Ocean automatic meteorological station abnormity monitoring method and device based on multi-dimensional configuration characteristics

Abstract

The embodiment of the invention provides a method and a device for monitoring abnormity of an automatic ocean weather station based on multi-dimensional configuration characteristics, which are applied to the technical field of meteorological monitoring. The method comprises the following steps: acquiring meteorological monitoring data and position calibration data in a target station area; grouping stations in the target station area according to the meteorological monitoring data and the calibration data; respectively inputting the meteorological monitoring data of the stations in each group into a preset performance health degree model, and correspondingly outputting performance health indexes; and / or respectively inputting the meteorological monitoring data of the stations in each group into a preset configuration health degree model, and correspondingly outputting configuration health indexes; and determining abnormal sites according to the performance health indexes and / or the configuration health indexes of the sites in each group. In this way, real-time monitoring of the operation health degree of the coastal automatic weather station equipment, abnormal early warning and automatic recognition of fault points can be achieved.

Description

Technical Field

[0001] The present disclosure relates to the technical field of data processing, in particular to the technical field of meteorological monitoring, and specifically relates to a method and device for abnormal monitoring of an ocean automatic weather station based on multi-dimensional configuration features. Background Art

[0002] Ocean automatic weather stations play a crucial role in aspects such as ocean meteorological monitoring, maritime traffic safety, ocean resource development, and ocean disaster warning. In recent years, the number of automatic weather stations in the ocean and coastal areas has increased rapidly, and the ocean meteorological service has gradually taken shape. However, a large number of ocean automatic weather stations adopt an unattended mode, and the operation quality monitoring and guarantee links of the automatic weather station equipment have not met the requirements of modernization construction, which has restricted the effectiveness of the observation equipment to a certain extent.

[0003] Specifically, Figure 1 FIG. shows a schematic diagram of the composition of a typical automatic weather station. An automatic weather station mainly consists of intelligent measuring instruments for each element, multiple intelligent node controllers adapted to wired and wireless modes, an intelligent integrated processor, a solar power supply system, peripheral devices, and supporting software, and realizes the automatic observation of meteorological elements such as air temperature, air pressure, relative humidity, wind direction, wind speed, precipitation, visibility, etc., and has characteristics such as low power consumption, high reliability, high precision, high stability, easy expansion, and easy maintenance. Due to the fact that an automatic weather station includes various sensors, various transmission devices, distributed and integrated controllers, and various transmission protocols, etc., combined with the unattended situation, it is difficult to supervise whether its operation is normal and whether the monitoring is accurate, and it is difficult to quickly locate when monitoring interruptions, fault alarms, etc. occur, which brings challenges to troubleshooting and early guarantee. Summary of the Invention

[0004] The present disclosure provides a method and device for abnormal monitoring of an ocean automatic weather station based on multi-dimensional configuration features.

[0005] According to a first aspect of the present disclosure, there is provided a method for abnormal monitoring of an ocean automatic weather station based on multi-dimensional configuration features. The method includes:

[0006] Obtain meteorological monitoring data and position calibration data within the target site area;

[0007] Group the sites within the target site area according to the meteorological monitoring data and the position calibration data;

[0008] Input the meteorological monitoring data of the sites in each group into a preset performance health model respectively, and correspondingly output performance health indicators; and / or input the meteorological monitoring data of the sites in each group into a preset configuration health model respectively, and correspondingly output configuration health indicators;

[0009] Determine abnormal sites according to the performance health indicators and / or configuration health indicators of the sites within each group.

[0010] For the aspects and any possible implementation manners as described above, a further implementation manner is provided. The grouping of the sites within the target site area according to the meteorological monitoring data and the position calibration data includes:

[0011] Group the sites within the target site area according to the meteorological monitoring data and the position calibration data based on the k-means clustering algorithm.

[0012] For the aspects and any possible implementation manners as described above, a further implementation manner is provided. The construction of the preset performance health model includes:

[0013] Calculate the corresponding parameter validity according to the number of valid values in the meteorological monitoring data of the sites within each group;

[0014] Calculate the corresponding parameter integrity according to the number of actual collected values and the theoretical collection number in the meteorological monitoring data of the sites within each group;

[0015] Construct a preset performance health model based on the parameter validity and the parameter integrity.

[0016] For the aspects and any possible implementation manners as described above, a further implementation manner is provided. The construction of the preset configuration health model includes:

[0017] Based on the principal component analysis algorithm, calculate the consistency of the same-type parameters and the correlation of the different-type parameters according to the meteorological monitoring data of the sites within each group;

[0018] Construct a preset configuration health model based on the consistency of the same-type parameters and the correlation of the different-type parameters.

