A highway wind and snow early warning method, device, equipment and readable storage medium

Abstract

The present invention provides a highway blizzard warning method, device, equipment and readable storage medium, which relates to the technical field of highway disaster prevention, including constructing a criterion layer weight matrix based on the relative importance between criterion layers, and constructing a sub-criterion layer weight matrix based on the relative importance between sub-criterion layers; calculating the weight of the blizzard intensity factor according to the criterion layer weight matrix and the sub-criterion layer weight matrix; inputting the blizzard monitoring data at the current moment into a prediction model to obtain the prediction data of the blizzard intensity factor at a preset moment, wherein the time difference between the preset moment and the current moment is a first duration; performing weighted summation on the prediction data of the blizzard intensity factor and the corresponding weight to calculate the blizzard intensity at the preset moment. The present invention is used to solve the technical problems that the existing warning method has high requirements for camera cleaning and maintenance, can only reflect the current snow accumulation situation on the road, and the warning information has a lag and cannot be fed back in time.

Description

Technical Field

[0001] The present invention relates to the technical field of highway disaster prevention, and in particular to a highway wind and snow early warning method, device, equipment and readable storage medium. Background Art

[0002] Blowing snow refers to the phenomenon that the wind blows snow particles on the ground or snow surface and suspends them in the air. Blowing snow can reduce visibility, block roads, affect driving safety, and even cause traffic accidents. It is one of the common natural disasters on high-altitude highways in cold regions. Timely and effective monitoring and early warning of blowing snow sections are conducive to ensuring highway traffic safety and reducing economic losses. Existing blowing snow monitoring equipment is mainly based on video image analysis methods. It captures blowing snow scenes through cameras and uses image processing technology to extract blowing snow characteristic parameters, such as visibility and snow particle concentration. However, the above method has high requirements for camera cleaning and maintenance, and can only reflect the current snow accumulation on the road. The warning information has a lag and cannot be fed back in time. Summary of the invention

[0003] The purpose of the present invention is to provide a highway wind and snow early warning method, device, equipment and readable storage medium to improve the above problems. In order to achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0004] In a first aspect, the present application provides a highway wind and snow early warning method, comprising:

[0005] By using the hierarchical analysis method, the intensity of wind and snow is set as the target layer, the influencing factors of wind and snow intensity are set as the criterion layer, and the wind and snow intensity factors included in the criterion layer are set as the sub-criterion layer.

[0006] Constructing a criterion layer weight matrix based on the relative importance between criterion layers, and constructing a sub-criterion layer weight matrix based on the relative importance between sub-criterion layers;

[0007] The weight of the wind and snow intensity factor is calculated according to the criterion layer weight matrix and the sub-criterion layer weight matrix;

[0008] Acquire monitoring data of the intensity factor of wind and snow on the highway at the current moment, input the monitoring data of wind and snow at the current moment into the prediction model, and obtain prediction data of the intensity factor of wind and snow at a preset moment, wherein the time difference between the preset moment and the current moment is a first duration;

[0009] The predicted data of the wind and snow intensity factor and the corresponding weight are weightedly summed to calculate the wind and snow intensity at the preset time, and it is determined whether the wind and snow intensity at the preset time exceeds the warning threshold.

[0010] In a second aspect, the present application also provides a highway wind and snow warning device, comprising:

[0011] Hierarchy analysis module: using the hierarchy analysis method, the intensity of wind and snow is set as the target layer, the influencing factors of wind and snow intensity are set as the criterion layer, and the wind and snow intensity factors included in the criterion layer are set as the sub-criterion layer;

[0012] Weight matrix construction module: constructs a criterion layer weight matrix based on the relative importance between criterion layers, and constructs a sub-criterion layer weight matrix based on the relative importance between sub-criterion layers;

[0013] Calculation module: The weight of the wind and snow intensity factor is calculated according to the criterion layer weight matrix and the sub-criterion layer weight matrix;

[0014] Prediction module: respectively obtain monitoring data of the intensity factor of wind and snow on the highway at the current moment, input the monitoring data of wind and snow at the current moment into the prediction model, and obtain prediction data of the intensity factor of wind and snow at a preset moment, wherein the time difference between the preset moment and the current moment is a first duration;

[0015] Judgment module: The predicted data of the wind and snow intensity factor and the corresponding weight are weighted and summed to calculate the wind and snow intensity at the preset time, and judge whether the wind and snow intensity at the preset time exceeds the warning threshold.

