CN113704558A - Automatic meteorological station automatic clustering method and system based on differential sequence - Google Patents
Automatic meteorological station automatic clustering method and system based on differential sequence Download PDFInfo
- Publication number
- CN113704558A CN113704558A CN202110982065.0A CN202110982065A CN113704558A CN 113704558 A CN113704558 A CN 113704558A CN 202110982065 A CN202110982065 A CN 202110982065A CN 113704558 A CN113704558 A CN 113704558A
- Authority
- CN
- China
- Prior art keywords
- differential sequence
- characteristic quantity
- automatic
- stations
- station
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides an automatic meteorological station automatic clustering method and system based on high comparability of a differential sequence, and belongs to the technical field of meteorology. According to the method, an atmospheric element value difference sequence between stations is established according to the near-stratum atmospheric energy change rule of the automatic meteorological station, the signal characteristic quantity of the difference sequence is extracted by adopting a digital signal analysis and processing method, and an automatic clustering algorithm of the automatic meteorological station with high underlying surface similarity is established. The invention utilizes the characteristic of high comparability of meteorological element values in a station group to establish a meteorological element value balance system, calculates stations with data drift caused by artificial change or seasonal change of an underlying surface according to the balance system, establishes a more precise meteorological data quality control algorithm through the balance system, and provides effective ground data for fusion analysis of multi-source meteorological data, air-ground integrated meteorological information and the like by utilizing the comparability.
Description
Technical Field
The invention belongs to the technical field of weather, and particularly relates to an automatic weather station clustering method based on a differential sequence.
Background
Under the vigorous development of meteorological observation services, the grid layout of automatic meteorological stations in China is completely built, and a large amount of meteorological data are accumulated for 10 years. In the process of vigorously promoting the modernization of meteorological socialization services, the function and the status of the data of the automatic meteorological station are continuously improved and strengthened. However, in the actual operation and daily management process, due to a plurality of factors such as remote positions, site dispersion, human and technical capabilities and the like, the availability, reliability and continuity of a large amount of automatic weather station data are not guaranteed. In a plurality of meteorological data applications such as gridding weather forecast, climate analysis, environment monitoring, refined meteorological service and scientific research, the data function of the automatic meteorological station is not played to the maximum. The reason for this is mainly as follows:
1. the management standards and standards of the automatic weather station and the national ground observation station are greatly different, the quality standard of the atmospheric data of the automatic weather station is reduced, and the data application range is limited.
2. Due to changes in the underlying surface of the site, the resulting data drift is not recognizable. The characteristics of the networked automatic weather station similar to the cushion surface have no full play in the quality control from space to property.
Disclosure of Invention
In view of the above-mentioned deficiencies in the prior art, the present invention provides a method and a system for automatic meteorological station automatic clustering based on differential sequence, so as to solve the above-mentioned problems.
In order to achieve the above purpose, the invention adopts the technical scheme that:
the scheme provides an automatic meteorological station automatic clustering method based on a differential sequence, which comprises the following steps:
s1, forming a first differential sequence by meteorological data of all national ground observation stations within 100 kilometers at the lowest temperature moment every day, and extracting signal characteristic quantity of the first differential sequence;
s2, respectively taking the maximum value and the minimum value of the first differential sequence signal characteristic quantity, taking the minimum value as a first cluster reference characteristic quantity, and taking the intermediate value between the minimum value and the maximum value as a second cluster reference characteristic quantity;
s3, judging whether the selection of all national ground observation stations within 100 kilometers is finished, if so, entering a step S6, otherwise, entering a step S4;
s4, selecting an automatic weather station which is 50 kilometers away from the national weather station, establishing a second differential sequence, and extracting signal characteristic quantity of the second differential sequence;
s5, collecting all automatic weather stations with the second difference sequence characteristic quantity smaller than the first cluster reference characteristic quantity to form a first site cluster, and collecting all automatic weather stations with the second difference sequence characteristic quantity larger than or equal to the first cluster reference characteristic quantity to form a second site cluster to finish automatic cluster processing of the automatic weather stations;
s6, judging whether the automatic weather station is selected in the selected area, if so, finishing the automatic clustering processing of the automatic weather station, otherwise, entering the step S7;
s7, selecting an automatic weather station which is 30 kilometers away from the national weather station, establishing a third differential sequence, and extracting signal characteristic quantity of the third differential sequence;
and S8, collecting all the automatic weather stations with the third differential sequence signal characteristic quantity more than or equal to the second cluster reference characteristic quantity to form a third station cluster, and finishing the automatic clustering processing of the automatic weather stations.
The invention has the beneficial effects that: the invention judges the stations with high underlay surface similarity and weather data comparability among the stations, establishes a differential data analysis sequence of the related atmospheric elements among the stations according to the near-stratum atmospheric energy change principle of the automatic weather station, and forms an analysis data sequence with 90 lengths by taking seasons as units. Therefore, the data sequence meets the normal distribution characteristic, the sequence signal characteristic quantity is calculated, the similarity of the underlying surfaces of the two stations is high and the comparability of meteorological data of the two stations is high according to the fact that the larger the characteristic quantity is, and the two stations are automatically gathered together. In the same cluster, a meteorological element value balance system among stations is established, the artificial change of the underlying surface or the data drift caused by a sensor is monitored, and a more refined quality control algorithm of the data of the respective dynamic meteorological stations in the cluster is established through the meteorological element value balance system among the stations.
Further, the processes of extracting the first differential sequence signal feature quantity, extracting the second differential sequence signal feature quantity and extracting the third differential sequence signal feature quantity are the same, and each of the processes includes the following steps:
a1, forming a differential sequence by meteorological data at the lowest temperature time every day between any two national ground observation stations or automatic meteorological stations in a fixed area, wherein the differential sequence is a first differential sequence, a second differential sequence or a third differential sequence;
a2, judging whether the differential sequence meets normal distribution, if so, entering the step A3, otherwise, outputting an error prompt, and ending the process;
and A3, calculating the interval probability of the differential sequence, and calculating to obtain a differential sequence signal characteristic quantity according to the interval probability, wherein the differential sequence signal characteristic quantity is a first differential sequence signal characteristic quantity, a second differential sequence signal characteristic quantity or a third differential sequence signal characteristic quantity.
