A self-maintenance method based on automatic weather station data dictionary

Through the self-maintenance method of automatic weather station data dictionary, the probability distance, part-of-speech radius and hash dictionary algorithm are used to self-repair the business data dictionary, which solves the problem of mismatch and inconsistency of the data dictionary, and improves the reliability of the data dictionary and Practicality reduces program error rate and troubleshooting time.

CN118733112BActive Publication Date: 2024-11-12SICHUAN INFORMATION TECH COLLEGE
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411231297.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-04
Publication Date
2024-11-12
Estimated Expiration
2044-09-04

AI Technical Summary

Technical Problem

It is difficult to detect mismatch and inconsistency before the operation of the business software, which makes it difficult to locate and troubleshoot program operation errors, which takes time, affecting the reliability and practicality of the data dictionary.

Method used

The self-maintenance method based on the automatic weather station data dictionary is adopted, including four steps: data preparation, data loading, precise search and self-repair. The elements in the business data dictionary are self-repaired through probabilistic distance, part-of-speech radius and hash dictionary algorithms. Ensure the accuracy and consistency of the data dictionary.

Benefits of technology

It effectively reduces the program error rate, shortens the error handling time, improves the reliability and practicality of the data dictionary, realizes self-repair and precise positioning of the data dictionary, and reduces the program coupling.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118733112B_ABST
    Figure CN118733112B_ABST
Patent Text Reader

Abstract

The present invention discloses a self-maintenance method based on an automatic weather station data dictionary. The method includes four parts: data preparation, data loading, accurate search, and self-maintenance algorithm. The present invention can self-check the data dictionary when the automatic weather station business software is started or executed alone, and self-repair the questionable or erroneous data dictionary according to the characteristics of the automatic weather station, thereby reducing unnecessary errors in program operation and programmers' troubleshooting time. Experiments have shown that this algorithm has excellent self-maintenance capabilities for automatic weather station data dictionaries, and the algorithm is highly portable.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of automatic weather station data dictionaries, in particular to a self-maintenance method based on automatic weather station data dictionaries. Background Art

[0002] A data dictionary is a document that defines and describes the data elements used in a system or application. It is a structured document that usually contains information about databases, data tables, column names, data types, lengths, relationships, and so on. A data dictionary can help developers understand the structure and meaning of data, and can also serve as one of the important documents for team collaboration, improving code maintenance and development efficiency. Data dictionaries are widely used, for example: (1) Database design and management: During the database design phase, the ER model is usually designed first, and then the table, column name, data type, length, relationship, and other information are defined through the data dictionary to help developers build and manage the database. In database management, the data dictionary can provide a detailed description of the data, including the meaning of the table and column, the source of the data, the data specification, etc., to facilitate administrators to perform database maintenance and data analysis; (2) System development and maintenance: The data dictionary can serve as an important document in the system development process, recording the definition and description of various data elements used in the system. It can help developers understand the data structure and business logic, and improve the readability and maintainability of the code. At the same time, during the system maintenance phase, the data dictionary can also help the development team understand the system functions and data flow, and quickly locate and solve problems; (3) Data quality management: The data dictionary can help manage and monitor data quality. By defining the specifications of data elements, validation rules, and data sources, data can be effectively checked and corrected to ensure data accuracy, completeness, and consistency; (4) Data analysis, processing, and reporting: The data dictionary can provide a detailed description of the data and its business meaning, helping analysts understand the data and extract, transform, and load data as needed. In the reporting and analysis process, the data dictionary can play a guiding and explanatory role to ensure the accuracy and consistency of data use.

[0003] In short, data dictionaries play an important role in database design, system development, data quality management and data analysis, helping to organize and manage data and improve data reliability and practicality. However, due to the diversity, importance and extensiveness of the role of data dictionaries themselves, there are few studies on how to ensure the accuracy of data dictionaries. The present invention takes the data dictionary of an automatic weather station as the background to study the self-maintenance method of data dictionaries. Through the automatic generation of data dictionaries, before the business software is run, the business data dictionary is tested with a self-maintenance method to solve the problems of data dictionary mismatch and disunity, which makes it difficult to find program running errors, and the troubleshooting program is complicated and time-consuming, so as to improve the reliability, practicality and intelligence of data dictionaries. Summary of the invention

[0004] The object of the present invention is to provide a self-maintenance method based on an automatic weather station data dictionary in order to solve the above-mentioned problems.

