Abnormal meteorological temperature data restoration method based on space-time correlation

Abstract

The invention discloses an abnormal meteorological temperature data restoration method based on space-time correlation, and belongs to the technical field of meteorological data processing, and the method comprises the steps: obtaining temperature data as a data processing object; converting the acquired temperature data into Kelvin temperature data; calculating an average time sequence difference value, a daily period difference value and an average space difference value of the temperature data at the current moment based on the Kelvin temperature data; on the basis of the average time sequence difference value, the daily period difference value, the average space difference value and the corresponding threshold value range, whether the temperature data at the current moment are abnormal or not is judged, abnormal temperature data are locked, and an abnormal temperature data deviation value is calculated; and repairing the locked abnormal temperature data based on the locked abnormal temperature data and the abnormal temperature data deviation value to obtain repaired complete temperature data. The reliability of the judgment result of the meteorological data analysis and prediction model can be improved.

Description

Technical Field

[0001] The invention relates to the technical field of meteorological data processing, and in particular to a method for repairing abnormal meteorological temperature data based on time-space correlation. Background Art

[0002] Driven by the continuous progress of meteorological monitoring technology, we are able to obtain more comprehensive and rich meteorological data, among which temperature data, as one of the key elements, is of indispensable value to many fields. However, it is worth noting that despite the continuous technological innovation, the process of meteorological data collection is still subject to the complex and changeable external environment, such as equipment failure, operational errors and natural interference, etc. These factors often lead to noise or errors in temperature data, which is abnormal data.

[0003] Abnormal data, if not detected and processed in a timely manner, will directly affect the quality of temperature data and the accuracy of subsequent data analysis, which may mislead the results of meteorological analysis and weather forecasting. For example, a temperature reading that is too high or too low due to equipment problems may lead to an erroneous weather forecast, affecting agricultural crop planting decisions or public response measures. In meteorology, accurate data is the key to making accurate forecasts. Processing abnormal data can improve the accuracy of data and thus improve the reliability of forecasting models. Fixing abnormal data helps improve the quality of meteorological data, ensure the consistency and reliability of data, and thus enhance the credibility of data. This enhances the effectiveness of analysis and research on phenomena such as climate change and extreme weather events. Therefore, processing abnormal data in meteorological data is of great significance to improve the accuracy, quality and reliability of data, as well as enhance the performance of forecasting models.

[0004] At present, although a variety of abnormal data processing and repair methods have been proposed and applied, including statistical methods, machine learning techniques, etc., to improve the accuracy and reliability of data, these methods often only process data from historical data that depends on the time series dimension, and their efficiency and effectiveness are still limited when processing high-dimensional and large-scale meteorological data. Summary of the invention

[0005] The purpose of the present invention is to provide a method for repairing abnormal meteorological temperature data based on spatiotemporal correlation, which locks abnormal temperature data and calculates the deviation value of abnormal temperature data according to the average time series difference, daily cycle difference, average spatial difference and corresponding threshold range of temperature data at the current moment. Finally, the repaired complete temperature data is calculated based on the repair formula to improve the reliability of the judgment results of meteorological data analysis and prediction models. The present invention is achieved through the following technical solutions.

[0006] A method for repairing abnormal meteorological temperature data based on spatiotemporal correlation, comprising:

[0007] Obtaining the surface temperature data of the target area as the data processing object;

[0008] Convert the acquired temperature data into Kelvin temperature data and perform data standardization;

[0009] Based on the standardized Kelvin temperature data, the average time series difference, daily cycle difference and average spatial difference of the temperature data at the current moment are calculated;

[0010] Based on the average time series difference, daily cycle difference, average spatial difference and corresponding threshold range, determine whether the temperature data at the current moment is abnormal, lock the abnormal temperature data, and calculate the abnormal temperature data deviation value;

[0011] The locked abnormal temperature data is repaired based on the locked abnormal temperature data and the abnormal temperature data deviation value to obtain the repaired complete temperature data.