The beneficial effects of the further scheme are as follows: according to the method, the difference is made between the corresponding atmospheric element values at the same time point between two stations to form a differential sequence signal, the similarity degree of changes of the two stations is calculated in an error range by calculating normal distribution, and if a similarity program is larger, the relevance of the two stations is larger, and the comparability is stronger.
Further, the step a1 includes the following steps:
a101, selecting meteorological observation element values of any two national ground observation stations or automatic meteorological stations at the lowest temperature in a fixed area according to the atmospheric energy change rule of the near-stratum of the national ground observation stations or the automatic meteorological stations, and determining the correlation between the near-stratum atmospheric energy and the meteorological observation element values;
and A102, calculating the difference value of every two meteorological observation element values at the same time point between two national ground observation stations or automatic meteorological stations according to the association, and forming a differential sequence by taking a quarter as a time length, wherein the differential sequence is a first differential sequence, a second differential sequence or a third differential sequence.
The beneficial effects of the further scheme are as follows: according to the method, an atmospheric element value differential sequence between stations is established according to the near-stratum atmospheric energy change rule of the automatic meteorological station, and the signal characteristics of the differential sequence are extracted by adopting a digital signal analysis and processing method so as to establish an automatic clustering algorithm of the automatic meteorological station with high underlying surface similarity.
Still further, the expression of the difference sequence in step a102 is as follows:
wherein, Σ Δ ηiRepresenting a differential sequence, said differential sequence being a first, a second or a third differential sequence, gammaiRepresents beta, k and RiNormalized parameter, RiRepresenting the number of sunshine hours per day, beta and k both representing constants, Hi1Represents the humidity value T of the ground observation station or automatic meteorological station of a certain country at the lowest temperature moment every daymin(i1)Represents the daily minimum temperature value H of a ground observation station or an automatic meteorological station of a certain countryi2Representing the humidity value, T, of the lowest temperature moment of day of a ground observation station or an automated weather station of another countrymin(i2)Representing the lowest temperature value of each day of another national ground observation station or automated weather station.
The beneficial effects of the further scheme are as follows: according to the invention, the meteorological element values at the lowest temperature moment every day between two automatic meteorological stations are taken as a difference, and the difference sequence reduces the influence on data caused by the change of the underlying surface environment due to long time, so that the characteristic quantity extracted by the difference sequence signal is more obvious and stable.
Still further, the expression of the differential sequence signal characteristic quantity in the step a3 is as follows:
wherein col represents a differential sequence signal characteristic quantity, the differential sequence signal characteristic quantity is a first differential sequence signal characteristic quantity, a second differential sequence signal characteristic quantity or a third differential sequence signal characteristic quantity, fΔηExpressing a normal distribution function, e expressing an observation error range, delta eta, the near-to-earth atmospheric energy difference of two automatic meteorological stations at the lowest temperature moment, dΔηRepresents the integral variable of Δ η, σ represents the variance, and a is the mean of the two-station difference data sequence.
The beneficial effects of the further scheme are as follows: the comparability is judged by calculating the change similarity, and the bigger the change similarity is, the bigger the relevance of the two stations is, and the stronger the comparability is.
The invention also provides an automatic meteorological station automatic clustering system based on the differential sequence, which comprises the following components:
the first differential sequence signal characteristic quantity extraction module is used for forming a first differential sequence by meteorological data of all national ground observation stations within 100 kilometers at the lowest temperature moment every day and extracting a first differential sequence signal characteristic quantity;
the cluster reference characteristic quantity confirmation module is used for respectively taking the maximum value and the minimum value of the first differential sequence signal characteristic quantity, taking the minimum value as a first cluster reference characteristic quantity, and taking the middle value between the minimum value and the maximum value as a second cluster reference characteristic quantity;
the first judgment module is used for judging whether the selection of all national ground observation stations within 100 kilometers is finished;
the second differential sequence signal characteristic quantity extraction module is used for selecting an automatic weather station which is 50 kilometers away from the national weather station, establishing a second differential sequence and extracting a second differential sequence signal characteristic quantity;
the first classification module is used for collecting all automatic weather stations with second difference sequence characteristic quantity smaller than the first cluster reference characteristic quantity to form a first site cluster, and collecting all automatic weather stations with second difference sequence characteristic quantity larger than or equal to the first cluster reference characteristic quantity to form a second site cluster to finish automatic cluster processing of the automatic weather stations;
the second judgment module is used for judging whether the automatic weather station in the selected area is selected completely;
the third differential sequence signal characteristic quantity extraction module is used for selecting an automatic weather station which is 30 kilometers away from the national weather station, establishing a third differential sequence and extracting a third differential sequence signal characteristic quantity;
and the second classification module is used for collecting all the automatic weather stations with the third differential sequence signal characteristic quantity greater than or equal to the second cluster reference characteristic quantity to form a third station cluster and finish the automatic clustering processing of the automatic weather stations.
The invention has the beneficial effects that: the invention takes the characteristic quantity between national ground observation stations as the basis of grouping, because of the distance, the characteristic quantity between the large stations has the maximum value and the minimum value, the minimum value is taken as the reference quantity of the component, each national ground observation station conducts a differential data analysis sequence on the automatic stations in the preset range, the characteristic quantity is extracted, the stations which are larger than the reference quantity form a station cluster, the stations which are smaller than the reference quantity form a station cluster, the stations which are divided into the same cluster have similar underlying surface properties, and the station-to-point meteorological data has high comparability. The invention utilizes the characteristic of high comparability of meteorological element values in a station set to provide effective ground station support for fusion analysis of ground or air-ground integrated multi-source meteorological data, establishes a meteorological element value balance system among stations in the same cluster, and monitors artificial change of an underlying surface or data drift caused by a sensor. And establishing a more refined quality control algorithm of the data of the respective dynamic meteorological stations in the cluster through a meteorological element value balance system among the stations.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
FIG. 2 is a schematic diagram of the system of the present invention.