[0005] The technical solution adopted by the present invention is as follows:

[0006] A self-maintenance method based on an automatic weather station data dictionary includes four steps: data preparation, data loading, precise search, and self-repair. Specifically,

[0007] S1: Data preparation: prepare full-factor data dictionary and business data dictionary;

[0008] S2: Data loading: Loading the full-factor data dictionary and business data dictionary;

[0009] S3: Accurate search: Find whether the elements and business processes in the business data dictionary exist in the full factor data dictionary;

[0010] S4: Self-repair: The elements in S3 that do not exist in the full-factor data dictionary are self-repaired through probability distance, part-of-speech radius, and hash dictionary algorithms.

[0011] Furthermore, S1 is to prepare the full factor data dictionary and the business data dictionary. The preparation process is to automatically form a uniformly specified xml file from the database or various forms of tables in the full factor data dictionary and the business data dictionary according to business requirements through a program. The full factor data dictionary contains all meteorological elements and features of the automatic weather station, and the business data dictionary needs to include: manual and semi-manual elements of the automatic weather station, elements that have been transformed into automation, newly added elements, collection methods of each element, processing methods of each element, and storage paths of element files according to business needs.

[0012] Furthermore, the S2 is loaded in a double loop, the first loop sequentially loads the business data dictionary from the first element, and the second loop sequentially loads the full-element data dictionary from the first element.

[0013] Furthermore, S3 is based on S2, and the second-level loop starts from the first element of the full-factor data dictionary, searches in sequence, and determines whether the element in the business data dictionary is in the full-factor dictionary, and whether it is exactly the same element. If so, continue the loop, if not, enter the S4 self-repair step.

[0014] Furthermore, the S4 mainly includes three steps of S41 calculating probability distance, S42 calculating part-of-speech distance, and S43 hash dictionary correction. These three steps need to be executed according to conditions.

[0015] Further, the formula of S41 is:

[0016]

[0017]

[0018]

[0019]

[0020] in, is the probability distance, P E Indicates the ratio of the number of characters in meteorological element name A in the business data dictionary and meteorological element name B in the full element data dictionary to the number of characters in A. P D Indicates the ratio of the number of characters in meteorological element name A in the business data dictionary to the number of characters in meteorological element name B in the full factor data dictionary, where is the number of characters in the meteorological element name A, is the number of characters in the meteorological element name B, Indicates the sum of the number of identical characters in the meteorological element names A and B. Pick and The maximum value of .

[0021] Further, the formula of S42 is:

[0022]

[0023] in, Indicates the part-of-speech distance, and also serves as the radius of the circle in the diagram. The larger the radius, the less similar the part-of-speech. If the part-of-speech is exactly the same, it converges to the center of the circle. In order to distinguish other points, the center of the circle is represented by a diamond. In the formula, β is the weighted value, which is usually determined based on experience. Here, When β is 1, because in the automatic weather station, more than 70% of the elements are distinguished by the last two words. If the last two words can be distinguished, there is no need to calculate the subsequent probabilities, except In this case, β is a specific value or a function, but the final β value must be less than or equal to 1; represents the probability of taking the part of speech of the last two characters of meteorological element names A and B relative to the last two characters of meteorological element name A; represents the probability of the part of speech when taking the last two characters of meteorological element names A and B relative to the last two characters of meteorological element name B; It represents the probability of taking the last 3, 4, and 5 characters of meteorological element names A and B respectively.

[0024] Further, the specific steps of S43 are as follows:

[0025] S431, setting probability distance determination conditions;

[0026] S432. Classification data correction method.