[0019] For the aspects and any possible implementation manners as described above, a further implementation manner is provided. The determination of abnormal sites according to the performance health indicators and / or configuration health indicators of the sites within each group includes:

[0020] Determine the sites with abnormal validity and / or integrity of the sensor measurement values according to the performance health indicators of the sites within each group;

[0021] Determine the sites with abnormal accuracy from the sites excluding the abnormal validity and / or integrity of the sensor measurement values according to the configuration health indicators of the sites within each group.

[0022] For the aspects and any possible implementation manners described above, a further implementation manner is provided. The meteorological monitoring data includes air pressure, temperature, humidity, wind speed, wind direction, precipitation, and visibility; the location calibration data includes longitude and latitude.

[0023] For the aspects and any possible implementation manners described above, a further implementation manner is provided. The method further includes:

[0024] Sending an early warning message for an abnormal site, where the early warning message for the abnormal site includes the station number of the abnormal site and the type of abnormal parameter.

[0025] According to a second aspect of the present disclosure, an abnormal monitoring device for an ocean automatic weather station based on multi-dimensional configuration features is provided. The device includes:

[0026] An acquisition module, configured to acquire meteorological monitoring data and location calibration data within a target site area;

[0027] A grouping module, configured to group the sites within the target site area according to the meteorological monitoring data and the location calibration data;

[0028] A generation module, configured to respectively input the meteorological monitoring data of the sites within each group into a preset performance health model, and correspondingly output performance health indicators; and / or respectively input the meteorological monitoring data of the sites within each group into a preset configuration health model, and correspondingly output configuration health indicators;

[0029] A determination module, configured to determine an abnormal site according to the performance health indicators and / or configuration health indicators of the sites within each group.

[0030] According to a third aspect of the present disclosure, an electronic device is provided. The electronic device includes: a memory and a processor. A computer program is stored on the memory, and when the processor executes the program, the method described above is implemented.

[0031] According to a fourth aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored. When the program is executed by a processor, the method described above is implemented.

[0032] An abnormal monitoring method and device for marine automatic weather stations based on multi-dimensional configuration features provided by an embodiment of the present application can obtain meteorological monitoring data and position calibration data within the target station area; then group the stations within the target station area according to the meteorological monitoring data and position calibration data; input the meteorological monitoring data of the stations in each group into a preset performance health model respectively, and correspondingly output performance health indicators; and / or input the meteorological monitoring data of the stations in each group into a preset configuration health model respectively, and correspondingly output configuration health indicators; then determine abnormal stations according to the performance health indicators and / or configuration health indicators of the stations in each group; based on this, in combination with the rapidly developing data analysis and artificial intelligence technologies, by constructing a health monitoring and abnormal identification model for coastal automatic weather stations, calculate and evaluate the health indicators of each element data of the observation stations, realize real-time monitoring of the operation health of coastal automatic weather station equipment, abnormal early warning and automatic identification of fault points, so as to provide support for improving the ability and efficiency of meteorological intelligent guarantee work.

[0033] It should be understood that the content described in the summary of the invention is not intended to limit the key or important features of the embodiments of the present disclosure, nor to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] With reference to the accompanying drawings and the following detailed description, the above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent. The drawings are used to better understand the solution and do not constitute a limitation to the present disclosure. In the drawings, the same or similar reference numerals represent the same or similar elements, where:

[0035] Figure 1 Shows a schematic diagram of the composition of an automatic weather station;

[0036] Figure 2 Shows a flowchart of an abnormal monitoring method for marine automatic weather stations based on multi-dimensional configuration features according to an embodiment of the present disclosure;

[0037] Figure 3 Shows a schematic diagram of the grouping result of stations within the target station area according to an embodiment of the present disclosure;

[0038] Figure 4 Shows a schematic diagram of an example of the validity result of station data according to an embodiment of the present disclosure;

[0039] Figure 5 Shows a schematic diagram of an example of the integrity result of station data according to an embodiment of the present disclosure;

[0040] Figure 6 Shows a schematic diagram of the consistency performance of the air temperature configuration of stations within a group according to an embodiment of the present disclosure;

[0041] Figure 7 Shows a schematic diagram of the deviation distribution of the in-group station air temperature configuration trend according to an embodiment of the present disclosure;

[0042] Figure 8 Shows a schematic diagram of the distribution of the in-group station air temperature configuration health index according to an embodiment of the present disclosure;

[0043] Figure 9 Shows a schematic diagram of the change trend of the in-group station configuration health index according to an embodiment of the present disclosure;

[0044] Figure 10 Shows a schematic diagram of the health monitoring and anomaly identification results of an automatic weather station according to an embodiment of the present disclosure;

[0045] Figure 11 Shows the monitoring algorithm framework of an ocean automatic weather station according to an embodiment of the present disclosure;

[0046] Figure 12 Shows a block diagram of an ocean automatic weather station anomaly monitoring device based on multi-dimensional configuration features according to an embodiment of the present disclosure;

[0047] Figure 13 Shows a block diagram of an exemplary electronic device capable of implementing the embodiments of the present disclosure. Detailed implementation manners

[0048] To make the objectives, technical solutions and advantages of the embodiments of the present disclosure clearer, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present disclosure. Obviously, the described embodiments are some but not all of the embodiments of the present disclosure. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present disclosure without creative efforts shall fall within the scope of protection of the present disclosure.