[0016] In a third aspect, the present application also provides a highway wind and snow warning device, including:

[0017] Memory for storing computer programs;

[0018] A processor is used to implement the steps of the highway wind and snow warning method when executing the computer program.

[0019] In a fourth aspect, the present application further provides a readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the above-mentioned highway wind and snow warning method are implemented.

[0020] The beneficial effects of the present invention are:

[0021] By combining LSTM with the Transformer model, the present invention can learn different levels of feature representation from the time series, thereby more comprehensively expressing the information in the sequence, better processing the trend, periodicity and noise in the sequence, and improving the model's prediction accuracy for the wind and snow intensity factor at future moments. At the same time, the AHP hierarchical analysis method is used to determine the weight of each wind and snow intensity factor, and the wind and snow intensity is calculated according to the proportion, and the wind and snow intensity development is predicted in advance, so as to facilitate road control and reduce the impact of wind and snow on road traffic in high-cold areas.

[0022] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or be understood by implementing the embodiments of the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0024] Figure 1 A schematic diagram of the process of the highway wind and snow early warning method described in an embodiment of the present invention;

[0025] Figure 2 It is a grid structure diagram of the hierarchical analysis method described in an embodiment of the present invention;

[0026] Figure 3 This is a schematic structural diagram of a highway wind and snow warning device according to an embodiment of the present invention;

[0027] Figure 4 It is a schematic diagram of the structure of the highway wind and snow monitoring system described in an embodiment of the present invention;

[0028] Figure 5 Schematic diagram of the structure of the highway wind and snow warning device described in an embodiment of the present invention.

[0029] Markings in the figure:

[0030] 800. Highway wind and snow warning equipment; 801. Processor; 802. Memory; 803. Multimedia component; 804. I / O interface; 805. Communication component. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0032] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings. At the same time, in the description of the present invention, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0033] Embodiment 1:

[0034] See also Figure 1 This embodiment provides a highway wind and snow early warning method, including:

[0035] S1. Using the analytic hierarchy process, the intensity of wind and snow is set as the target layer C, the influencing factors of wind and snow intensity are set as the criterion layer Bi, and the wind and snow intensity factors included in the criterion layer are set as the sub-criterion layer Cij;

[0036] For details, please refer to Figure 2 The factors affecting the intensity of wind and snow include meteorological conditions B1 and topographic conditions B2;

[0037] The wind and snow intensity factors included in the meteorological condition B1 are: snow particle concentration C11, snow depth C12, wind speed C13, wind direction C14, temperature C15, temperature C16;

[0038] The wind and snow intensity factors of the terrain condition B2 are: altitude C21, terrain flatness C22, and vegetation coverage on both sides of the road C23.

[0039] Based on the above embodiments, the method further includes:

[0040] S2. constructing a criterion layer weight matrix based on the relative importance between criterion layers, and constructing a sub-criterion layer weight matrix based on the relative importance between sub-criterion layers;

[0041] Specifically, step S2 includes:

[0042] S21. Under the target layer, construct a first judgment matrix according to the importance of the criterion layer relative to the criterion layer;

[0043] Specifically, under the scaling criteria of the target layer, the indirect advantage comparison method is used to compare the elements in the criterion layer in pairs to judge their relative importance. Preferably, a 1-5 scaling method is used, where 1 indicates that the importance of the two elements is the same, 3 indicates that the former is slightly more important than the latter, and 5 indicates that the former is very important than the latter. Thus, the first judgment matrix W can be obtained:

[0044]

[0045] In the formula, w 11 represents the importance of meteorological condition B1 relative to meteorological condition B2, that is, 1, w 12 Indicates the importance of meteorological condition B1 relative to topographic condition B2, and so on.

