first road range respectively to obtain a second time range and a second road range of each traffic meteorological disaster event; based on the second time range and the second road section range of each traffic meteorological disaster event, determining a test sample of each traffic meteorological disaster event from risk forecast data output by the to-be-tested highway traffic meteorological disaster risk forecast model; and based on the respective test samples of the plurality of traffic meteorological disaster events, evaluating the to-be-tested highway traffic meteorological disaster risk forecasting model. According to the invention, the obtained test sample can better conform to the actual situation, and the reliability of the model test result is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of traffic meteorological disaster risk forecasting, and in particular to a testing method and device for a highway traffic meteorological disaster risk forecasting model. Background Art

[0002] At present, the verification scheme of the transportation meteorological disaster risk forecast model is mainly achieved by statistically analyzing the distribution of disaster event samples within different risk levels, or calculating the relevant evaluation index values. However, its verification process requires that the disaster risk forecast results be fully matched with the time and location of the actual disaster event. It does not consider the impact of some non-meteorological factors (such as human subjective factors, etc.) on the actual occurrence of disaster events, thereby affecting the reliability of the test results. Summary of the invention

[0003] The embodiment of the present invention provides a method and device for testing a highway traffic meteorological disaster risk forecasting model, which are used to solve the technical problem of low reliability of the test results in the prior art.

[0004] In a first aspect, an embodiment of the present invention provides a method for testing a highway traffic meteorological disaster risk forecasting model, comprising: obtaining traffic meteorological disaster event information of a selected highway section within a preset time range; wherein the traffic meteorological disaster event information includes a first time range, a first section range and traffic meteorological disaster information of each of a plurality of traffic meteorological disaster events; based on spatiotemporal deviation data, respectively correcting the first time range and the first section range to obtain a second time range and a second section range of each of the traffic meteorological disaster events; wherein the spatiotemporal deviation data includes a time deviation amount and a space deviation amount corresponding to each of different traffic meteorological disaster types; the time deviation amount and the space deviation amount are respectively used to characterize the time deviation and space deviation between the actual meteorological conditions that trigger the traffic meteorological disaster event and the actual traffic meteorological disaster event; based on the second time range and the second section range of each of the traffic meteorological disaster events, determining a test sample of each of the traffic meteorological disaster events from the risk forecasting data output by the highway traffic meteorological disaster risk forecasting model to be tested; and evaluating the highway traffic meteorological disaster risk forecasting model to be tested based on the test samples of each of the plurality of traffic meteorological disaster events.

[0005] In some embodiments, based on the spatiotemporal deviation data, the first time range and the first road section range are respectively corrected to obtain the second time range and the second road section range of each of the traffic and meteorological disaster events, including: for each of the traffic and meteorological disaster events, based on the traffic and meteorological disaster information of the traffic and meteorological disaster event, determining the target time deviation amount and the target space deviation amount from the spatiotemporal deviation data; based on the target time deviation amount, respectively correcting the start time and the end time of the first time range of the traffic and meteorological disaster event to obtain the second time range of the traffic and meteorological disaster event; based on the target space deviation amount, correcting the start and end time of the first road section range of the traffic and meteorological disaster event to obtain the second road section range of the traffic and meteorological disaster event.

[0006] In some embodiments, the time deviation corresponding to each traffic and meteorological disaster type includes a leading time deviation and a trailing time deviation; the time deviation corresponding to any of the traffic and meteorological disaster types is determined by: obtaining historical traffic and meteorological disaster event information corresponding to the traffic and meteorological disaster type within a historical time period; wherein the historical traffic and meteorological disaster event information includes a historical time range and a historical road section range of a plurality of historical traffic and meteorological disaster events; for a first historical traffic and meteorological disaster event among the historical traffic and meteorological disaster events, based on the first historical road section range and the first historical time range of the first historical traffic and meteorological disaster event, as well as the search time, obtaining a first historical meteorological observation data set corresponding to the first historical traffic and meteorological disaster event; based on the traffic The meteorological disaster risk forecasting model and the historical meteorological observation data at each moment in the first historical meteorological observation data set are used to conduct risk search to determine the first risk start time and the first risk end time of the first historical traffic meteorological disaster event; based on the difference between the start time of the first historical time range and the start time of the first historical time range, the first time deviation of the first historical traffic meteorological disaster event is determined; based on the difference between the end time of the first historical time range and the first risk end time, the second time deviation of the first historical traffic meteorological disaster event is determined; based on the first time deviation and the second time deviation of each of the multiple historical traffic meteorological disaster events, the lead time deviation and the post time deviation corresponding to the traffic meteorological disaster type are determined.

[0007] In some embodiments, based on the first historical road section range and the first historical time range, as well as the search time, the first historical meteorological observation data set corresponding to the first historical traffic meteorological disaster event is obtained, including: based on the first historical road section range, determining the target meteorological site corresponding to the first historical traffic meteorological disaster event; based on the first historical time range and the search time, determining the first historical meteorological observation data set from the historical meteorological observation database of the target meteorological site.

[0008] In some embodiments, the risk search is performed based on the highway traffic meteorological disaster risk forecast model to be tested and the historical meteorological observation data at each moment in the first historical meteorological observation data set to determine the first risk start time and the first risk end time of the first historical traffic meteorological disaster event, including: taking the start time of the first historical time range as the first starting point, according to the preset step size and the search time, the historical meteorological observation data of each moment before the first starting point in the first historical meteorological observation data set is sequentially input into the highway traffic meteorological disaster risk forecast model to be tested, and the first risk start time is determined based on the output risk actual result; if the first risk start time is not searched from the various moments before the first starting point, the historical meteorological observation data of each moment after the first starting point in the first historical meteorological observation data set is sequentially input according to the step size and the search time. The observation data is input into the road traffic meteorological disaster risk forecasting model to be tested, and the start time of the first risk is determined based on the output risk actual result; taking the end time of the first historical time range as the second starting point, the historical meteorological observation data of each moment after the second starting point in the first historical meteorological observation data set is input into the road traffic meteorological disaster risk forecasting model to be tested in sequence according to the step size and the search time, and the end time of the first risk is determined based on the output risk actual result; if the end time of the first risk is not found from the moments after the second starting point, the historical meteorological observation data of each moment before the second starting point in the first historical meteorological observation data set is input into the road traffic meteorological disaster risk forecasting model to be tested in sequence according to the step size and the search time, and the end time of the first risk is determined based on the output risk actual result.

[0009] In some embodiments, based on the second time range and second road section range of each traffic and meteorological disaster event, the test sample of each traffic and meteorological disaster event is determined from the risk forecast data output by the highway traffic and meteorological disaster risk forecast model to be tested, including: for each traffic and meteorological disaster event, based on a preset road section unit table, determining a target road section unit that has an intersection with the second road section range of the traffic and meteorological disaster event; wherein the road section unit table includes the road section ranges of a plurality of road section units, and the correspondence between each road section unit and the forecast unit of the highway traffic and meteorological disaster risk forecast model to be tested; the plurality of road section units are obtained by segmenting each highway section according to a preset length; based on the road section unit table, determining a target forecast unit corresponding to the target road section unit; based on the target forecast unit and the second time range of the traffic and meteorological disaster event, determining target risk forecast data from the risk forecast data; the target risk forecast data is the risk forecast data whose forecast time is within the second time range and whose forecast unit is the target forecast unit; determining the target risk forecast data as the test sample of the traffic and meteorological disaster event.