[0012] In the actual collection process of meteorological temperature data, the complex and changeable external environment may cause abnormal values ​​in the collected temperature data. The present invention judges whether there is noise or error in the temperature data at the current moment, that is, whether there is an abnormality in the current temperature data and locks the abnormal temperature data based on the average time series difference of adjacent moments of historical temperature data, the daily cycle difference and the average spatial difference of adjacent spatial points at the same moment. For the abnormal temperature data existing in the collection process, the three difference parameters of the temperature data at the current moment are combined with the corresponding weights to calculate the abnormal temperature data deviation value. Finally, the abnormal temperature data and the abnormal temperature data deviation value are superimposed through the repair formula to obtain the repaired complete temperature data, which improves the reliability of the judgment results of the meteorological data analysis and prediction model.

[0013] Optionally, converting the acquired temperature data into Kelvin temperature data and performing data standardization includes:

[0014] If the temperature data is in Celsius, the formula for converting it to Kelvin temperature data and normalizing the data is:

[0015] ,

[0016] If the temperature data is obtained in Fahrenheit, the formula for converting it to Kelvin temperature data and normalizing the data is:

[0017] ,

[0018] In the formula, For the current moment Normalized Kelvin temperature data, is the temperature data in degrees Celsius, The temperature data is in degrees Fahrenheit.

[0019] Optionally, the average time series difference of the temperature data at the current moment is calculated based on the standardized Kelvin temperature data, and the formula is:

[0020] ,

[0021] In the formula, For the current moment The corresponding average timing difference is For the current moment Normalized Kelvin temperature data, and For the current moment As the starting point, the standardized Kelvin temperature data of the first two historical moments are selected. and For the current moment As the starting point, the standardized Kelvin temperature data of the last two future moments are selected.

[0022] Optionally, the daily cycle difference of the current temperature data is calculated based on the standardized Kelvin temperature data. The formula is:

[0023] ,

[0024] In the formula, For the current moment The daily cycle difference with the temperature data at the same time 24 hours ago, For the current moment Normalized Kelvin temperature data, This is the standardized Kelvin temperature data for the same time twenty-four hours ago.

[0025] Optionally, the average spatial difference of the temperature data at the current moment is calculated based on the standardized Kelvin temperature data, and the formula is:

[0026] ,

[0027] In the formula, For the current moment The corresponding average spatial difference is , , , These are the temperature data collected at the same time by adjacent collection points in the four directions of up, down, left and right of the spatial collection point.

[0028] Optionally, judging whether the temperature data at the current moment is abnormal and locking the abnormal temperature data based on the average time series difference, the daily cycle difference, the average spatial difference and the corresponding threshold range, and calculating the abnormal temperature data deviation value includes:

[0029] According to the average timing difference , daily cycle difference , average spatial difference The threshold range formula is:

[0030] ,

[0031] If the current moment Normalized Kelvin temperature data Corresponding , , If any of the values ​​in exceeds the corresponding threshold range, the temperature data at the current moment is considered abnormal, and the Kelvin temperature data at that moment is marked as abnormal temperature data and recorded, recorded as ;

[0032] In the formula, Represents the average timing difference The threshold value, c represents the daily cycle difference The threshold value, Represents the average spatial difference The threshold value.

[0033] The recorded abnormal temperature data can be further analyzed, processed or repaired as needed.

[0034] Optionally, the abnormal temperature data deviation value is calculated using the formula:

[0035] ,

[0036] In the formula, is the abnormal temperature data deviation value, , and The average time difference , daily cycle difference and the average spatial difference The corresponding weights can be determined using an intelligent algorithm.

[0037] Optionally, the locked abnormal temperature data is repaired based on the locked abnormal temperature data and the abnormal temperature data deviation value to obtain the repaired complete temperature data, and the formula is:

[0038] ,

[0039] In the formula, For the complete temperature data after repair, To lock abnormal temperature data, It is the abnormal temperature data deviation value.

[0040] Beneficial Effects

[0041] The abnormal meteorological temperature data repair method based on spatiotemporal correlation proposed in the present invention combines the time and space correlation of the meteorological data collection process, integrates the characteristics of multiple dimensions such as time series dimension, period dimension and space dimension, realizes the detection and repair of meteorological data, and effectively solves the abnormality problem in the meteorological temperature data collection process. It can provide technical support for timely capture and repair of abnormal data that deviates greatly from normal data during the perception of temperature data, improve the accuracy and reliability of meteorological temperature data, and provide more accurate data support for climate analysis and weather forecasting. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 The figure shows a flow chart of a method for repairing abnormal meteorological temperature data based on spatiotemporal correlation according to the present invention;

[0043] Figure 2 The figure shows the comparison effect between the temperature data and the temperature data after repair. DETAILED DESCRIPTION

[0044] The invention is further described below with reference to the accompanying drawings and specific embodiments.