[0046] S22. Using the eigenvalue method, perform eigendecomposition on the columns of the first judgment matrix to obtain a plurality of first eigenvectors, and standardize the plurality of eigenvectors into a plurality of first weight vectors;

[0047] Specifically, for w 11 and w 21 After feature decomposition and normalization, the first weight vector w is obtained a1 , for w 21 and w 22 After feature decomposition and normalization, the first weight vector w is obtained a2 ;

[0048] S23. Perform a consistency check on all first weight vectors. When the consistency check value is less than or equal to a preset check value, a criterion layer weight matrix consisting of multiple first weight vectors is obtained;

[0049] Specifically, the first weight vector is input into the YAANP software to calculate the consistency test value of the first judgment matrix W. If the consistency test value is less than or equal to the preset test value 0.1, it means that the first judgment matrix W meets the consistency requirement. Otherwise, the first weight vector w is readjusted. a1 , w a2 , so that it meets the consistency requirements, the criterion layer weight matrix w is composed of the first weight vector that meets the consistency requirements a =[w a1 , w a2 ] T ;

[0050] The step S2 further comprises:

[0051] S24. Under the criterion layer, construct a second judgment matrix according to the importance of the sub-criterion layer relative to the sub-criterion layer;

[0052] Specifically, under meteorological condition B1, the 1-5 scaling method is used to construct the second judgment matrix V1:

[0053]

[0054] Among them, v 11 Indicates the importance of snow particle concentration C11 relative to snow particle concentration C11, v 19 represents the importance of snow particle concentration C11 relative to the vegetation coverage C23 on both sides of the road, v 91 It indicates the importance of the vegetation coverage C23 on both sides of the road relative to the snow particle concentration C11, and so on.

[0055] Under the terrain condition B2, the 1-5 scaling method is used to construct the second judgment matrix V2:

[0056]

[0057] Similarly, v 11 ′ represents the importance of snow particle concentration C11 relative to snow particle concentration C11, and so on.

[0058] S25. Using the eigenvalue method, perform eigendecomposition on the columns of the second judgment matrix to obtain a plurality of second eigenvectors, and standardize the plurality of second eigenvectors into a plurality of second weight vectors;

[0059] Specifically, the second weight vector obtained by eigendecomposition of V1 is V a =[v a1 , v a2 …v a9 ] T ; The second weight vector obtained by eigendecomposition of V2 is V b =[v b1 , v b2 …v b9 ] T .

[0060] S26. Perform consistency check on all second weight vectors. Specifically, the consistency check method is the same as the check method in S23, which will not be repeated here. When the consistency check value is less than or equal to the preset check value, a sub-criteria layer weight matrix V consisting of multiple second weight vectors is obtained. ab =[V a ,V b ].

[0061] Based on the above embodiments, the method further includes:

[0062] S3. Calculate the weight of the wind and snow intensity factor according to the criterion layer weight matrix and the sub-criterion layer weight matrix;

[0063] Specifically, by W a *V ab =[w a1 , w a2 ] T *[V a ,V b ] The weights of the nine wind and snow intensity factors are calculated to be q 11 ,q 12 ,q 13 ,q 14 ,q 15 ,q 16 ,q 21 ,q 22 ,q 23 .