Detailed Description
The following description of the embodiments of the present invention is provided to facilitate the understanding of the present invention by those skilled in the art, but it should be understood that the present invention is not limited to the scope of the embodiments, and it will be apparent to those skilled in the art that various changes may be made without departing from the spirit and scope of the invention as defined and defined in the appended claims, and all matters produced by the invention using the inventive concept are protected.
Example 1
The atmospheric element values observed by the automatic meteorological station are mainly near-stratum atmospheric data and are reflected by a near-stratum atmospheric energy change rule, near-stratum atmospheric energy changes periodically along with sunrise and sunset, the magnitude of the near-stratum atmospheric energy is determined by energy balance conditions, the atmospheric energy approaches to a balance point at any moment according to an energy conservation principle, and the atmospheric meteorological element values tend to be stable at the balance point. In a day cycle, two balance points with most obvious atmospheric energy are the highest and the lowest near-formation atmospheric temperature, at the lowest temperature, the near-formation atmospheric energy is the lowest, the energy exchange of adjacent air masses is the least, the energy radiation of the underlying surface is the least, at the moment, the near-formation atmosphere of a station is the least interfered, the acquired atmospheric element value is stable, the accuracy of data is mainly determined by a sensor and an acquisition device processing algorithm, and the element value is directly related to the underlying surface characteristic.
The environment of the underlying surface of the automatic meteorological station changes with seasons, the near-stratum atmospheric energy also changes correspondingly, but the change of the similar underlying surface environment in the automatic meteorological station within a certain range has great tendency. The lowest energy difference between the automatic meteorological stations also tends to be constant, and the constancy of the difference is also reflected in the difference between the stations corresponding to the atmospheric element values. Therefore, a differential signal sequence is formed by adopting the corresponding meteorological element difference value between the stations at the time of the lowest energy of the automatic meteorological stations every day, the differential signal is analyzed and researched, the differential signal characteristics are found, the similarity of underlying surfaces of the stations is identified, an automatic station clustering algorithm is established, and a station set with high data comparability is formed.
Therefore, the invention discloses an automatic meteorological station clustering method based on a differential sequence, which forms a station set with strong data comparability according to an automatic meteorological station clustering algorithm, establishes a meteorological element value dynamic balance system adaptive to the change of underlying surface environment in the set, and provides support for application of meteorological element quality control, regional weather analysis, grid point meteorological prediction, meteorological big data service and the like. The data comparability is high, the gas element value changes little, and the change rules are similar. According to the method, the correlation atmospheric element differential sequence between the stations is established according to the stations with high similarity of the underlying surface and high weather data comparability and the near-stratum atmospheric energy change principle of the automatic weather station. The difference sequence takes seasons as units to form difference sequences with the lengths of 90, so that the difference sequences meet normal distribution characteristics, the signal characteristic quantity of the difference sequences is calculated, and according to the fact that the larger the characteristic quantity is, the higher the similarity of underlying surfaces of two stations is, the higher the comparability of meteorological data of the two stations is, the two stations are automatically gathered together. Meanwhile, the characteristic of high comparability of meteorological element values in a station set is utilized to provide effective ground station support for fusion analysis of ground or air-ground integrated multi-source meteorological data, an inter-station meteorological element value balance system is established in the same cluster, the artificial change of an underlying surface or data drift caused by a sensor is monitored, and a more precise quality control algorithm of the data of each dynamic meteorological station in the cluster is established through the inter-station meteorological element value balance system.
In view of the above, as shown in fig. 1, the present invention provides an automatic weather station clustering method based on differential sequence, which is implemented as follows:
s1, forming a first differential sequence by meteorological data of all national ground observation stations within 100 kilometers at the lowest temperature moment every day, and extracting signal characteristic quantity of the first differential sequence;
s2, respectively taking the maximum value and the minimum value of the first differential sequence signal characteristic quantity, taking the minimum value as a first cluster reference characteristic quantity, and taking the intermediate value between the minimum value and the maximum value as a second cluster reference characteristic quantity;
s3, judging whether the selection of all national ground observation stations within 100 kilometers is finished, if so, entering a step S6, otherwise, entering a step S4;
s4, selecting an automatic weather station which is 50 kilometers away from the national weather station, establishing a second differential sequence, and extracting signal characteristic quantity of the second differential sequence;
s5, collecting all automatic weather stations with the second difference sequence characteristic quantity smaller than the first cluster reference characteristic quantity to form a first site cluster, and collecting all automatic weather stations with the second difference sequence characteristic quantity larger than or equal to the first cluster reference characteristic quantity to form a second site cluster to finish automatic cluster processing of the automatic weather stations;
s6, judging whether the automatic weather station is selected in the selected area, if so, finishing the automatic clustering processing of the automatic weather station, otherwise, entering the step S7;
s7, selecting an automatic weather station which is 30 kilometers away from the national weather station, establishing a third differential sequence, and extracting signal characteristic quantity of the third differential sequence;
and S8, collecting all the automatic weather stations with the third differential sequence signal characteristic quantity more than or equal to the second cluster reference characteristic quantity to form a third station cluster, and finishing the automatic clustering processing of the automatic weather stations.
