Meteorological data transmission and sharing method for highways based on Internet of Things
By determining the mutation area in highway meteorological data and adjusting the sampling frequency according to its frequency domain changes, the problem of data redundancy in traditional sampling methods is solved, and more efficient data transmission is achieved.
Patent Information
- Application Number
- CN202411267505.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-11
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2044-09-11
AI Technical Summary
The mutation and trend of highway meteorological data lead to the traditional sampling method having data redundancy in areas without characteristic changes, which cannot effectively reduce the bandwidth required for data transmission.
Samples are performed by determining the mutation regions of the meteorological data sequence and determining the corresponding sampling frequency based on the changes in the frequency domain of these regions. Mutation regions adopt higher sampling frequency, while non-mutant regions adopt the original sampling frequency of Nyquist sampling theorem.
This method makes the sampling results closer to the original data characteristics, effectively reduces the bandwidth required for data transmission, and improves data transmission efficiency.
Smart Images

Figure CN118828254B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of data processing, and in particular to a method for transmitting and sharing meteorological data of a highway based on the Internet of Things. Background Art
[0002] The meteorological data of highways transmitted based on the Internet of Things can be mainly classified into the following categories, including temperature, humidity, precipitation and road conditions. In order to play the role of data transmission and sharing, such data needs to be accurately collected and compressed. For example, during the transmission process, the original data is sampled and processed to reduce the bandwidth occupied by the data and improve the transmission efficiency. By effectively transmitting and processing the meteorological data, highway management agencies can grasp the road and weather conditions in real time, effectively carry out traffic dispatch and safety management, and provide drivers with timely road information and driving suggestions.
[0003] In the process of transmitting highway meteorological data, in order to improve the transmission efficiency and increase the real-time performance of the transmitted data, it is necessary to sample the meteorological data to be transmitted and transmit the simplified data after sampling. Generally, in order to ensure that the sampled data has the complete characteristics of the original data, it is necessary to satisfy the Nyquist sampling theorem. However, due to the mutation and trend of highway meteorological data, the sampling rate determined by the mutation will have high data redundancy in areas where there are not many changes in data characteristics, which cannot effectively reduce the bandwidth required for data transmission and reduce the necessity of sampling operations. Summary of the invention
[0004] To solve the above problems, the present application provides a method for sharing meteorological data transmission on highways based on the Internet of Things. The method determines the mutation area according to the meteorological data value, and determines the corresponding sampling frequency to sample the mutation area according to the change of the mutation area in the frequency domain. This can make the sampling results closer to the original data characteristics, effectively reduce the bandwidth required for data transmission, and improve data transmission efficiency.
[0005] The present application provides a method for transmitting and sharing meteorological data of a highway based on the Internet of Things, which includes:
[0006] Acquire a meteorological data set, wherein the meteorological data set includes a plurality of meteorological data sequences, each meteorological data sequence representing a meteorological data value of a dimension;
[0007] Determine a mutation region corresponding to the meteorological data sequence, wherein the mutation region is a data region with drastic fluctuations in the meteorological data sequence;
[0008] Determine a change of each of the mutation regions in the frequency domain, and determine a first sampling frequency corresponding to the mutation region based on the change;
[0009] The mutation region is sampled based on the first sampling frequency, and the non-mutation region in the meteorological data sequence is sampled based on the second sampling frequency, wherein the second sampling frequency is the original sampling frequency of the Nyquist sampling theorem.
[0010] Wherein, determining the mutation region corresponding to the meteorological data sequence includes:
[0011] Calculate the meteorological mutation trend characteristics corresponding to each time node in the meteorological data sequence;
[0012] The time nodes when the meteorological mutation trend characteristics are in a preset interval are used as region division points, and the mutation regions corresponding to the meteorological data sequence are determined based on the region division points.