[0045] Example 1

[0046] This embodiment introduces a method for repairing abnormal meteorological temperature data based on spatiotemporal correlation, including:

[0047] Obtain the temperature data of two centimeters above the surface of the target area as the data processing object;

[0048] Convert the acquired temperature data into Kelvin temperature data for data standardization;

[0049] Based on the standardized Kelvin temperature data, the average time series difference, daily cycle difference and average spatial difference of the temperature data at the current moment are calculated;

[0050] Based on the average time series difference, daily cycle difference and average spatial difference, determine whether the temperature data at the current moment is abnormal, lock the abnormal temperature data and calculate the abnormal temperature data deviation value;

[0051] The locked abnormal temperature data is repaired based on the locked abnormal temperature data and the calculated abnormal temperature data deviation value to obtain the repaired complete temperature data.

[0052] In the actual collection process of meteorological temperature data, the complex and changeable external environment may cause abnormal values ​​in the collected temperature data. The present invention judges whether there is noise or error in the temperature data at the current moment, that is, whether there is an abnormality in the temperature value at the current moment, and locks the abnormal temperature data based on the average time series difference of the historical temperature data at adjacent moments, the daily cycle difference, and the average spatial difference of the adjacent spatial points at the same moment. For the abnormal temperature data existing in the collection process, the three difference parameters of the temperature data at the current moment are combined with the corresponding weights to calculate the abnormal temperature data deviation value. Finally, the abnormal temperature data and the abnormal temperature data deviation value are superimposed through the repair formula to obtain the repaired complete temperature data, which improves the reliability of the judgment results of the meteorological data analysis and prediction model.

[0053] Figure 1 The flowchart is a method for repairing abnormal meteorological temperature data based on spatiotemporal correlation. The flowchart only shows the logical sequence of the method described in this embodiment. In other possible embodiments of the present invention, different methods may be used without conflict. Figure 1 The steps shown or described are accomplished in the order shown.

[0054] Example 2

[0055] On the basis of Example 1, this example further introduces the implementation process of a method for repairing abnormal meteorological temperature data with spatiotemporal correlation, which specifically includes the following contents.

[0056] Step 1, obtain temperature data: In this step, the temperature data value sensed by the temperature sensor is used as the data processing object. Since there are differences in temperature sensors, the temperature data obtained by the temperature sensor are also different, and the obtained temperature data needs to be standardized.

[0057] Step 2, data standardization: Since the temperature data obtained by the sensor may be in various forms, including Celsius (°C) and Fahrenheit (°F), it is necessary to convert the obtained temperature data into Kelvin temperature data for data standardization.

[0058] If the temperature data is obtained in Celsius, it is converted to Kelvin temperature using the following formula:

[0059] (1),

[0060] If the temperature data is obtained in Fahrenheit, it is converted to Kelvin temperature using the following formula:

[0061] (2),

[0062] In the formula, For the current moment Normalized Kelvin temperature data, is the temperature data in degrees Celsius, The temperature data is in degrees Fahrenheit.

[0063] Step 3, calculating the average time series difference, daily cycle difference and average spatial difference, specifically includes the following steps.

[0064] Step 3.1, calculation of average timing difference:

[0065] In this embodiment, the meteorological temperature data is sampled at 1-hour intervals in a time series, and the Kelvin temperature data after standardization at the current time is recorded as , the standardized Kelvin temperature data at the historical moment is recorded as , the standardized Kelvin temperature data at the future time is recorded as There is a smooth transition between the temperature data at the current moment and its historical moment data and future moment data. The temperature will not change suddenly and drastically, but will change gradually over time. The change in temperature is not only affected by the current moment conditions, but is also closely related to the temperature data at its historical moment and future moment. This time dependency allows the present invention to analyze the accuracy of the temperature data at the current moment by combining historical data and future data. The present invention calculates the standardized Kelvin temperature data at the current moment by smoothing the historical and future moments. The average timing difference In practical applications, we need to calculate the standardized Kelvin temperature data at the current moment. The average timing difference in the sequence at each moment. Average timing difference Specifically, it is calculated by formula (3):

[0066] (3),

[0067] In the formula, is the average timing difference, For the current moment Normalized Kelvin temperature data, and For the current moment As the starting point, the standardized Kelvin temperature data of the first two historical moments are selected. and For the current moment As the starting point, the standardized Kelvin temperature data of the last two future moments are selected.