[0010] In a second aspect, an embodiment of the present invention provides a testing device for a highway traffic meteorological disaster risk forecasting model, comprising: a first acquisition module, used to obtain traffic meteorological disaster event information of a selected highway section within a preset time range; wherein the traffic meteorological disaster event information includes a first time range, a first section range and traffic meteorological disaster information of each of a plurality of traffic meteorological disaster events; a correction module, used to correct the first time range and the first section range respectively based on spatiotemporal deviation data, to obtain a second time range and a second section range for each of the traffic meteorological disaster events; wherein the spatiotemporal deviation data includes different traffic meteorological disaster types corresponding to each other The time deviation and space deviation are respectively used to characterize the time deviation and space deviation between the actual meteorological conditions that trigger traffic meteorological disaster events and the actual traffic meteorological disaster events; a first determination module is used to determine the inspection sample of each traffic meteorological disaster event from the risk forecast data output by the highway traffic meteorological disaster risk forecast model to be tested based on the second time range and the second road section range of each traffic meteorological disaster event; an evaluation module is used to evaluate the highway traffic meteorological disaster risk forecast model to be tested based on the inspection samples of each of the multiple traffic meteorological disaster events.

[0011] In a third aspect, an embodiment of the present invention provides an electronic device, comprising a processor and a memory storing a computer program, wherein when the processor executes the program, the method for verifying the highway traffic meteorological disaster risk forecasting model described in the first aspect is implemented.

[0012] In a fourth aspect, an embodiment of the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the verification method of the highway traffic meteorological disaster risk forecasting model described in the first aspect is implemented.

[0013] In a fifth aspect, an embodiment of the present invention provides a computer program product, including a computer program, which, when executed by a processor, implements the verification method of the highway traffic meteorological disaster risk forecasting model described in the first aspect.

[0014] According to a method and device for testing a highway traffic meteorological disaster risk forecast model provided by an embodiment of the present invention, traffic meteorological disaster event information of a selected highway section within a preset time range is obtained; wherein the traffic meteorological disaster event information includes a first time range, a first section range and traffic meteorological disaster information of each of a plurality of traffic meteorological disaster events; based on the spatiotemporal deviation data, the first time range and the first section range are respectively corrected to obtain a second time range and a second section range of each traffic meteorological disaster event; based on the second time range and the second section range of each traffic meteorological disaster event, a test sample of each traffic meteorological disaster event is determined from the risk forecast data output by the highway traffic meteorological disaster risk forecast model to be tested; based on the test samples of each of the plurality of traffic meteorological disaster events, the highway traffic meteorological disaster risk forecast model to be tested is evaluated. The present invention starts with traffic meteorological disaster events to implement the test of the highway traffic meteorological disaster risk forecast model, and considering the influence of non-meteorological factors on the occurrence of traffic meteorological disaster events, the time range and section range of traffic meteorological disaster events are corrected based on the spatiotemporal deviation data, so that the obtained test samples can be more in line with the actual situation, thereby improving the reliability of the model test results. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are 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.

[0016] Figure 1 One of the flow charts of the verification method of the highway traffic meteorological disaster risk forecast model provided by the embodiment of the present invention; Figure 2 A second flow chart of a method for testing a highway traffic meteorological disaster risk forecasting model provided by an embodiment of the present invention; Figure 3A third flow chart of a method for testing a highway traffic meteorological disaster risk forecasting model provided by an embodiment of the present invention; Figure 4 A fourth flow chart of a method for testing a highway traffic meteorological disaster risk forecasting model provided by an embodiment of the present invention; Figure 5 A schematic diagram of the structure of a testing device for a highway traffic meteorological disaster risk forecasting model provided by an embodiment of the present invention; Figure 6 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are 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.

[0018] Explanation of terms involved in the embodiments of the present invention: 1. Traffic meteorological disaster events: traffic events that affect the safety or operating efficiency of highway traffic due to some adverse meteorological conditions, such as traffic control caused by fog, traffic blockage caused by rainfall, traffic accidents caused by snowfall, etc.

[0019] 2. Highway traffic meteorological disaster risk: the possibility and magnitude of traffic meteorological disaster events such as traffic control, traffic blockage, traffic accidents, etc. due to the adverse impact of certain unfavorable meteorological conditions on highway traffic safety and operating efficiency.

[0020] In order to solve the above technical problems, the present invention proposes a method and device for testing a highway traffic meteorological disaster risk forecasting model.

[0021] Figure 1 The present invention provides a flow chart of a method for testing a highway traffic meteorological disaster risk forecasting model. Figure 1 As shown, the method may include the following steps.

[0022] Step 101, obtaining traffic and meteorological disaster event information of a selected highway section within a preset time range; wherein the traffic and meteorological disaster event information includes a first time range, a first road section range and traffic and meteorological disaster information of each of a plurality of traffic and meteorological disaster events.

[0023] In some embodiments, traffic and meteorological disaster events refer to actual traffic control, traffic blockage, traffic accidents and other events caused by some bad weather, such as traffic accidents caused by snowfall, traffic control events caused by fog, etc. Traffic and meteorological disaster event information can be obtained based on the traffic management system.

[0024] Among them, the selected highway section can be pre-set based on actual needs, which refers to the section where the highway traffic meteorological disaster risk forecast model needs to be verified. The selected highway section can include one highway section or multiple highway sections.

[0025] In some embodiments, the first time range of each traffic meteorological disaster event refers to the time period between the start time and the end time of the traffic meteorological disaster event, and the first section range of each traffic meteorological disaster event refers to the section range of the highway section where the traffic meteorological disaster event is located, which can be the section between the start pile number and the end pile number of the highway line where the traffic meteorological disaster event occurs. For example, the first section range of a certain traffic meteorological disaster event is the section range from the start pile number K100+316 to the end pile number K110+139 of the road code G1. The traffic meteorological disaster information of each traffic meteorological disaster event may include the type of traffic meteorological disaster and the disaster level, etc. The type of traffic meteorological disaster refers to the highway traffic meteorological disaster category corresponding to the traffic meteorological disaster event, such as traffic control caused by fog is a traffic meteorological disaster type, traffic accidents caused by snowfall are a traffic meteorological disaster type, and traffic blockage caused by rainfall is also a traffic meteorological disaster type.

[0026] It should be noted that similar traffic incidents caused by different meteorological conditions also belong to different types of traffic meteorological disasters. For example, traffic blockages caused by fog and traffic blockages caused by snowfall belong to different types of traffic meteorological disasters.

[0027] Step 102, based on the time-space deviation data, respectively correct the first time range and the first road section range to obtain the second time range and the second road section range for each traffic meteorological disaster event; wherein the time-space deviation data includes the time deviation amount and the space deviation amount corresponding to different types of traffic meteorological disasters; the time deviation amount and the space deviation amount are respectively used to characterize the time deviation and space deviation between the actual meteorological conditions that trigger the traffic meteorological disaster event and the actual traffic meteorological disaster event.

[0028] It is understandable that traffic blockage, traffic control and other traffic meteorological disaster events under bad weather will be affected by non-meteorological factors. For example, when dense fog comes, although there are already meteorological conditions for traffic control caused by fog, the traffic control department will not only consider the current and subsequent development trend of fog, but also comprehensively consider traffic flow, entrance and exit locations, personnel and equipment preparation, etc. to decide whether to carry out traffic control and the time, scope, and level of traffic control. Therefore, the actual time and location of the meteorological conditions that cause traffic meteorological disaster events may not be completely consistent with the actual time and location of traffic meteorological disaster events. The inconsistency here can be attributed to the influence of non-meteorological factors such as human subjectivity. Therefore, the difference between the risk forecast data output by the highway traffic meteorological disaster risk forecast model and the traffic meteorological disaster event information will be affected by some non-meteorological factors such as human subjectivity. Therefore, when testing the highway traffic meteorological disaster risk forecast model, it is necessary to exclude the influence of non-meteorological factors on the results to achieve effective testing of the highway traffic meteorological disaster risk forecast model.