[0013] The calculation of the meteorological mutation trend characteristics corresponding to each time node in the meteorological data sequence includes:
[0014] Establishing a time-meteorological data value coordinate system of the meteorological data sequence, and calculating the Euclidean distance between the current time node and the left adjacent time node based on the coordinate system; wherein the meteorological data sequence includes meteorological data values collected every second, and each minute is regarded as a time node;
[0015] Determine the weight coefficient corresponding to the meteorological data value of the current time node based on the change between the meteorological data value at time node t and the meteorological data value at time node t-1, and the change between the meteorological data value at time node t-1 and the meteorological data value at time node t-2;
[0016] The meteorological mutation trend characteristics corresponding to the current time node are determined based on the Euclidean distance between the current time node and the left adjacent time node and the weight coefficient corresponding to the meteorological data value of the current time node; the Euclidean distance is positively correlated with the meteorological mutation trend characteristics corresponding to the current time node.
[0017] The weight coefficient corresponding to the meteorological data value at the current time node is calculated using the following formula:
[0018] ;
[0019] in, Indicates the weight coefficient corresponding to the meteorological data value at time node t, Indicates the change in the meteorological data value at time node t relative to the meteorological data value at time node t-1, Indicates the change in the meteorological data value at time node t-1 relative to the meteorological data value at time node t-2, express and The mean of Represents an exponential function with the natural constant e as the base.
[0020] The meteorological mutation trend characteristics corresponding to the current time node are calculated using the following formula:
[0021] ;
[0022] in, Indicates the meteorological mutation trend characteristics corresponding to time node t, Indicates the weight coefficient corresponding to the meteorological data value at time node t, represents the horizontal axis of time node t, represents the horizontal coordinate of time node t-1, Indicates the ordinate corresponding to time node t, represents the ordinate corresponding to the time node t-1, [-c, c] represents the neighborhood data points involved in constructing the Euclidean distance between the current time node and the left adjacent time node, Represents the normalization function.
[0023] Wherein, determining the change of each mutation region in the frequency domain includes:
[0024] Performing short-time Fourier transform on the meteorological data values in adjacent time nodes in the mutation area to obtain amplitude information in a corresponding range of each time node in the mutation area;
[0025] The abnormal value of the mutation area is calculated based on the maximum amplitude information, the minimum amplitude information, the total number of time nodes in the mutation area, the meteorological mutation trend characteristics corresponding to each time node in the mutation area, and the discrete degree of the amplitude information in the mutation area. The abnormal value of the mutation area represents the change of the mutation area in the frequency domain.
[0026] The abnormal value of the mutation region is calculated using the following formula:
[0027] ;
[0028] in, Indicates the abnormal value of the mutation area M, represents the maximum amplitude information among the amplitude information of the time nodes in the mutation region M, represents the minimum amplitude information among the amplitude information of the time nodes in the mutation region M, represents the total number of time nodes in the mutation region M, Indicates the nth time node in the mutation region M The characteristics of the sudden change trend of meteorological Indicates the nth time node in the mutation region M Amplitude information, Represents the mean value of the amplitude information of all time nodes in the mutation region M.
[0029] Wherein, determining the first sampling frequency corresponding to the mutation region based on the change situation includes:
[0030] Normalizing the abnormal value of the mutation region;
[0031] A first sampling frequency corresponding to the mutation region is obtained according to the normalized anomaly value and the second sampling frequency; the first sampling frequency corresponding to the mutation region is positively correlated with the anomaly value of the mutation region.
[0032] The first sampling frequency corresponding to the mutation region is obtained by using the following formula:
[0033] ;
[0034] in, represents the first sampling frequency corresponding to the mutation region M, Indicates the abnormal value of the mutation area M, represents the normalization function, represents the second sampling frequency, where Represents the mean value of the amplitude information of all time nodes in the mutation region M.
[0035] Wherein, the method further comprises:
[0036] The meteorological data set is subjected to denoising processing.
[0037] The beneficial effects of the present application are different from those of the prior art. The meteorological data transmission and sharing method for highways based on the Internet of Things of the present application determines the mutation area corresponding to the meteorological data sequence, determines the first sampling frequency corresponding to the mutation area based on the change of the mutation area in the frequency domain, and samples the mutation area based on the first sampling frequency. This can make the sampling result closer to the original data characteristics, effectively reduce the bandwidth required for data transmission, and improve data transmission efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work, among which:
[0039] Figure 1It is a flow chart of an embodiment of a method for transmitting and sharing meteorological data of a highway based on the Internet of Things according to the present invention;
[0040] Figure 2 for Figure 1 A schematic flow chart of an embodiment of step S12;
[0041] Figure 3 for Figure 1 Schematic diagram of a flow chart of an embodiment of step S13 in FIG. DETAILED DESCRIPTION
[0042] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0043] See also Figure 1 , Figure 1 This is a flow chart of an embodiment of a method for transmitting and sharing meteorological data of a highway based on the Internet of Things of the present invention, which specifically includes:
[0044] Step S11: Acquire a meteorological data set, wherein the meteorological data set includes a plurality of meteorological data sequences, and each meteorological data sequence represents a meteorological data value of a dimension.