[0049] In addition, the term "and / or" in this article is merely a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this article generally represents an "or" relationship between the associated objects before and after.

[0050] In the present disclosure, in combination with the rapidly developing data analysis and artificial intelligence technologies, by constructing a coastal automatic weather station health monitoring and anomaly identification model, the health index of each element data of the observation station is calculated and evaluated, so as to realize the real-time monitoring of the operation health of the coastal automatic weather station equipment, anomaly early warning and automatic identification of fault points, thereby providing support for improving the ability and efficiency of meteorological intelligent guarantee work.

[0051] Figure 2 The flowchart of the abnormal monitoring method 200 for an ocean automatic weather station based on multi-dimensional configuration features according to an embodiment of the present disclosure is shown.

[0052] In block 210, meteorological monitoring data and position calibration data within the target station area are acquired.

[0053] In block 220, the stations within the target station area are grouped according to the meteorological monitoring data and the position calibration data.

[0054] In block 230, the meteorological monitoring data of the stations within each group are respectively input into a preset performance health model, and corresponding performance health indicators are output; and / or the meteorological monitoring data of the stations within each group are respectively input into a preset configuration health model, and corresponding configuration health indicators are output.

[0055] In block 240, abnormal stations are determined according to the performance health indicators and / or configuration health indicators of the stations within each group.

[0056] In block 210, the target station area can be set according to the actual needs of users. For example, according to the current distribution status of automatic weather stations, station areas such as coastal areas, islands, oil platforms, and moored buoys can be selected.

[0057] In some embodiments, the above-mentioned meteorological monitoring data includes air pressure, temperature, humidity, wind speed, wind direction, precipitation, and visibility; the above-mentioned position calibration data includes longitude and latitude.

[0058] In block 220, according to the current distribution status of automatic weather stations, a total of 1523 stations such as coastal areas, islands, oil platforms, and moored buoys can be selected for hourly monitoring data for relevant research. Based on the fact that the automatic weather stations are mainly located in coastal areas and are closely distributed, combined with the characteristics of meteorological distribution, stations with similar longitude and latitude are similar, and their air pressure, wind speed, temperature, and precipitation should have consistency, correlation, and time lag, but may be affected by the local environment. Therefore, the k-means clustering algorithm can be used to classify the automatic weather stations, that is, the stations within the target station area, according to the distance.

[0059] In some embodiments, the above-mentioned grouping of the stations within the target station area according to the meteorological monitoring data and the position calibration data specifically includes:

[0060] Based on the k-means clustering algorithm, the stations within the target station area are grouped according to the meteorological monitoring data and the position calibration data.

[0061] In some embodiments, the dataset, i.e., the meteorological monitoring data and location calibration data of the stations within the target site area, can be divided into K clusters based on the K-means clustering algorithm, so that the data points within each cluster have a relatively high similarity, while the data points between different clusters have a relatively low similarity. Among them, the similarity can be measured by a distance metric. For example, the Euclidean distance can be used as the measure of similarity. Considering the number and distribution characteristics of the stations within the target site area, the number of clusters of K-means clustering is taken as k = 150, and the stations within the target site area are grouped into 150 groups, and the grouping results are as Figure 3 shown. Judging from the grouping results, the number of stations in most groups is about 10 - 40, and the number of automatic weather stations in a small number of groups is small, mostly scattered automatic weather stations, which is consistent with the site distribution characteristics and observation requirements.

[0062] For example, next, the station group (22, 113) can be selected as a typical station group for the introduction of the calculation process. As shown in Table 1, this station group includes 31 automatic weather stations, which are relatively concentrated in the area near 22°N and 113°E.

[0063] Table 1: Information of Typical Grouped Stations

[0064] Station Number Location Area Station Number Location Area Station Number Location Area G1142 Area A G3531 Area C G3634 Area D G1155 Area B G3540 Area A G3641 Area A G1179 Area A G3546 Area A G3643 Area C G1189 Area A G3548 Area A G3645 Area D G1191 Area A G3551 Area B G3680 Area C G1192 Area C G3553 Area B G3704 Area C G1195 Area A G3555 Area A G3759 Area C G3501 Area D G3556 Area A G3768 Area C G3502 Area C G3561 Area A G3782 Area A G3521 Area A G3562 Area D G3530 Area C G3585 Area A

[0065] In block 230, the preset performance health model can be set according to the actual needs of the user.