In this embodiment, the processes of extracting the first differential sequence signal feature quantity, extracting the second differential sequence signal feature quantity, and extracting the third differential sequence signal feature quantity are the same, and each of the processes includes the following steps:
a1, forming a differential sequence by meteorological data at the lowest temperature time of each day between any two national ground observation stations or automatic meteorological stations in a fixed area, wherein the differential sequence is a first differential sequence, a second differential sequence or a third differential sequence, and the implementation method comprises the following steps:
a101, selecting meteorological observation element values of any two national ground observation stations or automatic meteorological stations at the lowest temperature in a fixed area according to the atmospheric energy change rule of the near-stratum of the national ground observation stations or the automatic meteorological stations, and determining the correlation between the near-stratum atmospheric energy and the meteorological observation element values;
a102, calculating difference values of weather observation element values at the same time point between two national ground observation stations or automatic weather stations according to the correlation, and forming a differential sequence by taking a quarter as a time length, wherein the differential sequence is a first differential sequence, a second differential sequence or a third differential sequence; the expression of the difference sequence is as follows:
wherein, Σ Δ ηiRepresenting a differential sequence, said differential sequence being a first, a second or a third differential sequence, gammaiRepresents beta, k and RiNormalized parameter, RiRepresenting the number of sunshine hours per day, beta and k both representing constants, Hi1Represents the humidity value T of the ground observation station or automatic meteorological station of a certain country at the lowest temperature moment every daymin(i1)Represents the daily minimum temperature value H of a ground observation station or an automatic meteorological station of a certain countryi2Representing the humidity value, T, of the lowest temperature moment of day of a ground observation station or an automated weather station of another countrymin(i2)Representing the lowest temperature value of each day of a ground observation station or an automatic weather station of another country
A2, judging whether the differential sequence meets normal distribution, if so, entering the step A3, otherwise, outputting an error prompt, and ending the process;
a3, calculating the interval probability of the differential sequence, and calculating to obtain a differential sequence signal characteristic quantity according to the interval probability, wherein the differential sequence signal characteristic quantity is a first differential sequence signal characteristic quantity, a second differential sequence signal characteristic quantity or a third differential sequence signal characteristic quantity, and the expression of the differential sequence signal characteristic quantity is as follows:
wherein col represents a differential sequence signal characteristic quantity, the differential sequence signal characteristic quantity is a first differential sequence signal characteristic quantity, a second differential sequence signal characteristic quantity or a third differential sequence signal characteristic quantity, fΔηExpressing a normal distribution function, e expressing an observation error range, delta eta, the near-to-earth atmospheric energy difference of two automatic meteorological stations at the lowest temperature moment, dΔηRepresenting Δ ηThe integral variable, σ, represents the variance, a two-station difference data sequence mean.
In this embodiment, the near-ground layer is a main site where material and energy exchange is performed between land and the atmosphere, and these exchanges directly affect the atmospheric motion and change the environmental conditions of the earth surface, thereby affecting the atmospheric circulation and weather climate change, and the periodic change of the near-ground layer atmospheric energy is also a main reason for the continuous change of the atmospheric element values collected by the automatic weather station. The main climate factors indicating the change of the near-stratum atmospheric energy are temperature, humidity, wind, rain, air pressure and the like, and the most core climate factors are temperature, humidity, sunlight, radiation and the like. For simple and convenient analysis, the law of the change of the near-stratum atmospheric energy is analyzed through the change of the temperature, the humidity and the sunshine of the climate element value, and the relationship between the near-stratum atmospheric energy U and the climate element value is shown as the following formula.
U∝(H、T、R、P)
The above equation indicates that atmospheric energy may be indicated by a correlation of H, T, R values. Wherein U is near-formation atmospheric energy, H is atmospheric humidity, T is atmospheric temperature, R is sunlight, and P is near-earth atmospheric pressure. In a conventional automatic weather station, no radiation signal is generally collected, and sunshine data of a national ground observation station is used for approximately representing the total energy acquired in the atmosphere within a certain area.
The amount of energy of the open air mass cannot be calculated at any time, but when the atmospheric factor value changes, the amount of energy change per volume of the air mass can be calculated. In any time period, the energy change of the near-earth air mass of the automatic meteorological station is reflected on the change of atmospheric factors such as temperature, humidity and air pressure, but the speed of the energy change is directly related to the underlying surface of the station. Therefore, the change rate of the near-earth air mass energy within a certain time can be used for reflecting the relation between the change of meteorological element values of meteorological stations in different automatic areas and the underlying surface. The energy change rate is shown as the following formula:
wherein, eta represents the rate of change of energy,λ represents a temperature energy conversion ratio, Δ T is a variation of temperature over a certain period of time, CψIs the wet heat capacity of the temperature, which is a function of the temperature h, CψAnd H is 1.01+1.88H, wherein H is the absolute temperature of air, the atmospheric humidity element collected greatly is the relative humidity, and the absolute humidity (H) corresponding to different temperatures and different relative humidities (H) can be obtained through a humidity diagram of the wet air. In practical cases, the effect of the moisture content in the near-formation atmosphere on the energy change is shown, and the moisture mass has the same temperature change and the moisture mass has a larger energy change than the dry mass. K in the denominator is the energy conversion coefficient of regional sunshine hours, R is the sunshine hours in a certain time, kR is approximately used as the total energy in the time period, eta represents different underlying surfaces, and under the same sunshine condition, the energy change rate is different. Since the near-earth atmospheric pressure P does not vary much within a certain area, it is not introduced in the calculation.
From the above formula, it can be seen that, within a certain time, the near-earth atmospheric energy changes in proportion to the temperature and humidity change amount. In the formula, the temperature is calculated from 0 ℃, and the temperature T at a certain moment can directly represent delta T; by customary representation of C by H (relative humidity)ψH (absolute temperature) since at one temperature T the only corresponding H is found from H by the psychrometric chart of the humid air. The above equation can evolve as:
in the above formula, beta is a and CψAnd (4) the change coefficient after integration. However, in the above formula, the influence of other elements on energy change is not considered, and it is difficult to ensure that the sunshine hours of different stations are the same, so when an automatic station automatic clustering algorithm is implemented, the automatic station automatic clustering algorithm cannot be directly designed by using the energy change rate. However, the above expression indicates that the energy change rate has a direct correlation with the atmospheric element value, in other words, the characteristics of the underlying surface of the automatic weather station can be reflected by the correlation characteristics of the atmospheric element values such as H (humidity), T (humidity), R (sunshine hours), and P (atmospheric pressure).