[0045] This embodiment aims to effectively transmit multi-dimensional meteorological data of highways. Specifically, this application mainly optimizes sampling to achieve the purpose of improving transmission efficiency, so meteorological data values of different dimensions need to be obtained before sampling processing. Including but not limited to: highway air temperature data, temperature data, local precipitation data, visibility, etc. This type of data is obtained through sensors, related professional collection equipment, etc., and the data form is continuous data.
[0046] For example, the meteorological data sequence A1 of temperature data is obtained by collecting air temperature data, the meteorological data sequence A2 of temperature data is obtained by collecting temperature data, the meteorological data sequence A3 of precipitation data is obtained by collecting precipitation data, and the meteorological data sequence A4 of visibility data is obtained by collecting visibility data. Meteorological data sequences A1, A2, A3, and A4 are meteorological data sequences of four different dimensions. In a meteorological data sequence, the data collected each time, such as air temperature data, temperature data, precipitation data, and visibility data, are meteorological data values of corresponding dimensions.
[0047] In one embodiment, mean filtering may be used to perform denoising on the meteorological data set, in order to prevent noise data from affecting the sampling iteration results.
[0048] Step S12: determining a mutation region corresponding to the meteorological data sequence, wherein the mutation region is a data region with drastic fluctuations in the meteorological data sequence.
[0049] The characteristics of meteorological data for highways are quite diverse, mainly including periodicity, trend, mutation and abnormality. Specifically, periodicity is manifested in that temperature and humidity data usually show periodic increase and decrease due to day and night changes; trend is manifested in that temperature data gradually increases or decreases with seasonal changes; mutation and abnormality are manifested in that wind speed and rainfall may have sudden changes, which are usually not easy to capture through conventional prediction models, and the probability of occurrence is relatively random, and this type of mutation data has more data characteristics.
[0050] Highway meteorological data, such as wind speed and precipitation in a short period of time, will have a certain degree of mutation. The mutation areas of data in different dimensions are also inconsistent. Therefore, the analysis here is only for a single dimension such as wind speed. The analysis methods for other dimensions are the same.
[0051] The change of wind speed is relatively random and has a certain degree of unpredictability, so when sampling it, the sampling frequency will be difficult to obtain due to the random change of data. In order to optimize the sampling results and improve the transmission efficiency, it is first necessary to determine the data range or data area with similar change characteristics. The data area with similar change characteristics is similar or identical in sampling frequency. When the data is sampled, the problem of inapplicability of the sampling frequency will not occur due to the randomness of the data characteristics; on the contrary, the mutational data with drastic fluctuations should conform to the data fluctuation characteristics in sampling frequency in order to optimize the sampling results. Therefore, it is necessary to determine the data mutation area with drastic fluctuations. In this embodiment, the meteorological mutation trend characteristics corresponding to the meteorological data sequence are analyzed to obtain the mutation area. It can be understood that the more similar the meteorological mutation trend characteristics of the data in a certain interval are, the more they can be used as the same mutation area.
[0052] Based on this, please combine Figure 2 , Figure 2 for Figure 1 The flowchart of an embodiment of step S12 in the embodiment specifically includes:
[0053] Step S21: Calculate the meteorological mutation trend characteristics corresponding to each time node in the meteorological data sequence.
[0054] Take any of the meteorological data sequences A1, A2, A3, and A4 as an example for illustration. In a specific embodiment, the meteorological data sequence corresponding to the wind speed can be taken as an example. The meteorological data sequence includes meteorological data values collected every second, and each minute is used as a time node. First, determine the meteorological mutation trend characteristics corresponding to each time node in the meteorological data sequence.