[0068] Step 3.2, daily cycle difference calculation:

[0069] Meteorological temperature data has periodic characteristics, especially daily periodicity, that is, the temperature will experience a process of rising and falling within a day, and this pattern will usually repeat over consecutive days. Therefore, the Kelvin temperature data after standardization at the current moment Normalized Kelvin temperature data at the same time as the previous day (i.e. 24 hours ago) There is often a correlation between them. Although temperature data is cyclical, the correlation between these cycles will gradually weaken over time. For example, the correlation between the temperature data at the current moment and the temperature data at the same moment a few days ago may decrease due to factors such as changes in weather systems and seasonal changes.

[0070] In order to determine the standardized Kelvin temperature data at the current moment Whether it is abnormal, the present invention can judge by calculating the difference between the current temperature data and the temperature data at the same time 24 hours ago. This difference is called the daily cycle difference in the present invention. . Daily cycle difference The size of helps to identify whether the current temperature deviates from the normal daily periodic fluctuation range, so as to determine whether there is an abnormal situation. In practical applications, to calculate the current temperature data The daily cycle difference in the time series. Daily cycle difference It is calculated by the following formula (4):

[0071] (4),

[0072] In the formula, For the current moment Normalized Kelvin temperature data, This is the standardized Kelvin temperature data for the same time 24 hours ago.

[0073] Step 3.3, calculation of average spatial difference:

[0074] In this embodiment, the meteorological temperature data collection points can be standardizedly divided into grids according to their latitude and longitude coordinates. Each grid node represents a specific geographical location, and the meteorological temperature data at this location is obtained through corresponding collection equipment. Comparing temperature data at the same time point but at different grid nodes reveals the spatial distribution characteristics of temperature data. Gridding the spatial representation of surface data is a routine operation in meteorological data processing.

[0075] In this embodiment, a certain collection point m in the space is selected, and the temperature data recorded at the collection point m is the standardized Kelvin temperature data collected at the current moment. The grid distribution of the four collection points in the space of the collection point m is shown. The temperature data of the four collection points in the space of the collection point m are recorded as , , as well as . On this section, you can see the temperature data on different grid nodes. The lateral relationship between these data points reflects the spatial continuity and referenceability of the surface meteorological temperature data. The surface meteorological temperature data in adjacent grid nodes has lateral referenceability, which means that the present invention can use the data of these adjacent nodes to verify and calibrate each other. For example, if the temperature data of a grid node fluctuates abnormally, you can refer to the data of its adjacent nodes to determine whether this fluctuation is real or caused by errors in the acquisition equipment. In actual applications, it is necessary to calculate the standardized Kelvin temperature data at the current moment. The average spatial difference of each acquisition point in the time series. The corresponding average spatial difference It is calculated by the following formula (5):

[0076] (5),

[0077] In the formula, , , as well as The temperature data collected at the same time by the adjacent collection points of the spatial collection point m, the temperature difference between the adjacent points in space, is recorded as the average spatial difference .

[0078] Step 4, locking the abnormal temperature data and calculating the abnormal temperature data deviation value, specifically includes the following steps.

[0079] Step 4.1, lock abnormal temperature data:

[0080] The average timing difference calculated according to step 3 of the present invention is , daily cycle difference , average spatial difference The three aspects are comprehensively considered to lock abnormal temperature data. As long as any of the three differences deviates from the preset deviation range, the current temperature data will be considered abnormal. An exception occurs. Specifically, according to the threshold range formula (6):

[0081] (6),

[0082] In the formula, q represents the average time difference The threshold value, c represents the daily cycle difference The threshold value, Represents the average spatial difference The threshold value of

[0083] If the Kelvin temperature data at the current moment is standardized Does not satisfy any of the above formulas (6), that is, , as well as If any value in exceeds its corresponding deviation range, the temperature data at the current moment is considered abnormal, and the temperature data at that moment is marked as abnormal temperature data and recorded, which is recorded as In step 5, the abnormal temperature data are further analyzed, processed or repaired as needed.