[0029] In some embodiments, the spatiotemporal deviation data includes time deviation and space deviation corresponding to different traffic and meteorological disaster types. The time deviation corresponding to a certain traffic and meteorological disaster type is used to characterize the deviation between the time range in which the meteorological conditions that trigger the traffic and meteorological disaster event of this type actually exist and the time range in which the traffic and meteorological disaster event actually occurs. The space deviation corresponding to a certain traffic and meteorological disaster type is used to characterize the deviation between the road section range in which the meteorological conditions that trigger the traffic and meteorological disaster event of this type actually exist and the road section range in which the traffic and meteorological disaster event actually occurs.

[0030] In other words, the time deviation and space deviation corresponding to each type of traffic and meteorological disaster can be regarded as the influence of non-meteorological factors such as human subjectivity on the time and space of traffic and meteorological disaster events. Based on the time and space deviation data, the first time range and the first road section range are corrected respectively to reduce the influence of non-meteorological factors such as human subjectivity on the time and space of traffic and meteorological disaster events, thereby improving the rationality and reliability of model testing.

[0031] In some embodiments, the time deviation amount and the space deviation amount can be calculated based on a large amount of data experiments, or can be pre-set based on empirical values.

[0032] In some embodiments, based on the spatiotemporal deviation data, the first time range and the first road section range are respectively corrected to obtain the second time range and the second road section range of each traffic and meteorological disaster event. The implementation process may include: for each traffic and meteorological disaster event, based on the traffic and meteorological disaster information of the traffic and meteorological disaster event, determining the target time deviation amount and the target space deviation amount from the spatiotemporal deviation data; based on the target time deviation amount, respectively correcting the start time and the end time of the first time range of the traffic and meteorological disaster event to obtain the second time range of the traffic and meteorological disaster event; based on the target space deviation amount, respectively correcting the start time and the end time of the first road section range of the traffic and meteorological disaster event to obtain the second road section range of the traffic and meteorological disaster event.

[0033] As a possible implementation, the time deviation of each traffic and meteorological disaster type may include a leading time deviation and a trailing time deviation. The leading time deviation refers to the deviation at the start time of the traffic and meteorological disaster event, and the trailing time deviation refers to the deviation at the end time of the traffic and meteorological disaster event. The spatial deviation of each traffic and meteorological disaster type may include a leading spatial deviation and a trailing spatial deviation, wherein the leading spatial deviation refers to the deviation at the starting point of the road section of the traffic and meteorological disaster event, and the trailing spatial deviation refers to the deviation at the end point of the road section of the traffic and meteorological disaster event.

[0034] As an example, both the lead time deviation and the post time deviation can be positive or negative. If the lead time deviation is negative, it means that the start time of the traffic meteorological disaster event lags behind the start time of the meteorological conditions that trigger the traffic meteorological disaster event; if the lead time deviation is positive, it means that the start time of the traffic meteorological disaster event is earlier than the start time of the meteorological conditions that trigger the traffic meteorological disaster event; if the post time deviation is negative, it means that the end time of the traffic meteorological disaster event lags behind the time when the meteorological conditions that trigger the traffic meteorological disaster event actually disappear; if the post time deviation is positive, it means that the end time of the traffic meteorological disaster event is earlier than the time when the meteorological conditions that trigger the traffic meteorological disaster event actually disappear. Similarly, if the leading spatial deviation is negative, it means that the starting point of the road section of the traffic meteorological disaster event is located after the starting point of the actual meteorological conditions that cause the traffic meteorological disaster event; if the leading spatial deviation is positive, it means that the starting point of the road section of the traffic meteorological disaster event is located before the starting point of the actual meteorological conditions that cause the traffic meteorological disaster event; if the trailing spatial deviation is negative, it means that the end point of the road section of the traffic meteorological disaster event is located after the end point of the actual meteorological conditions that cause the traffic meteorological disaster event; if the trailing spatial deviation is positive, it means that the end point of the road section of the traffic meteorological disaster event is located before the end point of the actual meteorological conditions that cause the traffic meteorological disaster event.

[0035] For example, if the first time range of traffic and meteorological disaster event A is from 08:00 on October 25, 2024 to 08:30 on October 25, 2024, then the start time of the first time range of traffic and meteorological disaster event A is 08:00 on October 25, 2024, and the end time is 08:30 on October 25, 2024; if the target time deviation includes a lead time deviation of -30min and a post time deviation of 20min, the correction process is equivalent to advancing the start time of the first time range of traffic and meteorological disaster event A by 30 minutes and delaying the end time of the first time range by 20 minutes, and obtaining the second time range of traffic and meteorological disaster event A from 07:30 on October 25, 2024 to 08:50 on October 25, 2024. The process of correcting the first road section range of a traffic meteorological disaster event is similar to that of correcting the time range, so that the corrected time range is close to the time range where the meteorological conditions that actually trigger the traffic meteorological disaster event exist, and the corrected road section range is close to the road section range where the meteorological conditions that actually trigger the traffic meteorological disaster event exist.

[0036] Step 103, based on the second time range and the second road section range of each traffic meteorological disaster event, determine the test sample of each traffic meteorological disaster event from the risk forecast data output by the highway traffic meteorological disaster risk forecast model to be tested.

[0037] That is to say, based on the second time range and second road section range of each traffic meteorological disaster event obtained after correction, the risk forecast data output by the highway traffic meteorological disaster risk forecast model to be tested is matched, and the matched risk forecast data is used as a test sample for each traffic meteorological disaster event, so that the occurrence of traffic meteorological disaster events can be used as evidence of the real existence of traffic meteorological disaster risks, and is used to evaluate whether the risk forecast data output by the highway traffic meteorological disaster risk forecast model to be tested is correct.

[0038] Since the second time range and second road section range of each traffic meteorological disaster event after correction exclude the influence of non-meteorological factors such as human subjectivity, the second time range and second road section range of each traffic meteorological disaster event are closer to the time range and road section range of the actual meteorological conditions that cause the traffic meteorological disaster event. Therefore, based on the second time range and second road section range of each traffic meteorological disaster event, the test sample of each traffic meteorological disaster event is determined from the risk forecast data output by the highway traffic meteorological disaster risk forecast model to be tested, which can improve the accuracy of the obtained test samples, and then improve the reliability of the model test.

[0039] In some embodiments, each forecast unit of the highway traffic meteorological disaster risk forecast model to be tested can be a point or a road section. For example, the forecast unit can be a point determined based on the pile number, a point determined based on the longitude and latitude, a road section determined based on the pile number, a road section determined based on the longitude and latitude, etc., and the risk forecast data outputted by it can be the risk forecast result of each forecast point, or the risk forecast result of each forecast road section. Based on the second time range and the second road section range of each traffic meteorological disaster event, the implementation process of determining the inspection sample of each traffic meteorological disaster event from the risk forecast data outputted by the highway traffic meteorological disaster risk forecast model to be tested can include: based on a preset conversion method, for each traffic meteorological disaster event, determining the target forecast unit contained in the second road section range of the traffic meteorological disaster event, and obtaining the target risk forecast data from the risk forecast data outputted by the highway traffic meteorological disaster risk forecast model to be tested; the forecast unit of the target risk forecast data is the target forecast unit, and the forecast time of the target risk forecast data is within the second time range of the traffic meteorological disaster event.