[0055] In one embodiment, a time-meteorological data value coordinate system of the meteorological data sequence is established. It should be noted that the meteorological data values of the corresponding dimension are collected with seconds as the time unit, so as to obtain the meteorological data sequence of the corresponding dimension, and each minute is recorded as a time node. When establishing the coordinate system, the time node can be used as the horizontal coordinate, and the meteorological data value of the time node can be used as the vertical coordinate, that is, only the meteorological data value collected every minute is recorded. It is also possible to use every second as the horizontal coordinate, and the meteorological data value collected every second as the vertical coordinate, that is, to record the meteorological data value collected every second.
[0056] The Euclidean distance between time node t and the left adjacent time node t-1 is calculated based on the coordinate system. In this embodiment, the neighborhood data points within the neighborhood range of time node t are used to participate in the calculation of the Euclidean distance between time node t and the left adjacent time node t-1. The neighborhood data points within the neighborhood range of time node t can be, for example, c neighborhood data points on the left side of time node t and c neighborhood data points on the right side of time node t, denoted as [tc, t+c]. The calculated Euclidean distance between time node t and the left adjacent time node t-1 is denoted as:
[0057] ;
[0058] in, represents the horizontal axis of time node t, Represents the horizontal coordinate of the time node t-1 adjacent to the left of the time node t, Indicates the ordinate corresponding to time node t, represents the ordinate corresponding to the time node t-1, and [-c, c] represents the neighborhood data points involved in constructing the Euclidean distance between the time node t and the left adjacent time node t-1.
[0059] For example, the current time node is recorded as time node t, and the weight coefficient corresponding to the meteorological data value of the current time node t is determined based on the change between the meteorological data value of the current time node t and the meteorological data value of time node t-1, and the change between the meteorological data value of time node t-1 and the meteorological data value of time node t-2.
[0060] In one embodiment, the weight coefficient corresponding to the meteorological data value at the current time node t is calculated using the following formula:
[0061] ;
[0062] in, Indicates the weight coefficient corresponding to the meteorological data value at time node t, Indicates the change in the meteorological data value at time node t relative to the meteorological data value at time node t-1, Indicates the change in the meteorological data value at time node t-1 relative to the meteorological data value at time node t-2, express and The mean of Represents an exponential function with the natural constant e as the base.
[0063] The meteorological mutation trend feature corresponding to the current time node is determined based on the Euclidean distance between the current time node and the left adjacent time node and the weight coefficient corresponding to the meteorological data value of the current time node; the Euclidean distance is positively correlated with the meteorological mutation trend feature corresponding to the current time node. In one embodiment, the meteorological mutation trend feature corresponding to the current time node is calculated using the following formula:
[0064] ;
[0065] in, Indicates the meteorological mutation trend characteristics corresponding to time node t, Indicates the weight coefficient corresponding to the meteorological data value at time node t, represents the horizontal axis of time node t, represents the horizontal coordinate of time node t-1, Indicates the ordinate corresponding to time node t, represents the ordinate corresponding to the time node t-1, [-c, c] represents the neighborhood data points involved in constructing the Euclidean distance between the current time node and the left adjacent time node, Represents the normalization function.
[0066] The above formula, It represents the Euclidean distance between time node t and the left adjacent time node t-1. The larger the distance, the greater the change in the meteorological data value between time node t and time node t-1, and the higher the corresponding meteorological mutation trend characteristics. The logic corresponding to the scene is: when the meteorological wind data surges or plummets within a certain time node, then within a fixed time, the larger the Euclidean distance between the initial time point and the end time point in the set of data points with large changes, the more obvious the wind data change is, and the closer it is to the mutation data. During the sampling process, the corresponding period of such mutation data is small (the mutation lasts shorter than the normal situation), and a higher sampling frequency is required to ensure that the original data characteristics are not lost, so the continuous data area of such mutation data can be regarded as a mutation area. Moreover, the change of one data point is not enough to explain the mutation of wind data, because the interval between time nodes is very small. Therefore, here we choose to use several neighborhood data points to participate in the Euclidean distance summation calculation to determine the mutation area. It can be understood that the larger the overall sum value, the higher the corresponding meteorological mutation trend characteristics.