[0084] Above, the average timing difference Threshold The variance of the temperature data calculated based on the time dimension data is calculated by using the temperature data of multiple consecutive moments to form a data sequence and calculating the variance value of the entire data sequence. Threshold The variance of the temperature data calculated based on the daily cycle dimension data is calculated by using the temperature data at the same time of multiple consecutive days to form a data sequence and calculating the variance value of the data sequence. Average spatial difference Threshold It is the mean of the temperature data variance calculated based on the spatial dimension data. The calculation method is as follows: for multiple points in the space, the temperature data variance of each point and its surrounding points is first obtained, and then the mean of the temperature data variance is obtained for multiple points.

[0085] Step 4.2, calculate the abnormal temperature data deviation value:

[0086] Since the deviation adjustment of abnormal temperature data needs to ensure compliance with the changing rules of time series dimension, period dimension and space dimension, the average time series difference should be used for fine-tuning. , daily cycle difference and the average spatial difference Determine the final adjustment parameter value. The abnormal temperature data deviation value is calculated by the following formula (7):

[0087] (7),

[0088] In the formula, is the abnormal temperature data deviation value, , and The average time difference , daily cycle difference and the average spatial difference The difference parameters cannot be simply processed by arithmetic averaging, because the amount of information and importance they carry at different time points are different. In order to accurately capture the dynamic changes of these difference parameters and obtain reasonable abnormal temperature data deviation values, different weights need to be assigned to them. These weights should be determined based on the time characteristics and impact of each data. Therefore, the weights in this embodiment are , and Determined according to the GA-PSO intelligent algorithm.

[0089] When solving single-objective optimization problems such as determining weights, both Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) have shown excellent performance. GA can effectively explore the solution space and maintain the diversity of the population by simulating the selection, crossover and mutation operations in the biological evolution process; while PSO can quickly converge to a better solution through information sharing and iterative update mechanism between particles. In order to combine the advantages of these two algorithms, a GA-PSO hybrid algorithm is designed. In terms of specific technical implementation, the GA-PSO algorithm regularly introduces the selection, crossover and mutation operations of GA on the basis of PSO. Specifically including:

[0090] (1) Initialization phase: First, the particle swarm is initialized according to the characteristics of the problem, including the positions and velocities of the particles, which represent potential solutions.

[0091] (2) Fitness evaluation: Calculate the fitness value of each particle based on the average time series difference of the above three difference parameters , daily cycle difference and the average spatial difference Weighted sum, where the weights are dynamically adjusted based on the importance of the above three difference parameters at a specific time point;

[0092] (3) PSO update mechanism: particles update their speed and position according to their own historical best position and global best position to achieve convergence to a better solution;

[0093] (4) Introduction of GA operation: At a specific moment in the PSO iteration process, every 10 generations, the particle swarm is subjected to GA selection, crossover and mutation operations. The selection operation retains excellent particles, the crossover operation generates new solutions, and the mutation operation introduces randomness to increase the exploration range of the solution space;

[0094] (5) Iteration and termination: Repeat the above steps until the preset number of iterations is reached, and finally the optimal influence weight values, i.e., λ, β, and γ, are calculated.

[0095] Through this GA-PSO hybrid algorithm, we can fully utilize the advantages of GA in exploring the solution space and maintaining diversity, as well as the ability of PSO in fast convergence, so as to more effectively solve the fine-tuning optimization problem and obtain more accurate and stable abnormal deviation values.

[0096] Step 5, abnormal temperature data repair: According to the weight value determined in the above step, the abnormal temperature data deviation value P of the true value can be calculated. The abnormal temperature data deviation value P is compared with the abnormal temperature data E n By superimposing, the repaired value can be obtained, and the original data can be filled in, and the complete temperature data after repair can be obtained through the following formula (8):

[0097] (8),

[0098] Where T is the complete temperature data after repair, is the abnormal temperature data locked in step 4.1, and P is the abnormal temperature data deviation value determined in step 4.2.

[0099] like Figure 2 The figure shows the comparison between the ground true temperature data and the temperature data before and after the repair process. MSE is an indicator to measure the difference or error of data. Compared with the noise data, the MSE value of the repaired data has been significantly reduced, such as the MSE value of 12:00:00 has dropped from 15.99 to 6.24, which shows that the abnormal meteorological temperature data repair method of the present invention can effectively remove noise, make the temperature data closer to the true value, and improve the accuracy.