[0027] Further, the setting of the probability distance determination condition is specifically classified into (where is the distance from the probability distance to the coordinate (1, 1), and its formula is: ):

[0028] C1. When there is an element , prioritize the part-of-speech distance of the element, sort the elements in the range from small to large according to the part-of-speech distance, select the element with the smallest part-of-speech distance, if more than two elements have the same part-of-speech distance, select the element with the smallest probability distance; if the minimum part-of-speech distance exceeds 0.1, only consider the probability distance, sort the probability distance, take the element with the smallest value, and correct the selected value with a hash dictionary;

[0029] C2. When all elements If the part-of-speech distance of one element is 0, the current element is selected. If the part-of-speech distance of more than two elements is 0, the element with the smallest probability distance is selected; otherwise, the two values ​​with the smallest probability distance and part-of-speech distance are selected and displayed in a pop-up prompt box on the program interface, requesting manual assistance, waiting for the manual response and then continuing to run the program. According to the result of manual assistance, the selected value is updated to the hash dictionary for correction.

[0030] Furthermore, the classification data correction method is specifically classified into:

[0031] C1.1. For the C1 classification, the selected element is first searched for a key-value pair in the hash dictionary, and then replaced. The replaced element value is replaced in the business data dictionary, and then an accurate search is performed starting from this element. Usually, an accurate match can be successfully made. If it does not exist in the hash dictionary, the current two values ​​are used as the key value pair of the hash dictionary to update the hash dictionary;

[0032] C2.1. For the C2 classification, if an element can be found, first look for this key-value pair in the hash dictionary. If the key-value pair does not exist in the hash dictionary, the hash dictionary cannot be updated with the key value at this time, but requires manual processing. If the result of manual assistance is confirmed, the hash table is updated with this value, otherwise the manual processing result is directly replaced in the business data dictionary.

[0033] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0034] The present invention provides a self-maintenance method based on the automatic weather station data dictionary, which automatically generates the automatic weather station full-factor data dictionary and the business data dictionary, and then detects each business data dictionary each time the business software is started, and self-repairs the elements in S3 that do not exist in the full-factor data dictionary through algorithms such as probability distance, part-of-speech radius, and hash dictionary, so as to achieve the purpose of reducing program error rate, reducing troubleshooting time, accurately locating, and reducing program coupling. Experiments have proved that this method can self-repair the business data dictionary to a large extent with high accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The present invention will now be described by way of example with reference to the accompanying drawings, in which:

[0036] Figure 1 It is a schematic diagram of the overall process;

[0037] Figure 2 It is a schematic diagram of the self-repair process;

[0038] Figure 3 It is the probability distance and part-of-speech distance graph of "5cm ground temperature". DETAILED DESCRIPTION

[0039] The present invention will be described in detail below in conjunction with the accompanying drawings.

[0040] All features disclosed in this specification, or steps in all methods or processes disclosed, except mutually exclusive features or steps, can be combined in any manner.

[0041] Any feature disclosed in this specification, unless otherwise stated, can be replaced by other equivalent or alternative features with similar purposes. Unless otherwise stated, each feature is only an example of a series of equivalent or similar features.

[0042] The automatic weather station data dictionary function application flow chart of the present invention needs to be able to accurately and correctly direct the formation of data files. Therefore, it is very necessary to test the data dictionary before starting the business software. If there is a problem, it can be learned and self-repaired. Please refer to Figure 1This application proposes a self-maintenance method based on the automatic weather station data dictionary, which includes four steps: data preparation, data loading, precise search, and self-repair. Specifically:

[0043] S1: Data preparation: Prepare the generated full-factor data dictionary and business data dictionary. For further information, please refer to Figure 1 , S1 is to prepare the full-factor data dictionary and the business data dictionary. The preparation process is to automatically form a unified xml file from the database or various forms of tables in the full-factor data dictionary and the business data dictionary according to business requirements through a program. The full-factor data dictionary contains all the meteorological elements and features of the automatic weather station, while the business data dictionary is based on business needs, but needs to include: manual and semi-manual elements of the automatic weather station, elements that have been transformed into automation, newly added elements, collection methods of each element, processing methods of each element, and element file storage paths. In order to illustrate this embodiment in more detail here, different element fragments of the full-factor data dictionary and the business data dictionary are intercepted in the figure ("5 cm ground temperature" and "5cm ground temperature").

[0044] S2: Data loading: Loading the full-factor data dictionary and the business data dictionary. Furthermore, the S2 is loaded in a double loop, the first loop loads the business data dictionary from the first element in sequence, and the second loop loads the full-factor data dictionary elements from the first one in sequence, and each element needs to be compared with the elements of the business data dictionary. For example, at this time, the business data dictionary has loaded "5cm ground temperature", and then it will be calculated and compared with each element name in the full-factor data dictionary.

[0045] S3: Precise search: Find whether the elements and business processes in the business data dictionary exist in the full factor data dictionary. For further information, see Figure 1 , S3 is based on S2, starting from the second layer of loop to search the elements of the full-factor data dictionary in turn, to determine whether the elements in the business data dictionary are in the full-factor dictionary, and whether they are exactly the same elements. If so, continue the loop, if not, enter the S4 step. In this embodiment, when running to "5cm ground temperature", no element name exactly the same as it is obtained, it is necessary to enter the S4 step.

[0046] S4: Self-repair: self-repair the elements in S3 that do not exist in the full factor data dictionary through probability distance, part-of-speech radius, and hash dictionary algorithms. For further information, please refer to Figure 1-3, S4 mainly includes three steps: finding probability distance, part-of-speech distance, and hash dictionary correction. These three steps need to be executed according to conditions, and the probability distance and part-of-speech distance can also be displayed by diagrams, such as Figure 3 As shown, and defined: The X-axis represents the probability relative to the business data dictionary P E , the Y axis represents the probability relative to the full factor data dictionary P D , A position indicates a complete match, the distance from other points to A is called probability distance, and the radius of the circle indicates the part-of-speech distance. The shorter the probability distance, the more similar the two are, and the more likely they are to represent the same element. The smaller the circle, the more similar the part-of-speech is, and the more likely they are to be in the same category. If the part-of-speech is exactly the same, it is represented by a diamond. At the same time, in this algorithm, the part-of-speech distance is specified to have the highest priority. If the part-of-speech distance is smaller within a certain distance range, the element is corrected first.

[0047] Further, the formula of S41 is:

[0048]

[0049]

[0050]

[0051]

[0052] in, is the probability distance, P E Indicates the ratio of the number of characters in meteorological element name A in the business data dictionary and meteorological element name B in the full element data dictionary to the number of characters in A. P D Indicates the ratio of the number of characters in meteorological element name A in the business data dictionary to the number of characters in meteorological element name B in the full factor data dictionary, where is the number of characters in the meteorological element name A, is the number of characters in the meteorological element name B, Indicates the sum of the number of identical characters in the meteorological element names A and B. Pick and The maximum value of .

[0053] Furthermore, The value of is based on the following formula:

[0054]

[0055] in, Represents the first characters, Represents the first characters, = hour( Take 1~ , Take 1~ ), that is, when the characters in the names of the two meteorological elements A and B are the same The value is 1, otherwise the value is 0.

[0056] Therefore, for the element "5cm ground temperature" in the business data dictionary at this time, it is necessary to calculate the probability distance between each element name in the full factor dictionary. According to the S41 formula, take the three elements of "5cm ground temperature", "15cm ground temperature" and "maximum temperature occurrence time" in the full factor data dictionary as examples:

[0057] In this embodiment, "5cm ground temperature" is the meteorological element name A in the business data dictionary, and "5cm ground temperature", "15cm ground temperature" and "time of maximum temperature occurrence" are the meteorological element names B1, B2 and B3 in the full factor data dictionary respectively.

[0058] First, the "5cm ground temperature" in the business data dictionary and the "5cm ground temperature" in the full element dictionary can be obtained according to the conditions ,when hour Take the number 1, otherwise, Take the number 0. Through "5cm ground temperature" and "5 cm ground temperature", we can find The value of is 3, because there are 3 identical characters "5", "ground", and "temperature". It can be concluded that n=5, h=5 at this time, so , ;

[0059] Continue the second loop. At this time, the business data dictionary element "5cm ground temperature" encounters the "15cm ground temperature" element in the full element dictionary. Therefore, the same conditions can be used to find ,when hour Take the number 1, otherwise, Take the number 0. Through "5cm ground temperature" and "15cm ground temperature", we can find The value of is 3, because there are 3 identical characters "5", "land", and "temperature". It can be concluded that n=5, h=6, so , ;

[0060] Continue with the second-layer loop. At this time, the business data dictionary element "5 cm ground temperature" encounters the element "time of occurrence of the highest temperature" in the all-elements dictionary. Therefore, similarly, according to the conditions, it can be calculated that , when time take the number 1, in other cases, take the number 0. Through "5 cm ground temperature" and "time of occurrence of the highest temperature", the value is 1 because there is 1 same character "temperature". Furthermore, it can be obtained that at this time n = 5, h = 8, so , ;

[0061] Other elements are calculated in turn according to the same method.

[0062] Furthermore, for S42, its formula is:

[0063]

[0064] where represents the part-of-speech distance and also serves as the Figure 3 radius of the circle in. The larger the radius, the less similar the parts of speech are. If the parts of speech are exactly the same, they converge to the center of the circle. To distinguish other points, the center of the circle is represented by a rhombus at this time; in the formula, β is the weighting value, which is usually determined according to empirical values. Here, when β takes 1 because in an automatic weather station, more than 70% of the elements are distinguished by the last two characters in terms of parts of speech. If they can be distinguished by the last two characters, there is no need to calculate the subsequent probabilities anymore, except for this situation, β takes a specific value or a function, but ultimately the value of β needs to satisfy being less than or equal to 1; represents the part-of-speech probability of taking the last 2 characters of the meteorological element names A and B relative to the last 2 characters of the meteorological element name A; represents the part-of-speech probability of taking the last 2 characters of the meteorological element names A and B relative to the last 2 characters of the meteorological element name B; represents the probabilities of taking the last 3, 4, and 5 characters of the meteorological element names A and B respectively for the last 3, 4, and 5 characters of the meteorological element name A and the meteorological element name B.

[0065] Similarly, taking the three elements of "5 cm ground temperature" in the business data dictionary and "5 cm ground temperature", "15 cm ground temperature", and "time of occurrence of the highest temperature" in the all-elements data dictionary as examples. According to S41, the probability distances have been calculated respectively. At this time, calculate the part-of-speech distance. and The calculation is still performed according to the S41 formula, but the numbers now represent the characters in reverse order from the element name, such as , is to take the "ground temperature" in "5cm ground temperature", at this time Take the "ground temperature" in "5 cm ground temperature", so , so β ​​is 1, Similarly, the difference between "5cm ground temperature" and "15cm ground temperature" in the full factor data dictionary is . Then find the business data dictionary "5cm ground temperature" and the full factor data dictionary "highest temperature occurrence time", and =0, because and The characters extracted are "ground temperature" and "time". and , and The values ​​obtained are all 0, because this automatic weather station mainly uses the last two characters, so except for the case of 1, the values ​​of other cases are: hour, ,so ,roll out ≈0.1.

[0066] Based on the above inference, please refer to Table 1 and Figure 3 In this embodiment, the data obtained by using the formula for "5cm ground temperature" and some elements in the full factor data dictionary are shown in Table 1, where is the distance from the probability distance to the coordinate (1, 1), and its formula is: According to the data, we can draw Figure 3 The meteorological element part-of-speech distance and probability distance diagram shown are convenient for analysis and observation.

[0067] Table 1 Parameter values ​​calculated by the meteorological element formula of "5cm ground temperature"

[0068]

[0069] For further information, see Figure 2 , the specific steps of S43 are as follows:

[0070] S431, setting probability distance determination conditions;

[0071] S432. Classification data correction method.

[0072] Furthermore, the setting of the probability distance determination condition is specifically classified into:

[0073] C1. When there is an element , that is, the element is located in Figure 3In the fan-shaped area in the upper right corner, the part-of-speech distance of the element is determined first, and the elements in the range are sorted from small to large according to the part-of-speech distance, and the element with the smallest part-of-speech distance is selected. If the part-of-speech distances of more than two elements are the same, the element with the smallest probability distance is selected; if the minimum part-of-speech distance exceeds 0.1, only the probability distance is considered, the probability distance is sorted, and the minimum value element is selected. The selected value is corrected using a hash dictionary. In this embodiment, please refer to Figure 3 , "5cm ground temperature" and "15cm ground temperature" elements belong to the C1 condition, and the part-of-speech distances of "5cm ground temperature" and "15cm ground temperature" are both 0, so the element "5cm ground temperature" with a smaller probability distance is selected in this area according to the condition;

[0074] C2. When all elements If the part-of-speech distance of one element is 0, the current element is selected. If the part-of-speech distance of more than two elements is 0, the element with the smallest probability distance is selected; otherwise, the two values ​​with the smallest probability distance and part-of-speech distance are selected and displayed in a pop-up prompt box on the program interface, requesting manual assistance, waiting for the manual response and then continuing to run the program. According to the result of manual assistance, the selected value is updated to the hash dictionary for correction.

[0075] Furthermore, the classification data correction method is specifically classified into:

[0076] C1.1. Please refer to Figure 2 , for the C1 classification, the selected element here first searches for a key-value pair in the hash dictionary, and then replaces it. The replaced element value is replaced in the business data dictionary, and then an accurate search is performed starting from this element. Usually, an accurate match can be successfully achieved. If it does not exist in the hash dictionary, the current two values ​​are used as the key-value pair of the hash dictionary to update the hash dictionary. In this embodiment, "5cm ground temperature" and "5 centimeters ground temperature" need to be corrected with the hash dictionary. The key-value pair of "cm" and "centimeter" is found in the hash dictionary. At this time, "cm" is replaced with "centimeter", and then the full element data dictionary is searched to obtain an accurate match value;

[0077] C2.1. For the C2 classification, if an element can be found, first look for this key-value pair in the hash dictionary. If the key-value pair does not exist in the hash dictionary, the hash dictionary cannot be updated with this value at this time, but requires manual processing. If the result of manual assistance is certain, the hash table can be updated with this value. Otherwise, the manual processing result is directly replaced in the business data dictionary. The present invention is not limited to the aforementioned specific implementation methods.

[0078] See also Figure 1In the above four steps, in practical applications, taking the automatic weather station data dictionary as an example, data preparation is first performed, and all the collected elements, processing module paths, and interface information are automatically generated from the document table or database as required to generate the full-factor data dictionary, and then the business data dictionary is automatically generated according to the process and processing requirements of the automatic weather station business. Before the business program is started, the business data dictionary is checked using the exact matching method. If there is a problem with the business data dictionary, the self-repair process is entered to screen the full-factor data dictionary elements by calculating the probability distance and part-of-speech radius, and the self-repair function of the business data dictionary is completed with the assistance of the hash dictionary. At the same time, the content of the hash dictionary can also be expanded for continuous self-learning and improvement.

[0079] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A self-maintenance method based on an automatic weather station data dictionary, characterized in that: It includes four steps: data preparation, data loading, precise search, and self-repair. Specifically: S1: Data preparation: prepare full-factor data dictionary and business data dictionary; S2: Data loading: Load the full-factor data dictionary and business data dictionary elements in sequence; S3: Accurate search: Find whether the elements and business processes in the business data dictionary exist in the full factor data dictionary; S4: Self-repair: self-repair the elements in S3 that do not exist in the full-factor data dictionary through probability distance, part-of-speech radius, and hash dictionary algorithms; The S4 comprises the following steps which are executed in sequence according to the conditions: S41, find the probability distance; S42, find the part of speech distance; S43, hash dictionary correction; The formula of S41 is: in, is the probability distance, P E Indicates the ratio of the number of characters in meteorological element name A in the business data dictionary and meteorological element name B in the full element data dictionary to the number of characters in A. P D Indicates the ratio of the number of characters in meteorological element name A in the business data dictionary to the number of characters in meteorological element name B in the full factor data dictionary, where is the number of characters in the meteorological element name A, is the number of characters in the meteorological element name B, Indicates the sum of the number of identical characters in the meteorological element names A and B. Pick and The maximum value of The formula of S42 is: in, represents the part-of-speech distance; β is the weighted value, when When β takes 1, , β is a specific value or function less than 1; It represents the ratio of the last two characters of meteorological element names A and B to the last two characters of meteorological element name A, i.e., the part-of-speech probability; It represents the proportion of the last two characters of meteorological element names A and B that are identical to the last two characters of meteorological element name B, i.e., the part-of-speech probability; Indicates the percentage of the same characters in the last 3, 4, and 5 characters of meteorological element names A and B, respectively; The specific steps of S43 are as follows: S431, setting probability distance determination conditions; S432, classification data correction method; The S431 is specifically classified as follows: C1. When there is an element , prioritize the part-of-speech distance of the element, sort the elements in the range from small to large according to the part-of-speech distance, select the element with the smallest part-of-speech distance, if more than two elements have the same part-of-speech distance, select the element with the smallest probability distance; if the minimum part-of-speech distance exceeds 0.1, only consider the probability distance, sort the probability distance, take the element with the smallest value, and correct the selected value with a hash dictionary; C2. When all elements If the part-of-speech distance of one element is 0, the current element is selected. If the part-of-speech distance of two or more elements is 0, the element with the smallest probability distance is selected; otherwise, the two values ​​with the smallest probability distance and part-of-speech distance are selected and displayed in a pop-up prompt box on the program interface, requesting manual assistance, waiting for the manual response and then continuing to run the program. According to the manual assistance result, the selected value is updated to the hash dictionary for correction; The S432 is specifically classified as follows: C1.

1. For the C1 classification, the selected element is first searched for a key-value pair in the hash dictionary, and then replaced. The replaced element value is replaced in the business data dictionary, and then an accurate search is performed starting from this element. If it does not exist in the hash dictionary, that is, the match is not successful, the current two values ​​are used as the key value pair of the hash dictionary to update the hash dictionary; C2.

1. For the C2 classification, if an element can be found, first look for this key-value pair in the hash dictionary. If the key-value pair does not exist in the hash dictionary, the hash dictionary cannot be updated with this value at this time, but requires manual processing. If the result of manual assistance is confirmed, the hash table is updated with this value, otherwise the manual processing result is directly replaced in the business data dictionary.

2. The self-maintenance method based on the automatic weather station data dictionary according to claim 1, characterized in that: The S1 is to prepare the full-factor data dictionary and the business data dictionary. The preparation process is to automatically form a unified file with each element in the full-factor data dictionary and the business data dictionary according to business requirements through a program; wherein the full-factor data dictionary contains all meteorological elements and features of the automatic weather station, and the business data dictionary contains the manual and semi-manual elements of the automatic weather station, the elements that have been transformed into automation, the newly added elements, the collection method of each element, the processing method of each element, and the storage path of the element file.

3. The self-maintenance method based on the automatic weather station data dictionary according to claim 1, characterized in that: The S2 is loaded in a double loop, the first loop loads the business data dictionary from the first element in sequence, and the second loop loads the full-factor data dictionary from the first element in sequence.

4. The self-maintenance method based on the automatic weather station data dictionary according to claim 3, characterized in that: The S3 is based on the S2. The second-level loop starts from the first element of the full-factor data dictionary and searches in sequence to determine whether the element in the business data dictionary is in the full-factor dictionary and whether it is exactly the same element. If so, the loop continues. If not, enter the S4 self-repair step.