[0064] Based on the above embodiments, the method further includes:

[0065] S4. respectively obtain monitoring data of the intensity factor of the wind and snow on the highway at the current moment, input the monitoring data of the wind and snow at the current moment into the prediction model, and obtain the prediction data of the wind and snow intensity factor at the preset moment, wherein the time difference between the preset moment and the current moment is the first duration;

[0066] Specifically, the training process of the prediction model includes:

[0067] S41. Acquire multiple sets of monitoring data of the wind and snow intensity factor of the highway within a preset period of time, and pre-process the multiple sets of monitoring data. In this embodiment, acquire multiple sets of monitoring data of the wind and snow intensity factor of the highway within 20 days, and each set of monitoring data includes snow particle concentration, wind direction, wind speed, snow depth, temperature, and humidity;

[0068] The pre-processing comprises:

[0069] Encode the non-data types in the monitoring data and convert them into data types. For example, encode the wind directions in sequence: east [0], south [1], west [2], north [3], southeast [4], northeast [5], southwest [6], northwest [7];

[0070] Normalize the monitoring data and convert the normalized monitoring data into supervised learning data through the sliding window algorithm in, Indicated in Supervised learning data at each moment.

[0071] S42. The preprocessed monitoring data constitutes a data set, and the data set is divided into a training set and a test set according to a preset ratio;

[0072] Specifically, the preset time period is divided into a first time period and a second time period in a ratio of 7:3;

[0073] The monitoring data corresponding to the first period is used as the training set, and the monitoring data corresponding to the second period is used as the test set, that is, the monitoring data within the first 14 days is used as the training set, and the monitoring data within the last 6 days is used as the test set;

[0074] S43. Construct a bidirectional long short-term memory model LSTM, train and test the bidirectional long short-term memory model using a training set and a test set to obtain a prediction model;

[0075] Specifically, step S43 includes:

[0076] S431. Divide the first period into several moments;

[0077] S432. Determine an input time and a predicted time among several times, wherein the predicted time is after the input time;

[0078] For example, the consecutive n The first moment is used as the input moment, and the (t1+k)th moment is used as the predicted moment, wherein the time difference between t1 and the (t1+k) moment is the first duration k, and the first duration k can be 1 day, 2 days or 1 week.

[0079] S433. The monitoring data of the input time in the first period After inputting into the bidirectional long short-term memory model for learning, the prediction data of the prediction time is generated in chronological order.

[0080] Specifically, the bidirectional long short-term memory model (LSTM) includes four unit gates, namely a forget gate, an input gate, an update gate and an output gate. The four unit gates are used to selectively transmit information in a training set.

[0081] Preferably, the bidirectional long short-term memory model (LSTM) is added with a Transformer model, and the self-attention mechanism in the Transformer model is used to perform weighted aggregation on each position of the input wind and snow intensity factor monitoring data, and give a higher weight to the factors affecting the wind and snow intensity. The Transformer consists of an encoder (Encoder) and a decoder (Decoder). The Encoder encodes the monitoring data of the wind and snow intensity factor into a high-dimensional feature vector, so that the Decoder can make more accurate predictions based on the upper and lower dependencies of the sequence when outputting the generated sequence.

[0082] Supervised learning data The input is into the bidirectional long short-term memory model (LSTM) and the Transformer model. The relationship between the wind and snow data is mined and learned through LSTM and Transformer respectively. Then the output values ​​of the LSTM network and Transformer are input into the fully connected layer. The nonlinear relationship between the time series data of the two models is learned through the fully connected layer, and finally the prediction data at the prediction time is generated.

[0083] S434. Calculate the root mean square between the predicted data and the monitoring data at the prediction time. When the root mean square is less than the detection accuracy, obtain the prediction model:

[0084]

[0085] In the formula, RMSE means root mean square, is the predicted data of the i-th sample, is the real monitoring data of the ith sample, and m represents the total number of samples.

[0086] Based on the above embodiments, the method further includes:

[0087] S5. The predicted data of the wind and snow intensity factor and the corresponding weight are weighted and summed to calculate the wind and snow intensity at the preset time, and determine whether the wind and snow intensity at the preset time exceeds the warning threshold Q:

[0088] Q=C 11 *q 11 +C 12 *q 12 +…+C 23 *q 23 ;

[0089] When the warning threshold Q exceeds the warning threshold, control measures are taken in advance to control the road.