[0040] It should be noted that the risk forecast data output by the road traffic meteorological disaster risk forecast model to be tested can be for the next 2 hours, the next 12 hours, the next 24 hours, etc., and the forecast data each time can be divided into time periods. For example, the road traffic meteorological disaster risk forecast model to be tested outputs the risk forecast results for the next 12 hours each time, and each risk forecast result includes two data, one risk forecast result for the next 0-6 hours and one risk forecast result for the next 6-12 hours. When determining the test sample of a certain traffic meteorological disaster event, as long as the forecast time corresponding to a certain risk forecast result intersects with the second time range of the traffic meteorological disaster event, the risk forecast result will be used as a data in the test sample of the traffic meteorological disaster event.

[0041] For example, if the highway traffic meteorological disaster risk forecast model to be tested outputs risk forecast data at 08:00 on October 25, 2024, including a risk forecast result A for the next 0-6 hours and a risk forecast result B for the next 6-12 hours, then the starting time of risk forecast result A and risk forecast result B is 08:00 on October 25, 2024, the forecast period is 6 hours, the forecast time of risk forecast result A is from 08:01 on October 25, 2024 to 14:00 on October 25, 2024, and the forecast time of risk forecast result B is from 14:01 on October 25, 2024 to 20:00 on October 25, 2024.

[0042] In some embodiments, in order to improve the matching efficiency between the second road section range of the traffic and meteorological disaster event and the forecast unit of the model, the efficiency of determining the test sample is improved, such as Figure 2 As shown, based on the second time range and second road section range of each traffic meteorological disaster event, the implementation process of determining the test sample of each traffic meteorological disaster event from the risk forecast data output by the highway traffic meteorological disaster risk forecast model to be tested may include the following steps.

[0043] Step 201, for each traffic meteorological disaster event, based on a preset road section unit table, determine the target road section unit that intersects with the second road section range of the traffic meteorological disaster event; wherein the road section unit table includes the road section ranges of multiple road section units, and the corresponding relationship between each road section unit and the forecast unit of the highway traffic meteorological disaster risk forecast model to be tested; the multiple road section units are obtained by segmenting each highway section according to a preset length.

[0044] In some embodiments, the road section unit table may be pre-constructed, and its construction process may include: segmenting the selected highway section or all highway sections supported by the highway traffic meteorological disaster risk forecast model to be tested according to a preset length to obtain multiple road section units and the road section range of each road section unit; wherein the road section range of each road section unit may be expressed by a road section code and a pile number; based on the location information of each forecast unit of the highway traffic meteorological disaster risk forecast model to be tested and the road section range of each road section unit, determining the matching relationship between each road section unit and the forecast unit to obtain a road section unit table. The road section unit table may include the identification of each road section unit, the road section code, the starting pile number, the ending pile number of each road section unit, and the identification of the corresponding forecast unit.

[0045] When the location information of the forecast unit of the highway traffic meteorological disaster risk forecast model to be tested is determined by longitude and latitude, the road section unit is matched with the forecast unit. First, all highway sections can be geographically located in the GIS (Geographic Information System) map at 1km stake number; then all longitude and latitude information of the highway traffic meteorological disaster risk forecast model to be tested is read in for geographical coordinate location in the GIS map; finally, the nearest section code and stake number within 1km of each forecast unit are found by using the proximity distance method in GIS spatial analysis technology and marked as the section code and stake number of the forecast unit. Based on the section code and stake number corresponding to each forecast unit, as well as the section code and stake number contained in each section unit, the corresponding relationship between each section unit and each forecast unit is determined.

[0046] It should be noted that, usually, the division accuracy of the road section unit is greater than or equal to the accuracy of the forecast unit.

[0047] In some embodiments, when matching the second road section range of the traffic and meteorological disaster event with the road section unit, as long as the road section unit and the second road section range have an intersection, the road section unit is determined as the target road section unit. The number of target road section units can be one or more.

[0048] Step 202: Determine the target forecast unit corresponding to the target road section unit based on the road section unit table.

[0049] Step 203, based on the target forecast unit and the second time range of the traffic meteorological disaster event, determine the target risk forecast data from the risk forecast data; the target risk forecast data is the risk forecast data whose forecast time is within the second time range and the forecast unit is the target forecast unit.

[0050] Step 204: determine the target risk forecast data as a test sample of traffic meteorological disaster events.

[0051] After the test samples of each traffic meteorological disaster event are determined, step 104 is executed to evaluate the highway traffic meteorological disaster risk forecasting model to be tested.

[0052] Step 104 , based on the respective test samples of a plurality of traffic meteorological disaster events, the highway traffic meteorological disaster risk forecasting model to be tested is evaluated.

[0053] In some embodiments, the occurrence of each traffic meteorological disaster event indicates the existence of a corresponding traffic meteorological disaster risk, so the accuracy of the risk forecast data output by the model can be determined based on the test sample of each traffic meteorological disaster event, so that the highway traffic meteorological disaster risk forecast model to be tested can be evaluated. The test indicators may include the accuracy rate, the missed rate, the risk level accuracy rate, the risk level over-rate, the risk level under-rate, etc.

[0054] Specifically, the prediction accuracy of the risk or non-risk output by the highway traffic meteorological disaster risk forecasting model can be evaluated based on the test samples of multiple traffic meteorological disaster events. It is also possible to mark the risk level as being too large or too small based on the difference between the disaster level of the traffic meteorological disaster event and the risk level in the test sample, and evaluate the risk level accuracy, risk level too large rate, and risk level too small rate of the highway traffic meteorological disaster risk forecasting model.

[0055] In some embodiments, the test samples may also include data such as traffic meteorological disaster types, reporting time, and forecast timeliness, so the forecast results may also be horizontally counted based on the above data, such as the forecast accuracy of different traffic meteorological disaster types, the forecast accuracy of different reporting timeliness, and the forecast accuracy of different forecast timeliness. In specific implementation, a multi-dimensional and comprehensive evaluation of the highway traffic meteorological disaster risk forecast model to be tested may be conducted based on the needs of the actual scenario.

[0056] According to the testing method of the highway traffic meteorological disaster risk forecasting model of the embodiment of the present invention, the traffic meteorological disaster event information of the selected highway section within a preset time range is obtained; wherein the traffic meteorological disaster event information includes the first time range, the first section range and the traffic meteorological disaster information of each of the multiple traffic meteorological disaster events; based on the time-space deviation data, the first time range and the first section range are respectively corrected to obtain the second time range and the second section range of each traffic meteorological disaster event; based on the second time range and the second section range of each traffic meteorological disaster event, the test sample of each traffic meteorological disaster event is determined from the risk forecast data output by the highway traffic meteorological disaster risk forecasting model to be tested; based on the test samples of each of the multiple traffic meteorological disaster events, the highway traffic meteorological disaster risk forecasting model to be tested is evaluated. The present invention starts with highway traffic meteorological disaster events to realize the verification of the highway traffic meteorological disaster risk forecasting model, and takes into account the impact of non-meteorological factors on the occurrence of traffic meteorological disaster events. The time range and road section range of traffic meteorological disaster events are corrected based on spatiotemporal deviation data, so that the obtained test samples can be more in line with the actual situation, thereby improving the reliability of the model test results, and the evaluation results can also be used as the basis for model optimization to improve the accuracy of highway traffic meteorological disaster risk forecasting.

[0057] In some embodiments, the time deviation corresponding to each traffic and meteorological disaster type includes a leading time deviation and a trailing time deviation. Next, the process of determining the time deviation corresponding to any traffic and meteorological disaster type is introduced.

[0058] Figure 3 The third flow chart of the method for testing the highway traffic meteorological disaster risk forecast model provided by the embodiment of the present invention is as follows. Figure 3 As shown, the time deviation corresponding to any traffic meteorological disaster type is determined by the following steps.

[0059] Step 301, obtaining historical traffic and meteorological disaster event information corresponding to the traffic and meteorological disaster type within a historical time period; wherein the historical traffic and meteorological disaster event information includes the historical time range and historical road section range of each of a plurality of historical traffic and meteorological disaster events.