[0066] In some embodiments, the performance health model of the marine automatic weather station in terms of performance dimensions includes two elements: parameter validity and parameter integrity. For the purpose of identifying whether the sensor and data transmission device have normal acquisition and transmission functions, the performance evaluation is carried out by analyzing the validity and integrity of the seven types of monitoring parameters of air pressure, temperature, humidity, precipitation, wind speed, wind direction, and visibility collected.

[0067] In some embodiments, the construction of the above-mentioned preset performance health model includes:

[0068] Calculate the corresponding parameter validity according to the number of valid values in the meteorological monitoring data of the stations within each group;

[0069] Calculate the corresponding parameter integrity according to the number of actual collected values and the theoretical collection number in the meteorological monitoring data of the stations within each group;

[0070] Based on the parameter validity and parameter integrity, construct a preset performance health model.

[0071] In some embodiments, according to the technical indicators of the observation equipment, the effective numerical ranges of various sensors can be sorted out, as shown in Table 2.

[0072] Table 2: Effective Numerical Range Measured by Automatic Weather Station

[0073] Measurement Field Measurement Type Unit Lower Limit Upper Limit Air Pressure Air Pressure hPa 500 1100 Temperature Air Temperature ℃ -50 50 Relative Humidity Humidity %RH 5 100 Precipitation in the Past 1 Hour Precipitation mm 0 - Ten-Minute Average Wind Direction Wind ° 0 360 Ten-Minute Average Wind Speed Wind m / s 0 60 1-Hour Average Visibility Visibility m 10 50000

[0074] In some embodiments, it is possible to identify whether various sensors are abnormal according to the parameter range, and count the number of valid values of each monitoring parameter within a time range. The calculation formula for the data validity of a single monitoring parameter is as follows:

[0075]

[0076] Among them, x v represents the parameter validity, n 有效 represents the number of valid values of the parameter, and n 总 represents the total number of the parameter.

[0077] For example, taking the automatic weather station No. G1195 as an example, the evaluation result of its data validity is as Figure 4 shown. The data validities of meteorological monitoring parameters such as wind speed, wind direction, air pressure, temperature, and humidity are all relatively high, while the data validity of visibility is basically zero, indicating that there is no visibility observation at this station.

[0078] In some embodiments, according to the technical indicators of the observation equipment, the theoretical acquisition frequency of each monitoring parameter is 1 hour. Therefore, the ratio of the number of actual acquisition values of the monitoring parameter within the time range to the theoretical acquisition number can be calculated as the parameter integrity. The calculation formula for the data integrity of a single monitoring parameter is as follows:

[0079]

[0080] Among them, x i represents the parameter integrity, n 总 represents the number of actual acquisition values of the parameter, and n 理论 represents the theoretical acquisition number of the parameter.

[0081] For example, taking the automatic weather station No. G1195 as an example, the evaluation result of its data integrity is as Figure 5 shown. The data integrity of each parameter is at the same level in the daily statistical results. If data is missing, all monitoring values are lost simultaneously. Therefore, all data missing can be considered as an abnormality in the data transmission module or the power supply system.

[0082] In some embodiments, based on the parameter validity and the parameter integrity, it is possible to establish a performance health index for detecting sensors of the automatic weather station and complete the construction of the performance health model. For the seven monitoring parameters of air pressure, temperature, humidity, wind speed, wind direction, precipitation, and visibility, the performance health index h of each parameter is evaluated according to the following formula p .

[0083]

[0084] Among them, x is the parameter integrity or parameter validity index of a certain measurement value, and x ref is the lower limit of the distribution of the historical statistics of this index x, is the mean value of the historical statistics of this index, and σ x is the standard deviation of the historical statistics of this index. k is a penalty coefficient, and its value can be ln0.6, indicating that when the data integrity or validity is lower than the lower limit of the historical statistical distribution, the result of the performance health index is 0.6.

[0085] Based on the data integrity and validity evaluation indexes of each measurement value, the performance health index H of the automatic weather station is calculated using the following formula according to the weights in Table 3 p

[0086]

[0087] Among them, k i is the index weight of the i-th parameter, and h pi is the parameter performance health index of the i-th parameter.

[0088] Table 3: Weights of Automatic Weather Station Parameter Indexes

[0089]

[0090] For example, taking the G1195 automatic weather station as an example, analyzing its data from January to July 2023, it is identified that its performance health index drops to 0.58 on January 17, 2023, and drops to 0.66 on April 25, 2023, indicating that the performance of the automatic weather station equipment is abnormal during this period.

[0091] In some embodiments, the construction of the above-mentioned preset configuration health model includes:

[0092] Based on the principal component analysis algorithm, according to the meteorological monitoring data of each site in the group, calculate the consistency of the same type of parameters and the correlation of different types of parameters;

[0093] Based on the consistency of the same type of parameters and the correlation of different types of parameters, construct a preset configuration health model.