[0090] Embodiment 2:

[0091] like Figure 3 As shown, this embodiment provides a highway wind and snow warning device, the device comprising:

[0092] Hierarchy analysis module: using the hierarchy analysis method, the intensity of wind and snow is set as the target layer, the influencing factors of wind and snow intensity are set as the criterion layer, and the wind and snow intensity factors included in the criterion layer are set as the sub-criterion layer;

[0093] Weight matrix construction module: constructs a criterion layer weight matrix based on the relative importance between criterion layers, and constructs a sub-criterion layer weight matrix based on the relative importance between sub-criterion layers;

[0094] Calculation module: The weight of the wind and snow intensity factor is calculated according to the criterion layer weight matrix and the sub-criterion layer weight matrix;

[0095] Prediction module: respectively obtain monitoring data of the intensity factor of wind and snow on the highway at the current moment, input the monitoring data of wind and snow at the current moment into the prediction model, and obtain prediction data of the intensity factor of wind and snow at a preset moment, wherein the time difference between the preset moment and the current moment is a first duration;

[0096] Judgment module: The predicted data of the wind and snow intensity factor and the corresponding weight are weighted and summed to calculate the wind and snow intensity at the preset time, and judge whether the wind and snow intensity at the preset time exceeds the warning threshold.

[0097] Based on the above embodiments, the weight matrix construction module includes:

[0098] The first construction unit: under the target layer, constructing a first judgment matrix according to the importance of the criterion layer relative to the criterion layer;

[0099] A first eigendecomposition unit: using an eigenroot method, performing eigendecomposition on the columns of the first judgment matrix to obtain a plurality of first eigenvectors, and standardizing the plurality of eigenvectors into a plurality of first weight vectors;

[0100] The first checking unit performs a consistency check on all first weight vectors. When the consistency check value is less than or equal to a preset check value, a criterion layer weight matrix consisting of a plurality of first weight vectors is obtained.

[0101] Based on the above embodiments, the weight matrix construction module also includes:

[0102] The second construction unit: under the criterion layer, constructing a second judgment matrix according to the importance of the sub-criterion layer relative to the sub-criterion layer;

[0103] A second eigendecomposition unit: using an eigenroot method, performing eigendecomposition on the columns of the second judgment matrix to obtain a plurality of second eigenvectors, and standardizing the plurality of second eigenvectors into a plurality of second weight vectors;

[0104] The second checking unit performs a consistency check on all second weight vectors. When the consistency check value is less than or equal to a preset check value, a sub-criteria layer weight matrix consisting of a plurality of second weight vectors is obtained.

[0105] Based on the above embodiments, the prediction module includes:

[0106] Preprocessing unit: obtain multiple sets of monitoring data of wind and snow intensity factors of the highway within a preset period of time, and preprocess the multiple sets of monitoring data;

[0107] A first division unit: forming a data set from the preprocessed monitoring data, and dividing the data set into a training set and a test set according to a preset ratio;

[0108] Model training unit: construct a bidirectional long short-term memory model, use the training set and test set to train and test the bidirectional long short-term memory model, and obtain a prediction model.

[0109] Based on the above embodiment, the preprocessing unit includes:

[0110] Coding unit: encodes the non-data types in the monitoring data and converts them into data types;

[0111] Normalization processing unit: normalizes the monitoring data and converts the normalized monitoring data into supervised learning data through a sliding window algorithm.

[0112] Based on the above embodiment, the preprocessed monitoring data constitutes a data set, and the data set is divided into a training set and a test set according to a preset ratio, including:

[0113] Divide the preset time period into the first time period and the second time period in a ratio of 7:3;

[0114] The monitoring data corresponding to the first period is used as the training set, and the monitoring data corresponding to the second period is used as the test set.