[0067] In the above formula, the weight coefficient corresponding to the meteorological data value at time node t is used The meteorological data value at time node t Weighting. The slower the change of data points, the lower the weight in calculating the Euclidean distance, because the slower the change and the more evenly the increase and decrease of data points are caused by the original characteristics of the original data, such as the temperature difference between day and night. The meteorological mutation trend characteristics of such data are very low, so the corresponding weight coefficient is lower, and the calculated meteorological mutation trend characteristics are. The specific logic of the formula is: and The closer they are, the more regular the changes. The closer it is to 1, the The closer it is to 1, and the slower the overall data changes, the smaller the difference in mean The smaller the weight coefficient is, the The lower it is, the lower the characteristics of the meteorological mutation trend will be.
[0068] In summary, the meteorological mutation trend characteristics corresponding to each time node can be calculated.
[0069] Step S22: taking the time nodes when the meteorological mutation trend characteristics are in a preset interval as region division points, and determining the mutation region corresponding to the meteorological data sequence based on the region division points.
[0070] The closer the meteorological mutation trend characteristic is to 1, the higher the mutation is. The time nodes are used as regional division points, and the regional division points participate in regional division, thereby determining the mutation area corresponding to the meteorological data series. The remaining undivided areas are used as other periodic segments. It should be noted here that for any region with mutation, after calculating the meteorological mutation trend characteristics, the segment belonging to the mutation area must have two points that meet the mutation characteristics, and the area composed of these two points is the entire mutation area.
[0071] Step S13: determining a change of each of the mutation regions in the frequency domain, and determining a first sampling frequency corresponding to the mutation region based on the change.
[0072] The data in the mutation area fluctuates more dramatically and contains more data feature information. In order to retain the characteristics of the meteorological data before sampling, a higher sampling frequency is required in the mutation area during sampling.
[0073] In one embodiment, please combine Figure 3 , Figure 3 for Figure 1 The flowchart of an embodiment of step S13 is shown in FIG. 1 , wherein step S13 specifically includes:
[0074] Step S31: performing short-time Fourier transform on the meteorological data values in adjacent time nodes in the mutation area to obtain amplitude information in a corresponding range of each time node in the mutation area.
[0075] In order to accurately obtain the periodic information of the mutation region, it is necessary to perform short-time Fourier transform on the meteorological data values in adjacent time nodes in the mutation region to obtain the amplitude information within the corresponding range of each time node in the mutation region.
[0076] Specifically, the data range divided by two adjacent time nodes in the mutation area is subjected to short-time Fourier transform to obtain its spectrum information, including amplitude information and phase information. Therefore, there are several amplitude information and phase information here. For example, in the mutation area, there are 100 time nodes, then there are 99 data ranges, corresponding to 99 amplitude information and phase information. Get amplitude information: ,here That is, it indicates the amplitude information size of the short-time Fourier transform within each time node.
[0077] Step S32: Based on the maximum amplitude information, the minimum amplitude information, the total number of time nodes in the mutation area, the meteorological mutation trend characteristics corresponding to each time node in the mutation area, and the discrete degree of the amplitude information in the mutation area, the abnormal value of the mutation area is calculated to obtain the abnormal value of the mutation area, which represents the change of the mutation area in the frequency domain.
[0078] In one embodiment, the abnormal value of the mutation region is calculated using the following formula:
[0079] ;
[0080] in, Indicates the abnormal value of the mutation area M, represents the maximum amplitude information among the amplitude information of the time nodes in the mutation region M, represents the minimum amplitude information among the amplitude information of the time nodes in the mutation region M, represents the total number of time nodes in the mutation region M, Indicates the nth time node in the mutation region M The characteristics of meteorological mutation trend, Indicates the nth time node in the mutation region M Amplitude information, Represents the mean value of the amplitude information of all time nodes in the mutation region M.
[0081] In the above formula, the abnormal value of the mutation area is intended to quantify the change of the meteorological data value in the mutation area in the frequency domain. The greater the difference in amplitude information, that is, The larger it is, the more it indicates that the data changes in the mutation area are irregular, because the amplitude reflects the frequency of the data, that is, the greater the difference in the frequency domain characteristics of the data, the less correlation there is in the time domain. Corresponding to the scene, for example, wind data, the wind force and wind direction change greatly in a short period of time, and the change is irregular, then such change information needs to be transmitted as feature data to play the role of highway weather warning, so The larger the result, the greater the anomaly value of the mutation area.