[0060] That is to say, for any type of traffic and meteorological disaster, the historical traffic and meteorological disaster events corresponding to the type of traffic and meteorological disaster in the historical time period are obtained.

[0061] Among them, the historical time period can be pre-set, can be a continuous time period, and the length of the historical time period can be 1 year, 2 years, etc. The historical time range of each historical traffic and meteorological disaster event refers to the time range between the start time and the end time of the historical traffic and meteorological disaster event, which is consistent with the first time range of the traffic and meteorological disaster event in the above embodiment. The historical section range of each historical traffic and meteorological disaster event refers to the section range of the historical traffic and meteorological disaster event, which can be the section range from the starting pile number to the ending pile number of the highway line, which is consistent with the first section range of the traffic and meteorological disaster event in the above embodiment.

[0062] Step 302, for the first historical traffic and meteorological disaster event in the historical traffic and meteorological disaster events, based on the first historical road section range and the first historical time range of the first historical traffic and meteorological disaster event, and the search time, obtain the first historical meteorological observation data set corresponding to the first historical traffic and meteorological disaster event.

[0063] That is to say, the historical traffic meteorological observation data at each time within the first historical road section and within the first historical time range is obtained and used as the historical meteorological observation data set corresponding to the historical traffic meteorological disaster event.

[0064] In some embodiments, meteorological observations within different road sections are performed by different meteorological stations, so it is necessary to determine the target meteorological station corresponding to the first historical traffic meteorological disaster event and then obtain the first historical meteorological observation data set.

[0065] As a possible implementation method, based on the first historical road section range and the first historical time range of the first historical traffic meteorological disaster event, as well as the search time, the first historical meteorological observation data set corresponding to the first historical traffic meteorological disaster event is obtained: based on the first historical road section range, the target meteorological station corresponding to the first historical traffic meteorological disaster event is determined; based on the first historical time range and the search time, the first historical meteorological observation data set is determined from the historical meteorological observation database of the target meteorological station.

[0066] Among them, the historical road section range of historical traffic meteorological disaster events is usually located by pile number, and the meteorological station is usually located by longitude and latitude. When matching meteorological stations based on the historical road section range, the longitude and latitude information corresponding to each highway pile number can be queried based on the relevant database of the transportation department, and a matching table between meteorological stations and highway pile numbers can be established. Then, based on the matching table and the highway pile number range included in the historical road section range of historical traffic meteorological disaster events, the target meteorological station corresponding to the historical traffic meteorological disaster event is determined.

[0067] In some embodiments, if the longitude and latitude information corresponding to each highway pile number cannot be obtained, the "Create Routes" tool in the ArcGIS linear referencing function can be used to merge the original highway information to create a single route (such as connecting the originally segmented Beijing-Harbin Expressway from the starting point to the end point to form a route); use the "Make Route Event Layer" tool in the ArcGIS linear referencing function to read in the 1km pile number information of the target highway to achieve the geographic coordinate positioning of the 1km pile number in the GIS map; read in the longitude and latitude information of the meteorological station, and use the proximity distance method in the GIS spatial analysis technology to determine the nearest meteorological station within 1km of each pile number, and establish a matching table between the meteorological station and the highway pile number.

[0068] In some embodiments, the search time can be pre-set, for example, it can be determined based on the research of traffic safety management work experience of each type of traffic meteorological disaster. As an example, the time deviation setting value corresponding to each type of traffic meteorological disaster is determined by the traffic control department through research, and the search time can be determined based on the time deviation setting value. For example, the time deviation setting value of the traffic control disaster type caused by fog is 2h, and the search time can be determined as 2×2h=4h. The search time refers to the search range before and after the first risk start time and the first risk end time around the first historical time range. Based on the first historical time range and the search time, the implementation process of determining the first historical meteorological observation data set from the historical meteorological observation database of the target meteorological site includes: subtracting the search time from the start time of the first historical time range to obtain the earliest time of the search time range, adding the search time to the end time of the first historical time range to obtain the latest time of the search time range; obtaining the historical meteorological observation data of each time within the search time range from the historical meteorological observation database of the target meteorological site, and determining it as the first historical meteorological observation data set.

[0069] Step 303, based on the highway traffic meteorological disaster risk forecasting model to be tested and the historical meteorological observation data at each time in the first historical meteorological observation data set, a risk search is performed to determine the first risk start time and the first risk end time of the first historical traffic meteorological disaster event.

[0070] In some embodiments, the historical meteorological observation data at each moment in the first historical meteorological observation data set can be input into the highway traffic meteorological disaster risk forecasting model to be tested in sequence to obtain the actual risk results at each moment, and based on the actual risk results at each moment, the first moment when the traffic meteorological disaster risk changes from nothing to something, and the second moment when the traffic meteorological disaster risk changes from something to nothing are determined; if there is only one first moment within the search time range, the first moment is determined as the start moment of the first risk; if there are multiple first moments, the first moment closest to the start moment of the first historical time range is determined as the start moment of the first risk; if there is only one second moment within the search time range, the second moment is determined as the end moment of the first risk; if there are multiple second moments, the second moment closest to the end moment of the first historical time range is determined as the end moment of the first risk.

[0071] In some embodiments, in order to improve the efficiency of determining the first risk start time and the first risk end time, the following method can be used: Figure 4 The steps shown determine a first risk start time and a first risk end time.

[0072] Step 401, taking the start time of the first historical time range as the first starting point, according to the preset step size and search time, the historical meteorological observation data of each moment before the first starting point in the first historical meteorological observation data set is input into the highway traffic meteorological disaster risk forecasting model to be tested, and based on the output risk actual results, the first risk start time is determined.

[0073] Among them, the actual risk results refer to the risk assessment results output by the highway traffic meteorological disaster risk forecast model to be tested based on the input meteorological observation data.

[0074] As an example, the road traffic meteorological disaster risk forecasting model to be tested can be used to forecast the road traffic meteorological disaster risk in the next 24 hours. The step size can be set to 1 hour, that is, when sampling historical meteorological observation data, not all historical meteorological observation data are collected at every moment, but the time interval between two adjacent sampling points is 1 hour.

[0075] Usually, due to the influence of non-meteorological factors such as human subjectivity, the time range for the occurrence of traffic meteorological disaster events is mostly smaller than the actual time range for the meteorological conditions that cause traffic meteorological disaster events. Therefore, when determining the start time of the first risk, you can give priority to searching from the time before the first starting point. When determining the end time of the first risk, you can give priority to searching from the time after the second starting point.

[0076] As a possible implementation method, the start time of the first historical time range can be used as the first starting point, and the historical meteorological observation data of the first starting point (first sampling point), the historical meteorological observation data of the second sampling point located before the first sampling point and separated by a preset step length, and the historical meteorological observation data of the third sampling point located before the second sampling point and separated by a preset step length are sequentially input into the highway traffic meteorological disaster risk forecast model to be tested to obtain the risk actual result corresponding to the current sampling point; if the current sampling point is the nth sampling point, the risk actual result corresponding to the current sampling point is no risk and the risk actual result corresponding to the n-1th sampling point is risky, then the time of the n-1th sampling point is determined as the first risk start time, and the search is stopped. If the search for the time before the first starting point has reached the earliest time of the search time range, and the first risk start time has not been searched, then step 402 is continued.

[0077] Step 402: If the first risk start time is not searched from the various moments before the first starting point, the historical meteorological observation data of the various moments after the first starting point in the first historical meteorological observation data set are input into the highway traffic meteorological disaster risk forecasting model to be tested in sequence according to the step size and the search time, and the first risk start time is determined based on the output risk actual results.