[0115] Based on the above embodiment, the model training unit includes:

[0116] A second division unit: dividing the training set into a plurality of moments;

[0117] A selection unit: determining an input time and a predicted time from among a plurality of time instants, wherein the predicted time instant is after the input time instant;

[0118] Learning unit: After inputting the monitoring data of the input time in the training set into the bidirectional long short-term memory model for learning, the prediction data of the prediction time is generated in chronological order;

[0119] Calculation unit: Calculate the root mean square between the predicted data and the monitoring data at the prediction time. When the root mean square is less than the detection accuracy, a prediction model is obtained.

[0120] It should be noted that, regarding the device in the above embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment of the method, and will not be elaborated here.

[0121] Embodiment 3:

[0122] See also Figure 4 This embodiment also provides a highway wind and snow monitoring system, the system comprising:

[0123] A wind power generation module, used to use wind energy to provide power for the monitoring system;

[0124] A satellite communication module is used to exchange data with a cloud information processing module and a terminal receiving module via satellite signals;

[0125] A snow particle concentration monitoring module is used to monitor the concentration of snow particles in real time through a laser transmitter and a photoelectric detector, and send the snow particle concentration data to the satellite communication module;

[0126] A snow depth monitoring module is used to monitor the snow depth in real time through an ultrasonic transmitter and an ultrasonic receiver, and send the snow depth data to a satellite communication module;

[0127] Wind direction and speed monitoring module, used to monitor wind speed and direction in real time through wind speed sensor and wind direction sensor, and send wind speed and direction data to satellite communication module;

[0128] The cloud information processing module is used to receive the data transmitted by the satellite communication module, and predict the intensity of the wind and snow according to the highway wind and snow prediction method, and classify the signal level according to the wind and snow intensity, and send the signal level to the terminal receiving module;

[0129] The terminal receiving module is used to receive the signal sent by the cloud information processing module and take corresponding measures according to the signal level, including displaying the current signal level and related data on the display screen, the alarm emitting different sound and light prompts, and the road control device implementing different traffic control measures.

[0130] Embodiment 4:

[0131] Corresponding to the above method embodiment, this embodiment also provides a highway wind and snow warning device. The highway wind and snow warning device described below and the highway wind and snow warning method described above can refer to each other.

[0132] Figure 5 FIG. 8 is a block diagram of a highway wind and snow warning device 800 according to an exemplary embodiment. Figure 5 As shown, the highway wind and snow warning device 800 may include: a processor 801 and a memory 802. The highway wind and snow warning device 800 may also include one or more of a multimedia component 803, an I / O interface 804, and a communication component 805.

[0133] The processor 801 is used to control the overall operation of the highway wind and snow warning device 800 to complete all or part of the steps in the above-mentioned highway wind and snow warning method. The memory 802 is used to store various types of data to support the operation of the highway wind and snow warning device 800, and these data may include, for example, instructions for any application or method operated on the highway wind and snow warning device 800, and application-related data, such as contact data, sent and received messages, pictures, audio, video, etc. The memory 802 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, disk or optical disk. The multimedia component 803 may include a screen and an audio component. The screen may be, for example, a touch screen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signal may be further stored in the memory 802 or sent via the communication component 805. The audio component also includes at least one speaker for outputting audio signals. The I / O interface 804 provides an interface between the processor 801 and other interface modules, and the above-mentioned other interface modules can be keyboards, mice, buttons, etc. These buttons can be virtual buttons or physical buttons. The communication component 805 is used for wired or wireless communication between the highway wind and snow warning device 800 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G or 4G, or a combination of one or more of them, so the corresponding communication component 805 can include: Wi-Fi module, Bluetooth module, NFC module.

[0134] In an exemplary embodiment, the highway wind and snow warning device 800 can be implemented by one or more application specific integrated circuits (Application Specific Integrated Circuit, referred to as ASIC), digital signal processors (Digital Signal Processor, referred to as DSP), digital signal processing devices (Digital Signal Processing Device, referred to as DSPD), programmable logic devices (Programmable Logic Device, referred to as PLD), field programmable gate arrays (Field Programmable Gate Array, referred to as FPGA), controllers, microcontrollers, microprocessors or other electronic components to execute the above-mentioned highway wind and snow warning method.