[0082] In the above formula, the summation result is Characterizes the discreteness of amplitude information in the mutation area, aiming to quantify the data discreteness of the spectrum amplitude. The higher the discreteness, the more drastic the change of meteorological data values at different time nodes, the higher the irrelevance, and the higher the corresponding abnormal value, the more it needs to be transmitted as feature data, so the corresponding sampling frequency needs to be increased. The above summation formula is similar to the variance calculation formula, but Here, it is used as the weight value in the original variance calculation process to weight the amplitude. The higher the meteorological mutation trend characteristics at this time node, that is, the higher the weight, we believe that the value of the data when calculating the amplitude variance will be higher, and the corresponding abnormal value will be higher.
[0083] In the mutation area, the larger the range of the data spectrum amplitude information changes, the higher the abnormal value of the spectrum meteorological period segment, because in this mutation area, the spectrum information corresponding to different time node segments is quite different, proving that the meteorological data (such as wind force) has irregular changes, such as wind direction and wind speed, etc. There are large changes in a short period of time. In order to retain such change characteristics, the sampling frequency needs to be higher.
[0084] In one embodiment, the anomaly value of the mutation region is normalized; a first sampling frequency corresponding to the mutation region is obtained according to the normalized anomaly value and the second sampling frequency; and the first sampling frequency corresponding to the mutation region is positively correlated with the anomaly value of the mutation region.
[0085] In a specific embodiment, the first sampling frequency corresponding to the mutation region is obtained using the following formula:
[0086] ;
[0087] in, represents the first sampling frequency corresponding to the mutation region M, Indicates the abnormal value of the mutation area M, represents the normalization function, represents the second sampling frequency, where Represents the mean value of the amplitude information of all time nodes in the mutation region M.
[0088] For the first sampling frequency, the latter Represents the second sampling frequency, which is the original sampling frequency of the Nyquist sampling theorem, and represents the sampling frequency value that conforms to the Nyquist sampling theorem under the amplitude mean. The higher the abnormal value of the mutation area M, the higher the corresponding first sampling frequency, and the more it can ensure that the data characteristics are retained after sampling.
[0089] In summary, each mutation region has a first sampling frequency, and for non-mutation regions, their sampling frequencies only need to conform to the Nyquist sampling theorem.
[0090] Step S14: sampling the mutation region based on the first sampling frequency, and sampling the non-mutation region in the meteorological data sequence based on a second sampling frequency, wherein the second sampling frequency is the original sampling frequency of the Nyquist sampling theorem.
[0091] Specifically, the mutation region is sampled using a first sampling frequency corresponding to the mutation region, and the non-mutation region is sampled using the original sampling frequency of the Nyquist sampling theorem, that is, the second sampling frequency.
[0092] By sampling in the above manner, each data area has a sampling frequency obtained according to the data characteristics. The sampling frequency reduces the total amount of data on the basis of ensuring the original data characteristics, reduces the bandwidth required for data transmission, and increases the efficiency of highway meteorological data transmission. Only the sampling results are transmitted subsequently, which plays a role in data transmission sharing.
[0093] The above are only implementation methods of the present application, and are not intended to limit the scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the protection scope of the present application.