[0078] That is to say, if the first risk start time is not searched from the various moments before the first starting point, the start time of the first historical time range will continue to be used as the first starting point, and the historical meteorological observation data of the first sampling point located after the first starting point and at a preset step length, and the historical meteorological observation data of the second sampling point located after the first sampling point and at a preset step length will be input into the highway traffic meteorological disaster risk forecasting model to be tested in sequence to obtain the actual risk result corresponding to the current sampling point; if the actual risk result corresponding to the current sampling point is that there is a risk, the moment of the sampling point will be determined as the first risk start time, and the search will be stopped.

[0079] Step 403, taking the end time of the first historical time range as the second starting point, according to the step size and search time, the historical meteorological observation data of each moment after the second starting point in the first historical meteorological observation data set is input into the highway traffic meteorological disaster risk forecasting model to be tested, and based on the output risk actual results, the first risk end time is determined.

[0080] As a possible implementation method, the end time of the first historical time range can be used as the second starting point, and the historical meteorological observation data of the second starting point (first sampling point), the historical meteorological observation data of the second sampling point located after the first sampling point and separated by a preset step length, and the historical meteorological observation data of the third sampling point located after the second sampling point and separated by a preset step length are sequentially input into the highway traffic meteorological disaster risk forecast model to be tested to obtain the risk actual result corresponding to the current sampling point; if the current sampling point is the mth sampling point, the risk actual result corresponding to the current sampling point is no risk and the risk actual result corresponding to the m-1th sampling point is risky, then the moment of the mth sampling point is determined as the first risk end time, and the search is stopped. If the search for the moment after the second starting point has reached the latest moment of the search time range, and the first risk end time has not been found, then step 404 is continued.

[0081] Step 404: If the first risk end time is not searched from the moments after the second starting point, the historical meteorological observation data of the moments before the second starting point in the historical meteorological observation data set are input into the highway traffic meteorological disaster risk forecasting model to be tested in sequence according to the step size and the search time, and the first risk end time is determined based on the output risk actual results.

[0082] That is to say, if the end time of the first risk is not searched from the moments after the second starting point, the end time of the first historical time range is continued to be used as the second starting point, and the historical meteorological observation data of the first sampling point located before the second starting point and at a preset step length, and the historical meteorological observation data of the second sampling point located before the first sampling point and at a preset step length are sequentially input into the road traffic meteorological disaster risk forecasting model to be tested to obtain the actual risk result corresponding to the current sampling point; if the actual risk result corresponding to the current sampling point is no risk, the moment of the sampling point is determined as the end time of the first risk, and the search is stopped.

[0083] Step 304: Determine a first time deviation of the first historical traffic and meteorological disaster event based on the difference between the start time of the first historical time range and the first risk start time.

[0084] In some embodiments, the difference between the first risk start time and the start time of the first historical time range can be determined as the first time deviation of the first historical traffic and meteorological disaster event. If the first time deviation is positive, it means that the start time of the first historical traffic and meteorological disaster event is ahead of the start time of the actual occurrence of the meteorological conditions that trigger the first historical traffic and meteorological disaster event; if the first time deviation is negative, it means that the start time of the first historical traffic and meteorological disaster event lags behind the start time of the actual occurrence of the meteorological conditions that trigger the first historical traffic and meteorological disaster event.

[0085] Step 305: Determine a second time deviation of the first historical traffic and meteorological disaster event based on the difference between the end time of the first historical time range and the end time of the first risk.

[0086] In some embodiments, the difference between the end time of the first risk and the end time of the first historical time range can be determined as the second time deviation of the first historical traffic and meteorological disaster event. If the second time deviation is positive, it means that the end time of the first historical traffic and meteorological disaster event is ahead of the actual disappearance time of the meteorological conditions that caused the first historical traffic and meteorological disaster event; if the second time deviation is negative, it means that the end time of the first historical traffic and meteorological disaster event lags behind the actual disappearance time of the meteorological conditions that caused the first historical traffic and meteorological disaster event.

[0087] Step 306 , based on the first time deviation and the second time deviation of each of the plurality of historical traffic and meteorological disaster events, determine the leading time deviation and the trailing time deviation corresponding to the traffic and meteorological disaster type.

[0088] In some embodiments, the first time deviations of multiple historical traffic and meteorological disaster events can be averaged, and the obtained average value can be determined as the leading time deviation corresponding to the traffic and meteorological disaster type. The second time deviations of multiple historical traffic and meteorological disaster events can be averaged, and the obtained average value can be determined as the trailing time deviation corresponding to the traffic and meteorological disaster type.

[0089] In other embodiments, the first time deviation and the second time deviation of each of multiple historical traffic and meteorological disaster events can also be subjected to frequency distribution statistics according to a preset step interval, and the first time deviation with the highest frequency can be determined as the leading time deviation corresponding to the traffic and meteorological disaster type, and the second time deviation with the highest frequency can be determined as the post-time deviation corresponding to the traffic and meteorological disaster type.

[0090] It should be noted that the spatial deviation amount of each traffic meteorological disaster type in the embodiment of the present invention can also be determined in a similar manner as described above. Taking the spatial deviation amount of the traffic meteorological disaster type as an example, for the first historical traffic meteorological disaster event among multiple historical traffic meteorological disaster events, based on the starting point and end point of the first historical road section range, the search road section range, and the start and end times of the first historical time range, determine the second historical meteorological observation data set of each meteorological station related to the search road section range at the start time of the first historical time range, and the third historical meteorological observation data set of each meteorological station related to the search road section range at the end time of the first historical time range; based on the highway traffic meteorological disaster risk forecast model to be tested and the historical meteorological observation data of each meteorological station in the second historical meteorological observation data set, perform a risk search forward and then backward to determine the first risk starting highway of the first historical traffic meteorological disaster event. Pile number; based on the highway traffic meteorological disaster risk forecast model to be tested and the historical meteorological observation data of each meteorological station in the third historical meteorological observation data set, the risk search is performed backward and then forward to determine the first risk end highway pile number of the first historical traffic meteorological disaster event; based on the section length between the first risk start highway pile number and the starting point of the first historical road section range, the first spatial deviation of the first historical traffic meteorological disaster event is determined; based on the section length between the first risk end highway pile number and the end point of the first historical road section range, the second spatial deviation of the first historical traffic meteorological disaster event is determined; based on the first spatial deviation and the second spatial deviation of each of the multiple historical traffic meteorological disaster events, the leading spatial deviation and the trailing spatial deviation corresponding to the traffic meteorological disaster type are determined.

[0091] According to the verification method of the highway traffic meteorological disaster risk forecasting model of the embodiment of the present invention, a large amount of time information and road section information of historical traffic meteorological disaster events, as well as historical meteorological observation data, are combined, and the actual risk results are calculated through the highway traffic meteorological disaster risk forecasting model to be tested, and the risk start time and risk end time are determined through search, so as to obtain a more accurate time deviation corresponding to each traffic meteorological disaster type, thereby improving the reliability of the model verification.

[0092] The following describes a verification device for a highway traffic meteorological disaster risk forecasting model provided by an embodiment of the present invention. The verification device for a highway traffic meteorological disaster risk forecasting model described below and the verification method for a highway traffic meteorological disaster risk forecasting model described above can be referenced to each other.

[0093] Figure 5 The schematic diagram of the structure of the testing device of the highway traffic meteorological disaster risk forecast model provided by the embodiment of the present invention. Figure 5As shown, the device includes: a first acquisition module 510, a correction module 520, a first determination module 530 and an evaluation module 540.