[0094] In some embodiments, for the automatic weather stations in the same region, their weather characteristics are theoretically roughly similar. Therefore, the same measurement parameters of different sites in the group should have the same trend. If the parameter trend of a certain site is inconsistent with that of other sites in the same group, it is considered that there may be an accuracy problem with this parameter of this site. Therefore, parameter consistency analysis is used as the main technology for the configuration health analysis of the automatic weather station.

[0095] In some embodiments, the parameter consistency analysis method is the Principal Component Analysis (PCA) algorithm. Since the PCA algorithm is a mainstream data dimensionality reduction technology, it can project high-dimensional data into a low-dimensional space by extracting the main change directions of the data, thereby extracting the main trend features of the same group of parameters. Therefore, the deviation analysis can be performed between each parameter in the group and the main trend features after back-projection to obtain the evaluation result of parameter consistency.

[0096] In some embodiments, the configuration health model of the marine automatic weather station is designed based on PCA, reducing the original data to k dimensions and the dimensions being independent of each other after dimensionality reduction. The value of k is calculated by the following formula. The information retention ratio r is taken as 0.9 to retain most of the information to reflect the main trend performance of the parameters. n T It is calculated by the following formula. The information retention ratio r is taken as 0.9 to retain most of the information to reflect the main trend performance of the parameters.

[0097]

[0098] In some embodiments, based on the data after dimensionality reduction, the dimension is restored by the inverse matrix of the transformation matrix and the data after dimensionality reduction. Since information is lost, only the maximum consistent information of all parameters is retained, thereby obtaining the theoretical trend line of each parameter in the group trend. After extracting the group parameter trend by PCA dimensionality reduction, the difference between the actual value of the parameter and the theoretical trend line is calculated to identify the change of each parameter measurement point outside the overall trend, so as to evaluate whether there is a trend deviation.

[0099] In some embodiments, the calculation rule of the configuration health index includes: within a certain time window, the ratio of the number of data points where the group of parameters exceed the trend deviation distribution to the total number of data points within this time window. The value range of this ratio is 0 to 1. The closer it is to 1, the higher the risk. Subtracting this ratio from 1 is the parameter consistency health index. Thus, the calculation formula for the configuration health index of this group is:

[0100]

[0101] ΔX = X - X'

[0102] where, H c is the configuration health index, n (ΔX>3σΔX) is the number of data points where the parameter exceeds the trend deviation of 3σ distribution, n 总 is the total number of all data points, X is the original feature matrix, and X' is the theoretical trend line matrix after dimension restoration.

[0103] ​In some embodiments, the configuration health of an automatic weather station is divided into two types: the consistency of similar parameters and the correlation of dissimilar parameters, corresponding to the relationships of the same type of parameters at different stations, such as the consistency of barometric pressure parameters at all stations, and the relationships of different types of parameters, such as the correlation between air temperature and barometric pressure. Among them, the correlation of dissimilar parameters is further divided into two working conditions: normal weather processes (no precipitation) and precipitation weather processes, with the zero 1-hour precipitation amount as the basis for dividing the working conditions.

[0104] In some embodiments, the configuration correlation features are respectively trained and the health index is calculated, and the configuration health index is obtained through weighted normalization calculation according to the weight of 0.7 for the consistency of similar parameters, the weight of 0.2 for the correlation of dissimilar parameters in normal weather processes, and the weight of 0.1 for precipitation weather processes, so as to analyze whether there are accuracy anomalies through the analysis of the variation rules among parameters.

[0105] For example, taking the station group (22, 113) where G1195 is located as an example, and still using the observation data from January to July 2023 as the analysis object, taking the air temperature parameter as an example to analyze its parameter consistency performance, such as Figure 6 shown, the air temperature parameters of the stations within the group have obvious trend consistency.

[0106] In some embodiments, adopting the above calculation process of the configuration health index, after extracting the parameter trend within the group through PCA dimensionality reduction, the difference between the actual value of the parameter and the theoretical trend line is calculated as the trend deviation value of the parameter, and its deviation value time series performance and distribution are as Figure 7 shown. The trend deviation values of parameters at most stations are distributed near zero and show a normal distribution, and the distribution interval threshold of 3σ is used to identify abnormal parameter trend deviations.

[0107] In some embodiments, the accuracy health index results of the air temperature parameters of each station in this area can be obtained by analyzing the proportion of trend deviation points, as Figure 8 shown. It can be identified that the configuration health index of the mangrove station was as low as 0.44 on January 29, 2023, and the configuration health index of the Nanshan station was as low as 0.61, and there were significant anomalies in the air temperature measuring instruments of the automatic weather stations.

[0108] In some embodiments, the consistency of similar parameters, the correlation of dissimilar parameters, etc. of the stations within the group (22, 113) are comprehensively calculated, and the configuration health index of the automatic weather stations within the group is calculated according to the aforementioned weights, such as Figure 9As shown. The configuration health indicators of most sites are between 0.8 and 0.9. For a few sites such as G1179, G191, G3643, G3704, and G3782, the configuration health indicators are less than 0.6, indicating that there are accuracy anomalies during this period. An automatic weather station anomaly warning is issued and the types of abnormal parameters are synchronized, so as to assist in remotely diagnosing and locating the types of faulty instruments.

[0109] In block 240, the determining of the abnormal sites according to the performance health indicators and / or configuration health indicators of the sites in each group specifically includes:

[0110] Determining the abnormal sites according to the performance health indicators of the sites in each group.

[0111] In some embodiments, the determining of the abnormal sites according to the performance health indicators and / or configuration health indicators of the sites in each group specifically further includes:

[0112] Determining the abnormal sites according to the configuration health indicators of the sites in each group.

[0113] In some embodiments, the determining of the abnormal sites according to the performance health indicators and / or configuration health indicators of the sites in each group specifically further includes:

[0114] Determining the sites with abnormal sensor measurement value validity and / or abnormal sensor measurement value integrity according to the performance health indicators of the sites in each group;

[0115] Determining the sites with accuracy anomalies from the sites excluding the sites with abnormal sensor measurement value validity and / or abnormal sensor measurement value integrity according to the configuration health indicators of the sites in each group.

[0116] In some embodiments, after the site grouping, some key dense site areas are extracted. With the condition that the number of sites in the same group area is not less than 20, 18 group areas with a total of 266 automatic weather station sites are selected as the analysis objects, and the results of the automatic weather station health monitoring algorithm model are evaluated for the automatic weather station data from January to July 2023. By comparing and analyzing the actual determination results and the true fault results evaluated daily, the dataset for performance health analysis has a total of 90710 samples. After filtering out the data anomaly samples (abnormal sensor acquisition values), the dataset for configuration health analysis has a total of 84624 samples. The automatic station health assessment results are summarized, and the confusion matrix is used to evaluate the performance health analysis and the configuration health analysis respectively. The analysis results are as Figure 10 shown.

[0117] In some embodiments, the results are judged in combination with operation and maintenance records. The fault identification accuracy rate of performance health analysis is 22735 / (22735 + 367) = 98.4%, and the false alarm rate is 1977 / (22735 + 1977) = 8.0%. The fault identification accuracy rate of configuration health analysis is 240 / (240 + 19) = 92.7%, and the false alarm rate is 38 / (240 + 38) = 13.7%. The algorithm models of performance health analysis and configuration health analysis have good abnormal identification capabilities for automatic weather stations and have good discrimination capabilities for such typical unbalanced classification problems.

[0118] According to the embodiments of the present disclosure, the following technical effects are achieved:

[0119] It is possible to obtain meteorological monitoring data and position calibration data within the target site area; then group the sites within the target site area according to the meteorological monitoring data and position calibration data; input the meteorological monitoring data of the sites within each group into a preset performance health model respectively, and correspondingly output performance health indicators; and / or input the meteorological monitoring data of the sites within each group into a preset configuration health model respectively, and correspondingly output configuration health indicators; then determine abnormal sites according to the performance health indicators and / or configuration health indicators of the sites within each group. Based on this, in combination with the rapidly developing data analysis and artificial intelligence technologies, through the constructed coastal automatic weather station health monitoring and abnormal identification model, calculate and evaluate the health indicators of each element data of the observation station, realize real-time monitoring of the operation health of coastal automatic weather station equipment, abnormal early warning and automatic identification of fault points, so as to provide support for improving the ability and efficiency of meteorological intelligent guarantee work.

[0120] It should be further noted that in order to further improve the operation status monitoring ability of automatic weather station equipment and the abnormal identification ability of observation data, and fully explore the potential value of historical observation data of automatic weather stations, the above method combines the rapidly developing data analysis and artificial intelligence technologies, through methods such as data correlation analysis of grouped sites, extraction of data trend deviation characteristics, data-driven modeling based on machine learning models, and abnormal data analysis after meteorological data quality control, etc., to research and establish a coastal automatic weather station health monitoring and abnormal identification model. This model can calculate and evaluate the health indicators of each element data of the observation station and finally realize equipment fault early warning. At the same time, a corresponding algorithm program package is developed based on the algorithm research and iteratively optimized through big data training to improve the accuracy rate and early warning lead time, realize real-time monitoring of the operation health of coastal automatic weather station equipment, abnormal early warning and automatic identification of fault points, so as to provide support for improving the ability and efficiency of meteorological intelligent guarantee work.

[0121] In some embodiments, the above method further includes:

[0122] Send early warning information for abnormal sites, where the early warning information for abnormal sites includes the station numbers of abnormal sites and types of abnormal parameters.

[0123] In some embodiments, the types of abnormal parameters include anomalies in the validity of sensor measurement values, anomalies in the integrity of sensor measurement values, and anomalies in the accuracy of sensor measurement values.

[0124] In some embodiments, in combination with the parameter characteristics and analysis dimensions of the monitoring data of ocean automatic weather stations, it is also possible to design Figure 11 the monitoring algorithm framework of the ocean automatic weather station as shown.

[0125] Specifically, in combination with the observation objects, data types, element correlations, etc. of the observation data, analyze its health status and detect anomalies from the two dimensions of performance and configuration. Performance analysis mainly focuses on the validity and integrity of the sensor measurement values of the automatic weather station equipment, and evaluates whether the basic measurement functions of the sensors and the data transmission module are working properly. Configuration analysis, on the other hand, analyzes the correlation and performance between parameters by mining the characteristics of the monitoring data of the automatic weather station to judge the accuracy of the sensor measurement values. After completing the performance analysis and configuration analysis, integrate and calculate the performance health indicators and configuration health indicators of the automatic weather station, and issue corresponding anomaly warnings according to the performance of the health indicators.

[0126] In summary, relying on the historical observation data of ocean automatic weather stations, it is possible to construct the framework and model of the automatic weather station anomaly monitoring algorithm model. Through in-depth data mining, feature construction, deviation analysis, etc., data-driven monitoring and anomaly monitoring modeling based on machine learning are realized. On the basis of meteorological data quality control, the established health index model of the ocean automatic weather station comprehensively evaluates the health of each element. Through multi-dimensional evaluation and integrated analysis, equipment health monitoring and anomaly warning of the ocean automatic weather station are realized.

[0127] Verification through historical data testing shows that the algorithm model has a high fault recognition rate and a low false alarm rate, has a high fault recognition rate and a low false alarm rate, and already has the ability to meet the actual business requirements. At the same time, the above algorithm model can not only be used to support the operation monitoring and meteorological guarantee of ocean automatic weather station equipment, but also provide useful references for the health monitoring and intelligent operation and maintenance technology development of other meteorological observation equipment.

[0128] More notably, due to difficulties such as unattended operation and insufficient monitoring means, marine automatic weather stations have long faced problems of lagging alarm monitoring and status monitoring. To improve the health monitoring and anomaly recognition capabilities of automatic weather stations, the above method uses artificial intelligence technologies such as machine learning and multi-dimensional configuration, combines with the working characteristics of automatic weather stations, constructs an anomaly monitoring algorithm model for automatic weather stations, and realizes early anomaly detection and auxiliary rapid diagnosis of automatic weather stations through big data feature training, deviation intelligent recognition, and health index construction. Tests with actual data show that the comprehensive accuracy rate of the model is 95.2%, and the false alarm rate is less than 10%, meeting the requirements of business applications and providing support for the intelligent guarantee of marine automatic weather stations.

[0129] It should be noted that for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the present disclosure is not limited by the described action sequence, because according to the present disclosure, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to the present disclosure.

[0130] The above is the introduction of method embodiments. The following further illustrates the solution of the present disclosure through device embodiments.

[0131] Figure 12 The block diagram of the marine automatic weather station anomaly monitoring device 1200 based on multi-dimensional configuration features according to an embodiment of the present disclosure is shown. As Figure 12 shown, the device 1200 includes:

[0132] An acquisition module 1210, configured to acquire meteorological monitoring data and position calibration data within the target site area;

[0133] A grouping module 1220, configured to group the sites within the target site area according to the meteorological monitoring data and position calibration data;

[0134] A generation module 1230, configured to respectively input the meteorological monitoring data of the sites within each group into a preset performance health model, and correspondingly output performance health indicators; and / or respectively input the meteorological monitoring data of the sites within each group into a preset configuration health model, and correspondingly output configuration health indicators;

[0135] A determination module 1240, configured to determine abnormal sites according to the performance health indicators and / or configuration health indicators of the sites within each group.

[0136] Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the described modules can refer to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0137] In the technical solution of the present disclosure, the acquisition, storage, and application of the user's personal information involved all comply with the provisions of relevant laws and regulations and do not violate public order and good customs.

[0138] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0139] Figure 13 The block diagram of an exemplary electronic device 1300 capable of implementing the embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0140] The electronic device 1300 includes a computing unit 1301, which can perform various appropriate actions and processes according to the computer program stored in the ROM 1302 or the computer program loaded from the storage unit 1308 into the RAM 1303. In the RAM 1303, various programs and data required for the operation of the electronic device 1300 can also be stored. The computing unit 1301, the ROM 1302, and the RAM 1303 are connected to each other through a bus 1304. The I / O interface 1305 is also connected to the bus 1304.