Claims
1. A method for transmitting and sharing meteorological data on a highway based on the Internet of Things, characterized in that: include: Acquire a meteorological data set, wherein the meteorological data set includes a plurality of meteorological data sequences, each meteorological data sequence representing a meteorological data value of a dimension; Determine a mutation region corresponding to the meteorological data sequence, wherein the mutation region is a data region with drastic fluctuations in the meteorological data sequence; Determine a change in each of the mutation regions in the frequency domain, and determine a first sampling frequency corresponding to the mutation region based on the change; Sampling the mutation region based on the first sampling frequency, and sampling the non-mutation region in the meteorological data sequence based on a second sampling frequency, wherein the second sampling frequency is the original sampling frequency of the Nyquist sampling theorem; Determining the mutation region corresponding to the meteorological data sequence includes: Calculate the meteorological mutation trend characteristics corresponding to each time node in the meteorological data sequence; Taking the time node where the meteorological mutation trend feature is in a preset interval as the region division point, and determining the mutation region corresponding to the meteorological data sequence based on the region division point; Determining the change of each mutation region in the frequency domain includes: Performing short-time Fourier transform on the meteorological data values in adjacent time nodes in the mutation area to obtain amplitude information in a corresponding range of each time node in the mutation area; Based on the maximum amplitude information, the minimum amplitude information, the total number of time nodes in the mutation area, the meteorological mutation trend characteristics corresponding to each time node in the mutation area, and the discrete degree of the amplitude information in the mutation area, the abnormal value of the mutation area is calculated, and the abnormal value of the mutation area represents the change of the mutation area in the frequency domain; Determining a first sampling frequency corresponding to the mutation region based on the change condition includes: Normalizing the abnormal value of the mutation region; Obtaining a first sampling frequency corresponding to the mutation region according to the normalized anomaly value and the second sampling frequency; the first sampling frequency corresponding to the mutation region is positively correlated with the anomaly value of the mutation region; Calculating the meteorological mutation trend characteristics corresponding to each time node in the meteorological data sequence, including: Establishing a time-meteorological data value coordinate system of the meteorological data sequence, and calculating the Euclidean distance between the current time node and the left adjacent time node based on the coordinate system; wherein the meteorological data sequence includes meteorological data values collected every second, and each minute is regarded as a time node; Determine the weight coefficient corresponding to the meteorological data value of the current time node based on the change between the meteorological data value at time node t and the meteorological data value at time node t-1, and the change between the meteorological data value at time node t-1 and the meteorological data value at time node t-2; Determine the meteorological mutation trend feature corresponding to the current time node based on the Euclidean distance between the current time node and the left adjacent time node and the weight coefficient corresponding to the meteorological data value of the current time node; the Euclidean distance is positively correlated with the meteorological mutation trend feature corresponding to the current time node; The weight coefficient corresponding to the meteorological data value of the current time node is calculated using the following formula: ; in, Indicates the weight coefficient corresponding to the meteorological data value at time node t, Indicates the change in the meteorological data value at time node t relative to the meteorological data value at time node t-1, Indicates the change in the meteorological data value at time node t-1 relative to the meteorological data value at time node t-2, express and The mean of It represents the exponential function with the natural constant e as the base; The meteorological mutation trend characteristics corresponding to the current time node are calculated using the following formula: ; in, Indicates the meteorological mutation trend characteristics corresponding to time node t, Indicates the weight coefficient corresponding to the meteorological data value at time node t, represents the horizontal axis of time node t, represents the horizontal coordinate of time node t-1, Indicates the ordinate corresponding to time node t, represents the ordinate corresponding to the time node t-1, [-c, c] represents the neighborhood data points involved in constructing the Euclidean distance between the current time node and the left adjacent time node, represents the normalization function; The abnormal value of the mutation region is calculated using the following formula: ; in, Indicates the abnormal value of the mutation area M, represents the maximum amplitude information among the amplitude information of the time nodes in the mutation region M, represents the minimum amplitude information among the amplitude information of the time nodes in the mutation region M, represents the total number of time nodes in the mutation region M, Indicates the nth time node in the mutation region M The characteristics of the sudden change trend of meteorological Indicates the nth time node in the mutation region M Amplitude information, represents the mean value of the amplitude information of all time nodes in the mutation region M; The first sampling frequency corresponding to the mutation region is obtained using the following formula: ; in, represents the first sampling frequency corresponding to the mutation region M, Indicates the abnormal value of the mutation area M, represents the normalization function, represents the second sampling frequency, where Represents the mean value of the amplitude information of all time nodes in the mutation region M.
2. The method for transmitting and sharing meteorological data of a highway based on the Internet of Things according to claim 1 is characterized in that: The method further comprises: The meteorological data set is subjected to denoising processing.
Citation Information
Patent Citations
Big-data-drive-based air quality and resident trip visual analysis method and system
CN106991525A
A soil moisture data acquisition method and device
CN109711757A
Methods and systems for detection in industrial internet of things data collection environment with large data sets
CN110073301A
WGAN-based one-dimensional time series data augmentation method
CN113627594A
Data acquisition method, computer program product, equipment and computer medium
CN118260168A