[0135] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided, and when the program instructions are executed by a processor, the steps of the above-mentioned highway wind and snow warning method are implemented. For example, the computer-readable storage medium can be the above-mentioned memory 802 including program instructions, and the above-mentioned program instructions can be executed by the processor 801 of the highway wind and snow warning device 800 to complete the above-mentioned highway wind and snow warning method.

[0136] Embodiment 5:

[0137] Corresponding to the above method embodiment, a readable storage medium is also provided in this embodiment. The readable storage medium described below and the highway wind and snow warning method described above can refer to each other.

[0138] A readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the highway wind and snow warning method of the above method embodiment.

[0139] The readable storage medium may specifically be a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, or other readable storage medium that can store program codes.

[0140] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0141] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A highway wind and snow early warning method, characterized in that: include: By using the hierarchical analysis method, the intensity of wind and snow is set as the target layer, the influencing factors of wind and snow intensity are set as the criterion layer, and the wind and snow intensity factors included in the criterion layer are set as the sub-criterion layer. Constructing a criterion layer weight matrix based on the relative importance between criterion layers, and constructing a sub-criterion layer weight matrix based on the relative importance between sub-criterion layers; The weight of the wind and snow intensity factor is calculated according to the criterion layer weight matrix and the sub-criterion layer weight matrix; Acquire multiple sets of monitoring data of wind and snow intensity factors of the highway within a preset period of time, and pre-process the multiple sets of monitoring data; The preprocessed monitoring data is used to form a data set, and the data set is divided into a training set and a test set according to a preset ratio; A bidirectional long short-term memory model with a Transformer model added is constructed, and the bidirectional long short-term memory model is trained and tested using a training set and a test set to obtain a prediction model. The self-attention mechanism in the Transformer model is used to perform weighted aggregation on each position of the input wind and snow intensity factor monitoring data, and a higher weight is given to the factors affecting the wind and snow intensity. The monitoring data is input into the bidirectional long short-term memory model and the Transformer to respectively mine and learn the relationship between the wind and snow data, and then the output values ​​of the bidirectional long short-term memory model and the Transformer are input into the fully connected layer. The nonlinear relationship between the time series data of the two models is learned through the fully connected layer, and finally the prediction data at the prediction moment is generated; Acquire monitoring data of the intensity factor of wind and snow on the highway at the current moment, input the monitoring data of wind and snow at the current moment into the prediction model, and obtain prediction data of the intensity factor of wind and snow at a preset moment, wherein the time difference between the preset moment and the current moment is a first duration; The predicted data of the wind and snow intensity factor and the corresponding weight are weightedly summed to calculate the wind and snow intensity at the preset time, and it is determined whether the wind and snow intensity at the preset time exceeds the warning threshold.

2. The highway wind and snow early warning method according to claim 1 is characterized in that ,Based on the relative importance between criterion layers, a criterion layer weight matrix is ​​constructed, including: Under the target layer, a first judgment matrix is ​​constructed according to the importance of the criterion layer relative to the criterion layer; Using the eigenroot method, eigendecomposing the columns of the first judgment matrix to obtain a plurality of first eigenvectors, and standardizing the plurality of eigenvectors into a plurality of first weight vectors; A consistency check is performed on all first weight vectors. When the consistency check value is less than or equal to a preset check value, a criterion layer weight matrix consisting of multiple first weight vectors is obtained.

3. The highway snowstorm warning method according to claim 1 is characterized in that ,The bidirectional long short term memory model is trained and tested using the training set and the test set to obtain the prediction model, including: Dividing the training set into a number of moments; Determining an input time and a predicted time among a plurality of time instants, wherein the predicted time instant is after the input time instant; After the monitoring data of the input time in the training set is sequentially input into the bidirectional long short-term memory model for learning, the prediction data of the prediction time is generated in chronological order; The root mean square between the predicted data and the monitoring data at the prediction time is calculated, and when the root mean square is less than the detection accuracy, the prediction model is obtained.