[0141] A plurality of components in the electronic device 1300 are connected to the I / O interface 1305, including: an input unit 1306, such as a keyboard, a mouse, etc.; an output unit 1307, such as various types of displays, speakers, etc.; a storage unit 1308, such as a magnetic disk, an optical disc, etc.; and a communication unit 1309, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 1309 allows the electronic device 1300 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0142] The computing unit 1301 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 1301 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 1301 executes the various methods and processes described above, such as method 200. For example, in some embodiments, method 200 can be implemented as a computer software program that is tangibly contained in a machine-readable medium, such as the storage unit 1308.

[0143] In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 1300 via the ROM 1302 and / or the communication unit 1309. When the computer program is loaded into the RAM 1303 and executed by the computing unit 1301, one or more steps of the method 200 described above can be executed. Alternatively, in other embodiments, the computing unit 1301 can be configured to execute method 200 in any other suitable manner (e.g., by means of firmware).

[0144] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuitry, integrated circuit systems, field-programmable gate arrays (FPGA), application-specific integrated circuits (ASIC), application-specific standard products (ASSP), system-on-a-chip systems (SOC), complex programmable logic devices (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a dedicated or general-purpose programmable processor, that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0145] The program code for implementing the methods of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the program codes are executed by the processor or controller, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0146] In the context of this disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0147] In order to provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, speech input, or tactile input).

[0148] The systems and techniques described herein can be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0149] A computer system can include a client and a server. The client and the server are generally remote from each other and typically interact through a communication network. The relationship of the client and the server is generated by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, or a server of a distributed system, or a server incorporating a blockchain.

[0150] It should be understood that the various forms of processes shown above can be used, with steps reordered, added or deleted. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and no limitations are imposed herein.

[0151] The above specific embodiments do not constitute a limitation on the protection scope of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the protection scope of this disclosure.

Claims

1. An abnormal monitoring method for ocean automatic weather stations based on multi-dimensional configuration characteristics, characterized in that Including: Obtain meteorological monitoring data and position calibration data within the target site area; Group the sites within the target site area according to the meteorological monitoring data and the position calibration data; Input the meteorological monitoring data of the sites within each group into a preset performance health model respectively, and correspondingly output performance health indicators; and / or input the meteorological monitoring data of the sites within each group into a preset configuration health model respectively, and correspondingly output configuration health indicators; Determine abnormal sites according to the performance health indicators and / or configuration health indicators of the sites within each group.

2. The method according to claim 1, characterized in that, The grouping of the sites within the target site area according to the meteorological monitoring data and the position calibration data includes: Based on the k-means clustering algorithm, group the sites within the target site area according to the meteorological monitoring data and the position calibration data.

3. The method according to claim 1, characterized in that, The construction of the preset performance health model includes: Calculate the corresponding parameter validity according to the number of valid values in the meteorological monitoring data of the sites within each group; Calculate the corresponding parameter integrity according to the number of actual collected values and the theoretical collection number in the meteorological monitoring data of the sites within each group; Based on the parameter validity and the parameter integrity, construct a preset performance health model.

4. The method according to claim 1, wherein The construction of the preset configuration health model includes: Based on the principal component analysis algorithm, calculate the consistency of similar parameters and the correlation of dissimilar parameters according to the meteorological monitoring data of the sites within each group; Based on the consistency of similar parameters and the correlation of dissimilar parameters, construct a preset configuration health model.

5. The method according to claim 1, wherein The determination of abnormal sites according to the performance health indicators and / or configuration health indicators of the sites within each group includes: Determine the sites with abnormal sensor measurement value validity and / or abnormal sensor measurement value integrity according to the performance health indicators of the sites within each group; According to the configuration health indicators of the sites within each group, determine the sites with abnormal accuracy from the sites excluding the abnormal sensor measurement value validity and / or abnormal sensor measurement value integrity.

6. The method according to any one of claims 1 to 5, characterized in that, The meteorological monitoring data includes air pressure, temperature, humidity, wind speed, wind direction, precipitation, and visibility; the position calibration data includes longitude and latitude.

7. The method according to any one of claims 1 to 5, characterized in that, The method further includes: Send an early warning message for abnormal sites, and the early warning message for abnormal sites includes the station number of the abnormal site and the type of abnormal parameter.

8. An abnormal monitoring device for an ocean automatic weather station based on multi-dimensional configuration characteristics, characterized in that, Including: An acquisition module for obtaining meteorological monitoring data and position calibration data within the target site area; A grouping module for grouping the sites within the target site area according to the meteorological monitoring data and the position calibration data; A generation module for inputting the meteorological monitoring data of the sites within each group into a preset performance health model respectively, and correspondingly outputting performance health indicators; and / or inputting the meteorological monitoring data of the sites within each group into a preset configuration health model respectively, and correspondingly outputting configuration health indicators; A determination module for determining abnormal sites according to the performance health indicators and / or configuration health indicators of the sites within each group.

9. An electronic device, characterized in that, Including: At least one processor; And A memory communicatively connected to the at least one processor; The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method according to any one of claims 1-7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are for causing the computer to execute the method according to any one of claims 1-7.

Images