Due to the influence of data acquisition errors, atmospheric system changes and production activities, data characteristic analysis cannot be carried out on data of one point or data of a plurality of points, and the properties of the sub-surfaces of the points under different automatic meteorological stations can be distinguished. In order to eliminate the influence of various errors or accidental factors, data at different time points are selected to form a data sequence with a certain length, comprehensive analysis and calculation are carried out, sequence signal characteristic quantity is extracted, and different underlying surfaces are distinguished through the characteristic quantity.
The time length setting method comprises the following steps:
the data are determined and verified by a big data analysis method, the general data are selected in units of days, and the time length is determined according to seasonal changes.
The time point selection method comprises the following steps:
the data information filed by the automatic weather station comprises minute data, hour data and day data. And selecting the atmospheric element value corresponding to the time with the lowest temperature in each day in the day data when the environmental interference is small and the signal is more stable to form a data sequence. For this reason, the above formula is derived as:
in the above formula, β and k can be approximately regarded as a constant in a similar region, i is day, HiThe humidity value at the lowest temperature moment of each day, Tmin(i) Is the lowest temperature value per day, RiIs the number of sunshine hours per day, Σ ηiIs a sequence of energy differences.
In the above analysis, it can be known that the atmospheric element value of the near-stratum of the station has strong correlation with the underlying surface, and the data signal features can be extracted through the atmospheric element value sequence to distinguish different underlying surfaces. However, the time of the data sequence becomes long, the underlying surface environment condition of each site changes, the stability of the signal characteristics obtained from the data sequence of the single site also changes, and difficulties are brought to the automatic clustering algorithm of the similar underlying surface sites.
Within a certain range and in the same time period, the environmental changes of similar underlying surfaces have great tendency, that is, the lowest energy difference value between automatic stations in a certain area and a certain time period tends to be constant, and the constancy of the energy difference value also reflects the difference value between corresponding atmospheric elements in the same time period in the same area. Therefore, the difference is made between the corresponding atmospheric element values at the same time point between two stations to form a new differential data sequence, and the formula is shown as the following formula:
in the above formula, Σ Δ ηiFor the two-station energy difference value sequence, when the two stations are in a certain area, the above formula can be similarly calculated, and is changed into a formula:
in the above formula, the first and second carbon atoms are,
∑Δηifor the differential data analysis row, the meteorological element value at the time of the lowest temperature every day between two stations is taken as the difference. The differential data analysis method reduces the influence on data analysis caused by long time and the change of the environment of the underlying surface, and the characteristic quantity extracted by the differential data sequence signal is more obvious and stable. Research and analysis of sample data show that when the similarity of the underlying surfaces of the two stations is high, the difference value of the two stations is concentrated to a certain value. When the number of days i of data is taken to be a certain length, the differential sequence data follows a normal distribution. It can be proved that if two stations at the same place simultaneously collect the atmospheric data of the same element, the data difference between the two stations is only related to the collection error, the length of the difference sequence is very short, the normal distribution characteristic can be shown, but the normal distribution characteristic is only obvious when the data difference sequence reaches a certain length due to the influence of various factors between the automatic meteorological stations of the same underlying surface. At present, for the analysis of 10-year-hour data samples of 50-year and 40-year automatic meteorological stations of 10 national ground observation stations, the length of differential sequence data of underlying similar stations is about 90 (i total number of i)Taking about 90 days, namely taking season as a unit) to carry out data analysis, the normal distribution characteristic of the data is obvious.
In this embodiment, when the basic change rules of the two stations are consistent (i.e., comparable stations), the data difference value of the two stations should be only affected by the noise distribution of the sensor, so that the sequence should obey normal distribution, and the probability in the error range is calculated (the error of the current temperature sensor is ± 0.1, and the error of the humidity sensor is large, so that the calculation range can be expanded), so as to measure the degree of similarity between the changes of the two stations. Defining: let Δ η sequence mean be a and variance be σ, and its distribution function be:
the two-station similarity coefficient (feature quantity) Col is:
dΔηfor the integral variable of Δ η, e is the observation error range, and only the temperature takes 0.1, if the error of humidity and the like is considered, the error can be extended to 0.3. The larger col is, the greater the association between two stations is, and the stronger the comparability is.
The invention is further illustrated by the following examples
(1) Inter-site data differential signal characteristic quantity extraction process
Taking data of all national ground observation stations within 100 kilometers, taking data of H, T at the lowest temperature moment every day, carrying out difference in pairs, taking the quarter as the time length, forming an element difference sequence, calculating a difference sequence signal characteristic quantity corresponding to the national ground observation stations, taking the characteristic quantity between the national ground observation stations as a basis and a reference for grouping, and then respectively carrying out difference sequence characteristic quantity calculation on the national ground observation stations and automatic meteorological stations within 50 kilometers.
The inter-site data signal characteristic quantity extraction process comprises the following steps: firstly, selecting the meteorological element values of H, T at the lowest temperature time of the two stations for difference to form a differential sequence, and secondly, testing the distribution characteristic of the temperature-cast differential data sequence. And finally, counting the interval probability and providing the differential sequence signal characteristic quantity.
(2) Comparable station set classification algorithm
The method comprises the steps of forming data differential sequences among national ground observation stations, national ground observation stations and automatic meteorological stations based on seasonal data within a certain range, extracting characteristic quantities of differential sequence signals by adopting a signal analysis method to form a classified characteristic quantity set among stations, taking the characteristic quantities among the national ground observation stations as grouping basis, wherein the characteristic quantities among the large stations have the maximum value and the minimum value due to the distance, taking the minimum value as a reference quantity of components, each national ground observation station conducts differential sequence on the automatic stations within the range of 50 kilometers, extracting characteristic quantities of differential sequence signals, stations larger than the reference quantity form a station cluster, stations smaller than the reference quantity form a station cluster, stations divided into the same cluster have similar underlying surface properties, and station-to-point meteorological data are high in comparability.