4. A highway wind and snow warning device, characterized in that: include: Hierarchy analysis module: using the hierarchy analysis method, the intensity of wind and snow is set as the target layer, the influencing factors of wind and snow intensity are set as the criterion layer, and the wind and snow intensity factors included in the criterion layer are set as the sub-criterion layer; Weight matrix construction module: constructs a criterion layer weight matrix based on the relative importance between criterion layers, and constructs a sub-criterion layer weight matrix based on the relative importance between sub-criterion layers; Calculation module: The weight of the wind and snow intensity factor is calculated according to the criterion layer weight matrix and the sub-criterion layer weight matrix; Preprocessing unit: obtain multiple sets of monitoring data of wind and snow intensity factors of the highway within a preset period of time, and preprocess the multiple sets of monitoring data; A first division unit: forming a data set from the preprocessed monitoring data, and dividing the data set into a training set and a test set according to a preset ratio; Model training unit: construct a bidirectional long short-term memory model with a Transformer model added, train and test the bidirectional long short-term memory model using a training set and a test set to obtain a prediction model, wherein the self-attention mechanism in the Transformer model is used to perform weighted aggregation on each position of the input wind and snow intensity factor monitoring data, assign a higher weight to the factors affecting the wind and snow intensity, input the monitoring data into the bidirectional long short-term memory model and the Transformer to mine and learn the relationship between the wind and snow data, and then input the output values ​​of the bidirectional long short-term memory model and the Transformer into the fully connected layer, learn the nonlinear relationship between the time series data of the two models through the fully connected layer, and finally generate the prediction data at the prediction moment; Prediction module: respectively obtain monitoring data of the intensity factor of wind and snow on the highway at the current moment, input the monitoring data of wind and snow at the current moment into the prediction model, and obtain prediction data of the intensity factor of wind and snow at a preset moment, wherein the time difference between the preset moment and the current moment is a first duration; Judgment module: The predicted data of the wind and snow intensity factor and the corresponding weight are weighted and summed to calculate the wind and snow intensity at the preset time, and judge whether the wind and snow intensity at the preset time exceeds the warning threshold.

5. The highway wind and snow warning device according to claim 4 is characterized in that: The weight matrix building module includes: The first construction unit: under the target layer, constructing a first judgment matrix according to the importance of the criterion layer relative to the criterion layer; A first eigendecomposition unit: using an eigenroot method, performing eigendecomposition on the columns of the first judgment matrix to obtain a plurality of first eigenvectors, and standardizing the plurality of eigenvectors into a plurality of first weight vectors; The first checking unit performs a consistency check on all first weight vectors. When the consistency check value is less than or equal to a preset check value, a criterion layer weight matrix consisting of a plurality of first weight vectors is obtained.

6. The highway wind and snow warning device according to claim 4, characterized in that: The model training unit comprises: A second division unit: dividing the training set into a plurality of moments; A selection unit: selecting an input time and a predicted time from among several time points, wherein the predicted time point is after the input time point; Learning unit: After inputting the monitoring data of the input time in the training set into the bidirectional long short-term memory model for learning, the prediction data of the prediction time is generated in chronological order; Calculation unit: Calculate the root mean square between the predicted data and the monitoring data at the prediction time. When the root mean square is less than the detection accuracy, a prediction model is obtained.

7. A highway wind and snow warning device, characterized in that: include: Memory for storing computer programs; A processor is used to implement the steps of the highway wind and snow warning method as claimed in any one of claims 1 to 3 when executing the computer program.

8. A readable storage medium, characterized in that: The readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the highway wind and snow warning method as claimed in any one of claims 1 to 3 are implemented.

Images