[0100] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the purpose of the present invention and the claims, which all fall within the protection of the present invention.

Claims

1. A method for repairing abnormal meteorological temperature data based on spatiotemporal correlation, characterized in that: include: Acquire the surface temperature data of the target area as the data processing object; Convert the acquired temperature data into Kelvin temperature data and perform data standardization; Based on the standardized Kelvin temperature data, the average time series difference, daily cycle difference and average spatial difference of the temperature data at the current moment are calculated; Based on the average time series difference, daily cycle difference, average spatial difference and corresponding threshold range, determine whether the temperature data at the current moment is abnormal, lock the abnormal temperature data, and calculate the abnormal temperature data deviation value; The locked abnormal temperature data is repaired based on the locked abnormal temperature data and the abnormal temperature data deviation value to obtain the repaired complete temperature data.

2. The abnormal meteorological temperature data repair method according to claim 1 is characterized in that: The step of converting the acquired temperature data into Kelvin temperature data and performing data standardization includes: If the temperature data is obtained in Celsius, the formula for converting it to Kelvin temperature data and normalizing the data is: ， If the temperature data is obtained in Fahrenheit, the formula for converting it to Kelvin temperature data and normalizing the data is: ， In the formula, For the current moment Normalized Kelvin temperature data, is the temperature data in degrees Celsius, The temperature data is in degrees Fahrenheit.

3. The abnormal meteorological temperature data repair method according to claim 1 is characterized in that: The average time series difference of the temperature data at the current moment is calculated based on the standardized Kelvin temperature data, and the formula is: ， In the formula, For the current moment The corresponding average timing difference is For the current moment Normalized Kelvin temperature data, and For the current moment As the starting point, the standardized Kelvin temperature data of the first two historical moments are selected. and For the current moment As the starting point, the standardized Kelvin temperature data of the last two future moments are selected.

4. The abnormal meteorological temperature data repair method according to claim 1 is characterized in that: Based on the standardized Kelvin temperature data, the daily cycle difference of the current temperature data is calculated using the formula: ， In the formula, For the current moment The daily cycle difference with the temperature data at the same time 24 hours ago, For the current moment Normalized Kelvin temperature data, This is the standardized Kelvin temperature data for the same time twenty-four hours ago.

5. The abnormal meteorological temperature data repair method according to claim 1 is characterized in that: Based on the standardized Kelvin temperature data, the average spatial difference of the temperature data at the current moment is calculated using the formula: ， In the formula, For the current moment The corresponding average spatial difference is , , , They are the temperature data collected at the same time by adjacent collection points in the four directions of up, down, left and right of the spatial collection point.

6. The abnormal meteorological temperature data repair method according to any one of claims 3 to 5, characterized in that: The method of judging whether the temperature data at the current moment is abnormal based on the average time series difference, the daily cycle difference, the average spatial difference and the corresponding threshold range, locking the abnormal temperature data and calculating the abnormal temperature data deviation value includes: According to the average timing difference , daily cycle difference , average spatial difference The threshold range formula is: ， If the current moment Normalized Kelvin temperature data Corresponding , , If any of the values ​​in exceeds the corresponding threshold range, the temperature data at the current moment is considered abnormal, and the Kelvin temperature data at that moment is marked as abnormal temperature data and recorded, recorded as ; In the formula, Represents the average timing difference The threshold value, c represents the daily cycle difference The threshold value, Represents the average spatial difference The threshold value.

7. The abnormal meteorological temperature data repair method according to claim 6 is characterized in that: The abnormal temperature data deviation value is calculated using the formula: ， In the formula, is the abnormal temperature data deviation value, , and The average time difference , daily cycle difference and the average spatial difference The corresponding weight.

8. The abnormal meteorological temperature data repair method according to claim 7 is characterized in that: The locked abnormal temperature data is repaired based on the locked abnormal temperature data and the abnormal temperature data deviation value to obtain the repaired complete temperature data, and the formula is: ， Where T is the complete temperature data after repair, is the locked abnormal temperature data, and P is the abnormal temperature data deviation value.

Images