In summary, in the comparable site set classification, the method for calculating the features between sites is as follows:
the atmospheric elements (temperature, humidity, air pressure, sunshine and the like) at the lowest temperature moment of each day are taken between any two meteorological acquisition stations in a certain area to form a differential sequence. And the difference sequence is 90 data in unit of one season, whether the difference sequence meets the normal distribution characteristic is judged, if the difference does not meet the normal distribution characteristic, the difference of the underlying surfaces of the two stations is large, the data comparability is low, and if the difference sequence meets the normal characteristic, the signal characteristic quantity of the difference sequence of the two stations is calculated.
The basis of the site cluster is the size of the characteristic quantity of the differential sequence among the sites, and the national weather station and the automatic weather station are arranged in a certain range, have strict data quality management and maintenance standards and better data quality and are used as the standard for comparing and analyzing data of other automatic weather stations in the area. Therefore, a differential sequence is formed between all national ground observation stations within 100 kilometers of the area, and the characteristic quantity of the differential sequence is extracted. Due to the factors of the distance between the station stations, different buried positions and the like, the characteristic quantity between the national ground observation stations also has the maximum value and the minimum value, the minimum value is used as the first cluster reference characteristic quantity of the regional automatic observation station subset, and the minimum and maximum intermediate value is used as the second cluster reference characteristic quantity of the regional automatic observation station subset. Each national ground observation station carries out differential sequence on automatic stations within the range of 50 kilometers, differential sequence signal characteristic quantity is extracted, stations larger than the reference characteristic quantity of the first cluster form a station cluster, stations smaller than the reference characteristic quantity of the first cluster form another station cluster, the stations are divided into the same cluster, underlying surface properties of the stations are similar, and station-to-point meteorological data comparability is high. And (3) performing a differential data analysis sequence among automatic meteorological stations within 30 kilometers of the area, extracting characteristic quantity, wherein the characteristic quantity is greater than or equal to a second cluster reference characteristic quantity to form a station cluster as an auxiliary application cluster.
Example 2
As shown in fig. 2, the present invention provides an automatic weather station cluster system based on differential sequence, which includes:
the first differential sequence signal characteristic quantity extraction module is used for forming a first differential sequence by meteorological data of all national ground observation stations within 100 kilometers at the lowest temperature moment every day and extracting a first differential sequence signal characteristic quantity;
the cluster reference characteristic quantity confirmation module is used for respectively taking the maximum value and the minimum value of the first differential sequence signal characteristic quantity, taking the minimum value as a first cluster reference characteristic quantity, and taking the middle value between the minimum value and the maximum value as a second cluster reference characteristic quantity;
the first judgment module is used for judging whether the selection of all national ground observation stations within 100 kilometers is finished;
the second differential sequence signal characteristic quantity extraction module is used for selecting an automatic weather station which is 50 kilometers away from the national weather station, establishing a second differential sequence and extracting a second differential sequence signal characteristic quantity;
the first classification module is used for collecting all automatic weather stations with second difference sequence characteristic quantity smaller than the first cluster reference characteristic quantity to form a first site cluster, and collecting all automatic weather stations with second difference sequence characteristic quantity larger than or equal to the first cluster reference characteristic quantity to form a second site cluster to finish automatic cluster processing of the automatic weather stations;
the second judgment module is used for judging whether the automatic weather station in the selected area is selected completely;
the third differential sequence signal characteristic quantity extraction module is used for selecting an automatic weather station which is 30 kilometers away from the national weather station, establishing a third differential sequence and extracting a third differential sequence signal characteristic quantity;
and the second classification module is used for collecting all the automatic weather stations with the third differential sequence signal characteristic quantity greater than or equal to the second cluster reference characteristic quantity to form a third station cluster and finish the automatic clustering processing of the automatic weather stations.
As shown in fig. 2, the automatic weather station automatic clustering system provided in the embodiment can execute the technical solutions shown in the above method embodiments, and the implementation principles and beneficial effects thereof are similar, and are not described herein again.
Claims (6)
1. An automatic meteorological station clustering method based on a differential sequence is characterized by comprising the following steps:
s1, forming a first differential sequence by meteorological data of all national ground observation stations within 100 kilometers at the lowest temperature moment every day, and extracting signal characteristic quantity of the first differential sequence;
s2, respectively taking the maximum value and the minimum value of the first differential sequence signal characteristic quantity, taking the minimum value as a first cluster reference characteristic quantity, and taking the intermediate value between the minimum value and the maximum value as a second cluster reference characteristic quantity;
s3, judging whether the selection of all national ground observation stations within 100 kilometers is finished, if so, entering a step S6, otherwise, entering a step S4;
s4, selecting an automatic weather station which is 50 kilometers away from the national weather station, establishing a second differential sequence, and extracting signal characteristic quantity of the second differential sequence;
s5, collecting all automatic weather stations with the second difference sequence characteristic quantity smaller than the first cluster reference characteristic quantity to form a first site cluster, and collecting all automatic weather stations with the second difference sequence characteristic quantity larger than or equal to the first cluster reference characteristic quantity to form a second site cluster to finish automatic cluster processing of the automatic weather stations;
s6, judging whether the automatic weather station is selected in the selected area, if so, finishing the automatic clustering processing of the automatic weather station, otherwise, entering the step S7;
s7, selecting an automatic weather station which is 30 kilometers away from the national weather station, establishing a third differential sequence, and extracting signal characteristic quantity of the third differential sequence;
and S8, collecting all the automatic weather stations with the third differential sequence signal characteristic quantity more than or equal to the second cluster reference characteristic quantity to form a third station cluster, and finishing the automatic clustering processing of the automatic weather stations.