[0094] The first acquisition module 510 is used to acquire the traffic and meteorological disaster event information of the selected highway section within a preset time range; wherein the traffic and meteorological disaster event information includes the first time range, the first road section range and the traffic and meteorological disaster information of each of the multiple traffic and meteorological disaster events; The correction module 520 is used to correct the first time range and the first road section range respectively based on the time-space deviation data to obtain the second time range and the second road section range of each traffic meteorological disaster event; wherein the time-space deviation data includes the time deviation amount and the space deviation amount corresponding to different traffic meteorological disaster types; the time deviation amount and the space deviation amount are used to characterize the time deviation and space deviation between the actual existing meteorological conditions that cause the traffic meteorological disaster event and the actual traffic meteorological disaster event respectively; The first determination module 530 is used to determine the test sample of each traffic meteorological disaster event from the risk forecast data output by the highway traffic meteorological disaster risk forecast model to be tested based on the second time range and the second road section range of each traffic meteorological disaster event; The evaluation module 540 is used to evaluate the highway traffic meteorological disaster risk forecast model to be tested based on the test samples of each of the multiple traffic meteorological disaster events.

[0095] In some embodiments, the correction module 520 is specifically used to: for each traffic and meteorological disaster event, based on the traffic and meteorological disaster information of the traffic and meteorological disaster event, determine the target time deviation and the target space deviation from the time and space deviation data; based on the target time deviation, respectively correct the start time and end time of the first time range of the traffic and meteorological disaster event to obtain the second time range of the traffic and meteorological disaster event; based on the target space deviation, correct the start and end points of the first section range of the traffic and meteorological disaster event to obtain the second section range of the traffic and meteorological disaster event.

[0096] In some embodiments, the time deviation corresponding to each traffic and meteorological disaster type includes a leading time deviation and a trailing time deviation; the device also includes a second determination module 550, which is used to determine the time deviation of each traffic and meteorological disaster type.

[0097] As a possible implementation method, the second determination module 550 is specifically used to: obtain historical traffic meteorological disaster event information corresponding to the traffic meteorological disaster type within the historical time period; wherein the historical traffic meteorological disaster event information includes the historical time range and historical road section range of multiple historical traffic meteorological disaster events; for the first historical traffic meteorological disaster event in the historical traffic meteorological disaster events, based on the first historical road section range and the first historical time range of the first historical traffic meteorological disaster event, and the search time, obtain the first historical meteorological observation data set corresponding to the first historical traffic meteorological disaster event; based on the highway traffic meteorological disaster risk forecasting model to be tested and the historical meteorological observation data at each moment in the first historical meteorological observation data set, perform risk search to determine the first risk start time and the first risk end time of the first historical traffic meteorological disaster event; based on the difference between the start time of the first historical time range and the first risk start time, determine the first time deviation of the first historical traffic meteorological disaster event; based on the difference between the end time of the first historical time range and the first risk end time, determine the second time deviation of the first historical traffic meteorological disaster event; based on the first time deviation and the second time deviation of each of the multiple historical traffic meteorological disaster events, determine the lead time deviation and the post time deviation corresponding to the traffic meteorological disaster type.

[0098] In some embodiments, the second determination module 550 is also used to: determine the target meteorological site corresponding to the first historical traffic meteorological disaster event based on the first historical road section range; determine the first historical meteorological observation data set from the historical meteorological observation database of the target meteorological site based on the first historical time range and search time.

[0099] In some embodiments, the second determination module 550 is also used to: take the start time of the first historical time range as the first starting point, and input the historical meteorological observation data of each moment before the first starting point in the first historical meteorological observation data set into the highway traffic meteorological disaster risk forecasting model to be tested in sequence according to the preset step size and search time, and determine the first risk start time based on the output risk actual situation result; if the first risk start time is not searched from the various moments before the first starting point, input the historical meteorological observation data of each moment after the first starting point in the first historical meteorological observation data set into the highway traffic meteorological disaster risk forecasting model to be tested in sequence according to the step size and search time, and based on the output risk actual situation result , determine the start time of the first risk; take the end time of the first historical time range as the second starting point, and input the historical meteorological observation data of each moment after the second starting point in the first historical meteorological observation data set into the highway traffic meteorological disaster risk forecasting model to be tested according to the step size and the search time, and determine the end time of the first risk based on the output risk actual situation result; if the end time of the first risk is not searched from the moments after the second starting point, input the historical meteorological observation data of each moment before the second starting point in the first historical meteorological observation data set into the highway traffic meteorological disaster risk forecasting model to be tested according to the step size and the search time, and determine the end time of the first risk based on the output risk actual situation result.

[0100] In some embodiments, the first determination module 530 is specifically used to: for each traffic and meteorological disaster event, based on a preset road section unit table, determine a target road section unit that intersects with the second road section range of the traffic and meteorological disaster event; wherein the road section unit table includes the road section ranges of multiple road section units, and the correspondence between each road section unit and the forecast unit of the highway traffic and meteorological disaster risk forecast model to be tested; multiple road section units are obtained by segmenting each highway section according to a preset length; based on the road section unit table, determine the target forecast unit corresponding to the target road section unit; based on the target forecast unit and the second time range of the traffic and meteorological disaster event, determine the target risk forecast data from the risk forecast data; the target risk forecast data is the risk forecast data whose forecast time is within the second time range and the forecast unit is the target forecast unit; determine the target risk forecast data as the test sample of the traffic and meteorological disaster event.

[0101] Figure 6 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 6As shown, the electronic device may include: a processor 610, a communication interface 620, a memory 630 and a communication bus 640, wherein the processor 610, the communication interface 620 and the memory 630 communicate with each other through the communication bus 640. The processor 610 may call the computer program in the memory 630 to execute the steps of the verification method of the highway traffic meteorological disaster risk forecast model.

[0102] For example, the method includes: obtaining traffic meteorological disaster event information of a selected highway section within a preset time range; wherein the traffic meteorological disaster event information includes the first time range, the first section range and the traffic meteorological disaster information of each of the multiple traffic meteorological disaster events; based on the time-space deviation data, respectively correcting the first time range and the first section range to obtain the second time range and the second section range of each traffic meteorological disaster event; wherein the time-space deviation data includes the time deviation amount and the space deviation amount corresponding to each of the different traffic meteorological disaster types; the time deviation amount and the space deviation amount are respectively used to characterize the time deviation and space deviation between the actual meteorological conditions that trigger the traffic meteorological disaster event and the actual traffic meteorological disaster event; based on the second time range and the second section range of each traffic meteorological disaster event, determining the test sample of each traffic meteorological disaster event from the risk forecast data output by the highway traffic meteorological disaster risk forecast model to be tested; based on the test samples of each of the multiple traffic meteorological disaster events, evaluating the highway traffic meteorological disaster risk forecast model to be tested.

[0103] In addition, the logic instructions in the above-mentioned memory 630 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a 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, etc. Various media that can store program codes.

[0104] On the other hand, an embodiment of the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the steps of the verification method of the highway traffic meteorological disaster risk forecasting model provided in the above-mentioned embodiments.

[0105] On the other hand, an embodiment of the present invention further provides a processor-readable storage medium, which stores a computer program, and the computer program is used to enable the processor to execute the steps of the verification method of the highway traffic meteorological disaster risk forecasting model provided in the above embodiments.

[0106] The processor-readable storage medium can be any available medium or data storage device that can be accessed by the processor, including but not limited to magnetic storage (such as floppy disks, hard disks, magnetic tapes, magneto-optical disks (MO), etc.), optical storage (such as CD, DVD, BD, HVD, etc.), and semiconductor storage (such as ROM, EPROM, EEPROM, non-volatile memory (NANDFLASH), solid-state drive (SSD)), etc.

[0107] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0108] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0109] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for testing a highway traffic meteorological disaster risk forecasting model, characterized in that: include: Acquire traffic and meteorological disaster event information of a selected highway section within a preset time range; wherein the traffic and meteorological disaster event information includes a first time range, a first road section range and traffic and meteorological disaster information of each of a plurality of traffic and meteorological disaster events; Based on the time-space deviation data, the first time range and the first road section range are respectively corrected to obtain the second time range and the second road section range of each traffic meteorological disaster event; wherein the time-space deviation data includes the time deviation amount and the space deviation amount corresponding to different traffic meteorological disaster types; the time deviation amount and the space deviation amount are respectively used to characterize the time deviation situation and the space deviation situation between the actual existing meteorological conditions that trigger the traffic meteorological disaster event and the actual traffic meteorological disaster event; Based on the second time range and the second road section range of each traffic meteorological disaster event, determining a test sample for each traffic meteorological disaster event from the risk forecast data output by the highway traffic meteorological disaster risk forecast model to be tested; Based on the test samples of each of the multiple traffic meteorological disaster events, the highway traffic meteorological disaster risk forecasting model to be tested is evaluated.

2. The method according to claim 1, characterized in that The first time range and the first road section range are respectively corrected based on the spatiotemporal deviation data to obtain the second time range and the second road section range of each traffic meteorological disaster event, including: For each of the traffic and meteorological disaster events, based on the traffic and meteorological disaster information of the traffic and meteorological disaster event, determining a target time deviation amount and a target space deviation amount from the time and space deviation data; Based on the target time deviation, respectively correct the start time and the end time of the first time range of the traffic and meteorological disaster event to obtain a second time range of the traffic and meteorological disaster event; Based on the target spatial deviation, the start point and the end point of the first road section range of the traffic and meteorological disaster event are respectively corrected to obtain the second road section range of the traffic and meteorological disaster event.

3. The method according to claim 1, characterized in that: The time deviation corresponding to each traffic and meteorological disaster type includes a leading time deviation and a trailing time deviation; the time deviation corresponding to any traffic and meteorological disaster type is determined by: Acquire historical traffic meteorological disaster event information corresponding to the traffic meteorological disaster type within a historical time period; wherein the historical traffic meteorological disaster event information includes the historical time range and historical road section range of each of a plurality of historical traffic meteorological disaster events; For a first historical traffic and meteorological disaster event among the historical traffic and meteorological disaster events, based on a first historical road section range and a first historical time range of the first historical traffic and meteorological disaster event, and a search time, obtaining a first historical meteorological observation data set corresponding to the first historical traffic and meteorological disaster event; Based on the highway traffic meteorological disaster risk forecasting model to be tested and the historical meteorological observation data at each time in the first historical meteorological observation data set, a risk search is performed to determine the first risk start time and the first risk end time of the first historical traffic meteorological disaster event; Determining a first time deviation of the first historical traffic and meteorological disaster event based on a difference between the start time of the first historical time range and the start time of the first risk; Determining a second time deviation of the first historical traffic and meteorological disaster event based on a difference between the end time of the first historical time range and the end time of the first risk; Based on the first time deviation and the second time deviation of each of the plurality of historical traffic and meteorological disaster events, the leading time deviation and the trailing time deviation corresponding to the traffic and meteorological disaster type are determined.

4. The method according to claim 3, characterized in that The acquiring of a first historical meteorological observation data set corresponding to the first historical traffic meteorological disaster event based on the first historical road section range and the first historical time range, as well as the search time, includes: Based on the first historical road section range, determining a target meteorological station corresponding to the first historical traffic meteorological disaster event; Based on the first historical time range and the search time, the first historical meteorological observation data set is determined from a historical meteorological observation database of the target meteorological site.

5. The method according to claim 3, characterized in that: The risk search is performed based on the highway traffic meteorological disaster risk forecast model to be tested and the historical meteorological observation data at each time in the first historical meteorological observation data set to determine the first risk start time and the first risk end time of the first historical traffic meteorological disaster event, including: Taking the start time of the first historical time range as the first starting point, according to the preset step size and the search time, the historical meteorological observation data of each time before the first starting point in the first historical meteorological observation data set is sequentially input into the highway traffic meteorological disaster risk forecasting model to be tested, and based on the output risk actual situation result, the first risk start time is determined; If the first risk start time is not found from the various moments before the first starting point, the historical meteorological observation data of each moment after the first starting point in the first historical meteorological observation data set are sequentially input into the highway traffic meteorological disaster risk forecasting model to be tested according to the step size and the search time, and the first risk start time is determined based on the output risk actual situation result; Taking the end time of the first historical time range as the second starting point, according to the step size and the search time, sequentially inputting the historical meteorological observation data at each time after the second starting point in the first historical meteorological observation data set into the highway traffic meteorological disaster risk forecasting model to be tested, and determining the first risk end time based on the output risk actual situation result; If the first risk end time is not searched from the various moments after the second starting point, the historical meteorological observation data of the various moments before the second starting point in the first historical meteorological observation data set are input into the road traffic meteorological disaster risk forecasting model to be tested in sequence according to the step size and the search time, and the first risk end time is determined based on the output risk actual situation results.

6. The method according to any one of claims 1 to 5, characterized in that The step of determining a test sample for each traffic meteorological disaster event based on the second time range and the second road section range of each traffic meteorological disaster event from the risk forecast data output by the road traffic meteorological disaster risk forecast model to be tested comprises: For each of the traffic meteorological disaster events, based on a preset road section unit table, a target road section unit that intersects with the second road section range of the traffic meteorological disaster event is determined; wherein the road section unit table includes the road section ranges of a plurality of road section units, and the corresponding relationship between each of the road section units and the forecast unit of the highway traffic meteorological disaster risk forecast model to be tested; the plurality of road section units are obtained by segmenting each highway section according to a preset length; Based on the road section unit table, determining a target forecast unit corresponding to the target road section unit; Based on the target forecast unit and the second time range of the traffic meteorological disaster event, target risk forecast data is determined from the risk forecast data; the target risk forecast data is the risk forecast data whose forecast time is within the second time range and the forecast unit is the target forecast unit; The target risk forecast data is determined as a test sample of the traffic meteorological disaster event.

7. A testing device for a highway traffic meteorological disaster risk forecasting model, characterized in that: include: A first acquisition module is used to acquire traffic and meteorological disaster event information of a selected highway section within a preset time range; wherein the traffic and meteorological disaster event information includes a first time range, a first road section range and traffic and meteorological disaster information of each of a plurality of traffic and meteorological disaster events; A correction module, for respectively correcting the first time range and the first road section range based on the time-space deviation data, to obtain a second time range and a second road section range for each of the traffic meteorological disaster events; wherein the time-space deviation data includes a time deviation amount and a space deviation amount corresponding to different types of traffic meteorological disasters; the time deviation amount and the space deviation amount are respectively used to characterize the time deviation and space deviation between the actual existing meteorological conditions that trigger the traffic meteorological disaster event and the actual traffic meteorological disaster event; A first determination module is used to determine a test sample for each traffic meteorological disaster event from the risk forecast data output by the highway traffic meteorological disaster risk forecast model to be tested based on the second time range and the second road section range of each traffic meteorological disaster event; An evaluation module is used to evaluate the highway traffic meteorological disaster risk forecasting model to be tested based on the test samples of each of the multiple traffic meteorological disaster events.

8. An electronic device comprising a processor and a memory storing a computer program, characterized in that: When the processor executes the computer program, the verification method of the highway traffic meteorological disaster risk forecasting model described in any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the verification method of the highway traffic meteorological disaster risk forecasting model according to any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the verification method of the highway traffic meteorological disaster risk forecasting model described in any one of claims 1 to 6 is implemented.

Images