2. The automatic meteorological station clustering method based on differential sequence as claimed in claim 1, wherein the processes of extracting the first differential sequence signal feature quantity, extracting the second differential sequence signal feature quantity and extracting the third differential sequence signal feature quantity are the same, and the method comprises the following steps:
a1, forming a differential sequence by meteorological data at the lowest temperature time every day between any two national ground observation stations or automatic meteorological stations in a fixed area, wherein the differential sequence is a first differential sequence, a second differential sequence or a third differential sequence;
a2, judging whether the differential sequence meets normal distribution, if so, entering the step A3, otherwise, outputting an error prompt, and ending the process;
and A3, calculating the interval probability of the differential sequence, and calculating to obtain a differential sequence signal characteristic quantity according to the interval probability, wherein the differential sequence signal characteristic quantity is a first differential sequence signal characteristic quantity, a second differential sequence signal characteristic quantity or a third differential sequence signal characteristic quantity.
3. The automatic weather station cluster method based on differential sequence as claimed in claim 2, wherein the step A1 includes the following steps:
a101, selecting meteorological observation element values of any two national ground observation stations or automatic meteorological stations at the lowest temperature in a fixed area according to the atmospheric energy change rule of the near-stratum of the national ground observation stations or the automatic meteorological stations, and determining the correlation between the near-stratum atmospheric energy and the meteorological observation element values;
and A102, calculating the difference value of every two meteorological observation element values at the same time point between two national ground observation stations or automatic meteorological stations according to the association, and forming a differential sequence by taking a quarter as a time length, wherein the differential sequence is a first differential sequence, a second differential sequence or a third differential sequence.
4. The automatic weather station clustering method based on differential sequences as claimed in claim 3, wherein the expression of the differential sequence in step A102 is as follows:
wherein, Σ Δ ηiRepresenting a differential sequence, said differential sequence being a first, a second or a third differential sequence, gammaiRepresents beta, k and RiNormalized parameter, RiRepresenting the number of sunshine hours per day, beta and k both representing constants, Hi1Represents the humidity value T of the ground observation station or automatic meteorological station of a certain country at the lowest temperature moment every daymin(i1)Represents the daily minimum temperature value H of a ground observation station or an automatic meteorological station of a certain countryi2Representing the humidity value, T, of the lowest temperature moment of day of a ground observation station or an automated weather station of another countrymin(i2)Representing the lowest temperature value of each day of another national ground observation station or automated weather station.
5. The automatic weather station cluster method based on differential sequence as claimed in claim 4, wherein the expression of the characteristic quantity of the differential sequence signal in the step A3 is as follows:
wherein col represents a differential sequence signal characteristic quantity, the differential sequence signal characteristic quantity is a first differential sequence signal characteristic quantity, a second differential sequence signal characteristic quantity or a third differential sequence signal characteristic quantity, fΔηExpressing a normal distribution function, e expressing an observation error range, delta eta, the near-to-earth atmospheric energy difference of two automatic meteorological stations at the lowest temperature moment, dΔηRepresents the integral variable of Δ η, σ represents the variance, and a is the mean of the two-station difference data sequence.
6. An automated weather station cluster system based on differential sequences, comprising:
the first differential sequence signal characteristic quantity extraction module is used for forming a first differential sequence by meteorological data of all national ground observation stations within 100 kilometers at the lowest temperature moment every day and extracting a first differential sequence signal characteristic quantity;
the cluster reference characteristic quantity confirmation module is used for respectively taking the maximum value and the minimum value of the first differential sequence signal characteristic quantity, taking the minimum value as a first cluster reference characteristic quantity, and taking the middle value between the minimum value and the maximum value as a second cluster reference characteristic quantity;
the first judgment module is used for judging whether the selection of all national ground observation stations within 100 kilometers is finished;
the second differential sequence signal characteristic quantity extraction module is used for selecting an automatic weather station which is 50 kilometers away from the national weather station, establishing a second differential sequence and extracting a second differential sequence signal characteristic quantity;
the first classification module is used for collecting all automatic weather stations with second difference sequence characteristic quantity smaller than the first cluster reference characteristic quantity to form a first site cluster, and collecting all automatic weather stations with second difference sequence characteristic quantity larger than or equal to the first cluster reference characteristic quantity to form a second site cluster to finish automatic cluster processing of the automatic weather stations;
the second judgment module is used for judging whether the automatic weather station in the selected area is selected completely;
the third differential sequence signal characteristic quantity extraction module is used for selecting an automatic weather station which is 30 kilometers away from the national weather station, establishing a third differential sequence and extracting a third differential sequence signal characteristic quantity;
and the second classification module is used for collecting all the automatic weather stations with the third differential sequence signal characteristic quantity greater than or equal to the second cluster reference characteristic quantity to form a third station cluster and finish the automatic clustering processing of the automatic weather stations.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110982065.0A CN113704558B (en) | 2021-08-25 | 2021-08-25 | Automatic meteorological station automatic clustering method and system based on differential sequence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110982065.0A CN113704558B (en) | 2021-08-25 | 2021-08-25 | Automatic meteorological station automatic clustering method and system based on differential sequence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113704558A true CN113704558A (en) | 2021-11-26 |
| CN113704558B CN113704558B (en) | 2022-10-14 |
Family
ID=78654724
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110982065.0A Active CN113704558B (en) | 2021-08-25 | 2021-08-25 | Automatic meteorological station automatic clustering method and system based on differential sequence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113704558B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130073387A1 (en) * | 2011-09-15 | 2013-03-21 | Stephan HEATH | System and method for providing educational related social/geo/promo link promotional data sets for end user display of interactive ad links, promotions and sale of products, goods, and/or services integrated with 3d spatial geomapping, company and local information for selected worldwide locations and social networking |
| CN104156777A (en) * | 2014-05-08 | 2014-11-19 | 南方电网科学研究院有限责任公司 | Low-cost photovoltaic power generation power prediction method based on urban weather forecast |
| CN107007290A (en) * | 2017-03-27 | 2017-08-04 | 广州视源电子科技股份有限公司 | Electroencephalogram relaxation degree identification method and device based on time domain and phase space |
| CN107507097A (en) * | 2017-07-03 | 2017-12-22 | 上海电力学院 | A kind of short-term wind power prediction method |
| CN109919353A (en) * | 2019-01-11 | 2019-06-21 | 杭州电子科技大学 | A kind of distributed photovoltaic prediction technique of the ARIMA model based on spatial coherence |
| US20200228872A1 (en) * | 2010-11-01 | 2020-07-16 | Gracenote, Inc. | Method and system for presenting additional content at a media system |
| CN112380499A (en) * | 2020-11-11 | 2021-02-19 | 成都市气象局 | Regional automatic weather station objective clustering method |
| CN112559588A (en) * | 2020-12-08 | 2021-03-26 | 天津市气象信息中心(天津市气象档案馆) | Construction method of one-hundred-year uniform air temperature daily value sequence observed by ground meteorological station |
| CN112880634A (en) * | 2021-01-15 | 2021-06-01 | 山东省国土测绘院 | Elevation reference frame dynamic monitoring method based on CORS station network |
| CN112907062A (en) * | 2021-02-08 | 2021-06-04 | 国网安徽省电力有限公司蚌埠供电公司 | Power grid electric quantity prediction method, device, medium and terminal integrating temperature characteristics |
-
2021
- 2021-08-25 CN CN202110982065.0A patent/CN113704558B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20200228872A1 (en) * | 2010-11-01 | 2020-07-16 | Gracenote, Inc. | Method and system for presenting additional content at a media system |
| US20130073387A1 (en) * | 2011-09-15 | 2013-03-21 | Stephan HEATH | System and method for providing educational related social/geo/promo link promotional data sets for end user display of interactive ad links, promotions and sale of products, goods, and/or services integrated with 3d spatial geomapping, company and local information for selected worldwide locations and social networking |
| CN104156777A (en) * | 2014-05-08 | 2014-11-19 | 南方电网科学研究院有限责任公司 | Low-cost photovoltaic power generation power prediction method based on urban weather forecast |
| CN107007290A (en) * | 2017-03-27 | 2017-08-04 | 广州视源电子科技股份有限公司 | Electroencephalogram relaxation degree identification method and device based on time domain and phase space |
| CN107507097A (en) * | 2017-07-03 | 2017-12-22 | 上海电力学院 | A kind of short-term wind power prediction method |
| CN109919353A (en) * | 2019-01-11 | 2019-06-21 | 杭州电子科技大学 | A kind of distributed photovoltaic prediction technique of the ARIMA model based on spatial coherence |
| CN112380499A (en) * | 2020-11-11 | 2021-02-19 | 成都市气象局 | Regional automatic weather station objective clustering method |
| CN112559588A (en) * | 2020-12-08 | 2021-03-26 | 天津市气象信息中心(天津市气象档案馆) | Construction method of one-hundred-year uniform air temperature daily value sequence observed by ground meteorological station |
| CN112880634A (en) * | 2021-01-15 | 2021-06-01 | 山东省国土测绘院 | Elevation reference frame dynamic monitoring method based on CORS station network |
| CN112907062A (en) * | 2021-02-08 | 2021-06-04 | 国网安徽省电力有限公司蚌埠供电公司 | Power grid electric quantity prediction method, device, medium and terminal integrating temperature characteristics |
Non-Patent Citations (3)
| Title |
|---|
| PENG YIN 等: "SeqSphereVLAD: Sequence Matching Enhanced Orientation-invariant Place Recognition", 《2020 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS)》 * |
| 张久山 等: "由地面温度极值观测差值序列分析谈自动气象站质量控制", 《第28届中国气象学会年会——S1第四届气象综合探测技术研讨会》 * |
| 贺南 等: "一种新的气象数据分析和误差诊断方法", 《成都信息工程大学学报》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113704558B (en) | 2022-10-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110008982B (en) | Meteorological monitoring point selection method based on photovoltaic power generation output clustering | |
| CN112862630B (en) | Photovoltaic power prediction method, terminal and medium based on weather type index interval | |
| CN109165693B (en) | Automatic identification method suitable for dew, frost and icing weather phenomena | |
| CN112380499B (en) | Regional automatic weather station objective clustering method | |
| CN107607692B (en) | Soil moisture monitoring and optimizing point distribution method based on maximum water storage capacity of soil | |
| CN109740195B (en) | Extreme value typhoon wind speed probability distribution model based on meteorological station observation data and evaluation method for designing typhoon wind speed | |
| CN110009385B (en) | Multi-factor mapping-based photovoltaic power generation user group division method | |
| CN110211325A (en) | A kind of area road icing high precision monitor early warning system based on meteorological big data | |
| CN114429053B (en) | Basin scale WEFE system adaptability simulation optimization method | |
| CN106897957B (en) | Automatic weather station real-time data quality control method based on PCA and PSO-E L M | |
| CN112348264A (en) | Carbon steel corrosion rate prediction method based on random forest algorithm | |
| CN113987912A (en) | Pollutant on-line monitoring system based on geographic information | |
| CN112114384A (en) | Power transmission line icing occurrence probability forecasting method | |
| CN114996943B (en) | Mesoscale numerical simulation method for reservoir storage climate effect evaluation | |
| CN114021830A (en) | Multi-time-range wind speed prediction method based on CNN-LSTM | |
| CN115575601A (en) | Vegetation drought index evaluation method and system based on water vapor flux divergence | |
| CN109543911B (en) | Sunlight radiation prediction method and system | |
| CN116800172B (en) | Photoelectric conversion method and energy storage system | |
| CN116961575B (en) | Tea light complementary photovoltaic power station monitoring system and method based on digital twin | |
| CN113704558B (en) | Automatic meteorological station automatic clustering method and system based on differential sequence | |
| CN108876060B (en) | Big data based prediction method for wind power output probability of sample collection | |
| CN117236720A (en) | Photovoltaic power station generated power prediction method utilizing multi-meteorological factor characteristics | |
| CN115456440A (en) | Method, device and system for evaluating power generation characteristics of distributed photovoltaic power station | |
| CN115166866A (en) | Citrus disease and insect pest occurrence forecasting method and system based on lattice point meteorological data | |
| CN118643468B (en) | Drought event identification method considering water balance |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |