Agricultural meteorological data processing method and system

Abstract

The invention discloses an agricultural meteorological data processing method and system, and relates to the technical field of agricultural meteorological data processing. The method comprises the following steps: acquiring agricultural meteorological time series data in a monitoring period T for fitting, acquiring a fitting value of ith type of agricultural meteorological data acquired by jth sensing equipment at a moment t, calculating a data deviation degree, screening out data of which the data deviation degree is not greater than a deviation degree threshold value, and calculating the data deviation degree of the ith type of agricultural meteorological data; describing the change trend of the operation reliability of the sensing equipment by using a Weibull distribution model, screening out data collected by the sensing equipment of which the operation reliability evaluation value is not less than a reliability threshold in the initial set, updating the initial set according to a data screening result, constructing a data iteration set, and calculating the operation reliability of the sensing equipment; and taking a normalized value of an operation reliability evaluation value of the sensing equipment corresponding to the k-th data in the iteration set as a data confidence coefficient, performing weighting processing on all the data in the iteration set, and taking the obtained weighted data as a sampling value of the i-th type of agricultural meteorological data at the t moment in the monitoring period T.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural meteorological data processing, and particularly relates to a method and system for processing agricultural meteorological data. Background Art

[0002] Agricultural meteorological data refers to meteorological observation data related to agricultural production, including meteorological elements such as temperature, humidity, precipitation, wind speed, and light intensity. These data play an extremely important role in aspects such as agricultural production, agricultural scientific research, resource utilization and disaster prevention, and agricultural policy formulation. By analyzing meteorological data, the optimal planting time and growth cycle of different crops can be determined to support precision agriculture; natural disasters such as droughts, floods, frosts, and heavy rains can be predicted; it can also be used to study the influence mechanism of meteorological factors on the growth, development, yield, and quality of crops, and accurate meteorological data is the basis for establishing an agricultural meteorological model; by real-time monitoring and processing meteorological data, early signs of meteorological disasters can be detected in a timely manner to improve the accuracy of disaster warnings.

[0003] In the process of agricultural meteorological data processing, it is necessary to organize the collected multi-source heterogeneous data into a data set according to regionality. The update and maintenance of the agricultural meteorological data set need to be in a dynamic and continuous state all the time, and the collection of various agricultural meteorological data depends on sensing devices. For example, temperature and humidity are obtained using temperature and humidity sensors, and wind speed is obtained using a wind speed sensing device. During the long-term use process, the reliability of electronic devices shows a decreasing trend with the change of use time. Therefore, it is difficult to ensure the accuracy of the obtained meteorological observation data during the operation and maintenance of the data set. For this reason, we propose a method and system for processing agricultural meteorological data. Summary of the Invention

[0004] The main purpose of the present invention is to provide a method and system for processing agricultural meteorological data, which can effectively solve the problems in the background art.

[0005] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0006] A method for processing agricultural meteorological data includes:

[0007] Step 1: Collect agricultural meteorological time series data within a monitoring period T, and use the collected time series data to construct an initial set of the i-th type of agricultural meteorological data Wherein, represents the i-th type of agricultural meteorological data collected by the m-th sensing device at time t; t ∈ T;

[0008] Step 2: Fit the time series data collected by the j-th sensing device to obtain the fitted value of the i-th type of agricultural meteorological data collected by the j-th sensing device at time t

[0009] Step 3: Calculate the initial set X i Among them, the i-th type of agricultural meteorological data collected by the j-th sensing device at time t Data deviation degree Set the data deviation degree threshold Screen out the initial set X i Among them, the data deviation degree of the data;

[0010] Step 4: Use the Weibull distribution model to describe the change trend of the operating reliability of the j-th sensing device during the monitoring period T, and obtain the operating reliability evaluation value R j (t) of the j-th sensing device at time t, and set the reliability threshold R(t) of the sensing device min , screen out the initial set X i Among them, the operating reliability evaluation value R j (t) ≥ R(t) min of the data collected by the sensing device;

[0011] Step 5: Update the initial set X according to the data screening result i , construct the iterative set X' of the i-th type of agricultural meteorological data at time t i , and use the normalized value of the operating reliability evaluation value of the sensing device corresponding to the k-th data in the iterative set X' i as the data confidence level θ k , perform weighted processing on all the data at time t in the iterative set X' i to obtain the weighted data as the sampling value of the i-th type of agricultural meteorological data at time t in the target area during the monitoring period T.

[0012] An agricultural meteorological data processing system, comprising:

[0013] A data acquisition module, configured to acquire the agricultural meteorological time series data in the target area during the monitoring period T, and construct the initial set of the i-th type of agricultural meteorological data by using the acquired time series data Among them, is expressed as the i-th type of agricultural meteorological data collected by the m-th sensing device at time t; t ∈ T; the agricultural meteorological data includes at least one of temperature, humidity, precipitation, wind speed, and light intensity;

[0014] A data fitting module, configured to fit the time series data collected by the j-th sensing device to obtain the fitting value of the i-th type of agricultural meteorological data collected by the j-th sensing device at time t Among them, j = 1, 2,..., m;

[0015] The first data screening module is used to calculate the initial set Xi Among them, the i-th type of agricultural meteorological data collected by the j-th sensing device at time t data deviation degree and set the data deviation degree threshold Screen out the initial set X i Among them, the data deviation degree of the data;

[0016] The second data screening module is used to describe the change trend of the operation reliability of the j-th sensing device during the monitoring period T by using the Weibull distribution model, and obtain the operation reliability evaluation value R j (t) of the j-th sensing device at time t, and set the reliability threshold R(t) of the sensing device min , screen out the initial set X i Among them, the operation reliability evaluation value R j (t) ≥ R(t) min data collected by the sensing device;

[0017] The dataset update module is used to update the initial set X according to the data screening result i , and construct the iterative set X' of the i-th type of agricultural meteorological data at time t i ;

[0018] The data processing module is used to use the normalized value of the operation reliability evaluation value of the sensing device corresponding to the k-th data in the iterative set X' i as the data confidence level θ k , perform weighted processing on all data at time t in the iterative set X' i to obtain the weighted data as the sampling value of the i-th type of agricultural meteorological data at time t in the target area during the monitoring period T.

[0019] The system includes a memory, a processor, and a computer program stored on the memory and executable on the processor.

[0020] Furthermore, the calculation method of the fitting value is:

[0021]

[0022] In the formula, represents the i-th type of agricultural meteorological data collected by the j-th sensing device at the t-r moment during the monitoring period T; is a constant coefficient, and

[0023] Furthermore, the calculation formula of the data deviation degree is:

[0024] Further, the expression of the Weibull distribution model for describing the reliability change trend of the j-th sensing device at time t is:

[0025]

[0026] In the formula, e is the base of the natural logarithm; η is the characteristic life, representing the average life of the sensing device; β is the shape parameter, used to describe the change trend of the failure rate.

[0027] Further, the data confidence level θ k The calculation formula is:

[0028]

[0029] In the formula, R k (t) represents the operation reliability evaluation value of the sensing device corresponding to the k-th data in the iteration set X'. i

[0030] Further, the weighted data The calculation formula is:

[0031]

[0032] In the formula, x ik represents the k-th data in the iteration set X'. i

[0033] Further, the agricultural meteorological data includes at least one of temperature, humidity, precipitation, wind speed, and light intensity.

[0034] The present invention has the following beneficial effects.

[0035] Compared with the prior art, by collecting the agricultural meteorological time-series data in the target area during the monitoring period T, using the collected time-series data to construct the initial set X of the i-th type of agricultural meteorological data i , fitting the time-series data collected by the j-th sensing device, obtaining the fitting value of the i-th type of agricultural meteorological data collected by the j-th sensing device at time t, calculating the data deviation degree i in the initial set X of the i-th type of agricultural meteorological data collected by the j-th sensing device at time t setting the data deviation degree threshold screening out the data in the initial set X i with a data deviation degree using the Weibull distribution model to describe the operation reliability change trend of the j-th sensing device during the monitoring period T, obtaining the operation reliability evaluation value R j (t) of the j-th sensing device at time t, setting the reliability threshold R(t) of the sensing devicemin , screen out the initial set X i Among them, the operation reliability evaluation value R j (t) ≥ R(t) min The data collected by the sensing devices, and update the initial set X according to the data screening results i , construct the iterative set X' of the i-th type of agricultural meteorological data at time t i , with the iterative set X' i The normalization value of the operation reliability evaluation value of the k-th data corresponding sensing device in it is used as the data confidence θ k , perform weighted processing on all the data at time t in the iterative set X' i to obtain the weighted data As the sampling value of the i-th type of agricultural meteorological data at time t in the target area during the monitoring period T, thus effectively solving the problem of inaccurate agricultural meteorological data caused by the reliability change of electronic devices during the long-term use process. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 is a schematic flow chart of a method for processing agricultural meteorological data according to the present invention;

[0037] Figure 2 is a schematic structural diagram of a system for processing agricultural meteorological data according to the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0038] The following further describes the present invention in conjunction with specific embodiments. Among them, the drawings are only used for illustrative purposes, showing only schematic diagrams, not physical diagrams, and should not be construed as limiting the present invention. In order to better illustrate the specific embodiments of the present invention, some components in the drawings will be omitted, enlarged or reduced, and do not represent the dimensions of actual products.

[0039] The specific implementation process of the technical solution of the present invention includes the following steps:

[0040] Step 1: Collect the agricultural meteorological time series data in the target area during the monitoring period T, and use the collected time series data to construct the initial set of the i-th type of agricultural meteorological data Among them, represents the i-th type of agricultural meteorological data collected by the m-th sensing device at time t; t ∈ T;

[0041] Among them, the agricultural meteorological data includes at least one of temperature, humidity, precipitation, wind speed, and light intensity. In this embodiment, the collected temperature data is taken as an example for illustration. Then the sensing device is a temperature sensor, and the initial set of agricultural meteorological data constructed with the collected temperature data can be denoted as denoted as the temperature value collected by the m-th temperature sensor at time t, initial set can be denoted as denoted as the temperature value collected by the m-th temperature sensor at time t + ε.

[0042] Step 2: Fit the time-series data collected by the j-th temperature sensor to obtain the fitted value of the temperature data collected by the j-th temperature sensor at time t where the fitted value is calculated as follows:

[0043]

[0044] In the formula, denotes the temperature value collected by the j-th temperature sensor at time t - r within the monitoring period T; is a constant coefficient, and

[0045] It should be noted that since the temperature is in dynamic change, therefore, for the value of ω, it is also in dynamic change. When fitting the time-series data collected by the j-th temperature sensor, the constant coefficient ω in the fitted value calculation formula should be determined first. For the value of the constant coefficient ω, the following method can be used to determine it:

[0046] a: Empirical judgment method

[0047] Determine the specific value of the constant coefficient ω according to the fluctuation characteristics and trend of the collected temperature time-series data:

[0048] Data fluctuation is stable: When the collected temperature time-series data shows a relatively stable horizontal trend, a smaller value should be selected, generally between 0.05 and 0.20;

[0049] Data has fluctuations but the trend change is not significant: When the collected temperature time-series data has fluctuations but the long-term trend change is not significant, a slightly larger value can be selected, usually between 0.1 and 0.4;

[0050] Data fluctuation is large and the trend change is obvious: When the collected temperature time-series data fluctuates greatly, the long-term trend change amplitude is large, showing an obvious and rapid upward or downward trend, a larger value should be selected, such as a value between 0.6 and 0.8 can be selected.

[0051] Specifically, the evaluation of the data fluctuation degree can be based on the temperature standard deviation of the temperature data collected by the j-th temperature sensor in the data sequence. It should be noted that since the data fitting is to obtain the predicted value, i.e., the fitting value, of the temperature data collected by the j-th temperature sensor at time t. Therefore, the calculated temperature standard deviation corresponds to the standard deviation of the temperature data collected by the j-th temperature sensor during the time period from the initial moment to time t within the monitoring period T. The calculation formula is as follows: In the formula, represents the temperature data collected by the j-th temperature sensor at time q, and q ∈ T, q < t; N represents the amount of temperature data collected by the j-th temperature sensor during the period from q = 1 to q = t; μ represents the mean value of the temperature data collected by the j-th temperature sensor during the period from q = 1 to q = t. Set the standard deviation level threshold to evaluate the fluctuation degree. For example, set the temperature standard deviation level thresholds to φ1 and φ2 respectively. When the calculated temperature standard deviation is less than φ1, it is determined that the data fluctuation state is stable; when the calculated temperature standard deviation is between φ1 and φ2, it is determined that the data fluctuation state has fluctuations but the trend change is small; when the calculated temperature standard deviation is greater than φ2, it is determined that the data fluctuation state has large fluctuations and the trend change is obvious.

[0052] b: Trial algorithm

[0053] According to the specific time series situation, select several different ω values for trial calculation, compare the prediction standard errors under different ω values, and select the ω value with the smallest prediction standard error.

[0054] c: Method of minimizing prediction error

[0055] It is determined by minimizing the prediction error (such as the mean square error MSE).

[0056] The specific steps are as follows:

[0057] Select different ω values (such as from ω = 0.1 to ω = 0.9, with a step size of 0.1);

[0058] For each ω value, calculate the error between the predicted value and the actual value. The calculation formula for the error is: (|predicted value - actual value|) / actual value × 100% or (|actual value - predicted value|) / actual value × 100%;

[0059] Select the ω value that minimizes the prediction error.

[0060] Step 3: Calculate the data deviation degree of the temperature data collected by the j-th temperature sensor at time t in the initial set X of the data The data deviation degree The calculation formula is:

[0061] Step 4: Set the data deviation threshold Screen out the data with data deviation in the initial set X.

[0062] It should be noted that for the value of the data deviation threshold , the average value method can be used. Calculate the data deviation at any moment respectively, and then calculate the average value of all calculated deviation values to obtain the average value as the data deviation threshold

[0063] For the data with data deviation , it indicates that there is a large deviation between its sampling value and the fitting value. For agricultural meteorological data, within the same observation time period, the observed data fluctuates up and down within a stable range, and the change trend of the data is relatively stable. Therefore, when the data deviation is large, it means that the possibility of this data being abnormal data is greater and needs to be screened out.

[0064] Step 5: Use the Weibull distribution model to describe the change trend of the operating reliability of the jth temperature sensor during the monitoring period T, and obtain the operating reliability evaluation value R j (t) of the jth temperature sensor at time t; where the expression of the Weibull distribution model is:

[0065]

[0066] In the formula, e is the base of the natural logarithm; η is the characteristic life, representing the average life of the temperature sensor; β is the shape parameter, used to describe the change trend of the failure rate.

[0067] It should be noted that for the values of the characteristic life η and the shape parameter β, the maximum likelihood estimation method can be used to determine them. By constructing a likelihood function, substituting the observed data into the probability density function of the Weibull distribution, and then maximizing this likelihood function to solve for the characteristic life η and the shape parameter β. The specific process is as follows:

[0068] Step S51: Obtain a complete sample data with a capacity of n, t1 ≤ t2 ≤... ≤ t n , where t1, t2,..., t n represent a set of observed life data of the temperature sensor. t1 represents the first observed object, that is, the service life or failure time of the first temperature sensor; t2 represents the second observed object, that is, the service life or failure time of the second temperature sensor; and so on, t n represents the nth observed object, that is, the service life or failure time of the nth temperature sensor; arrange the obtained observed data in ascending order;

[0069] Step S52: Obtain the log-likelihood function of the Weibull distribution for the complete sample data, where the function expression is:

[0070] Step S53: Take the partial derivatives of the log-likelihood function with respect to each parameter to obtain the following system of equations:

[0071]

[0072]

[0073] Step S54: By solving the above system of equations, the estimated values of the characteristic life η and the shape parameter β can be obtained.

[0074] Step 6: Set the reliability threshold R(t) of the temperature sensor min , and screen out the initial set X i in which the running reliability evaluation value R j (t) ≥ R(t) min for the data collected by the temperature sensor.

[0075] It should be noted that for the value of the reliability threshold R(t) min , the average value method can also be used to determine it. The specific method is the same as the method for determining the value of the data deviation degree threshold in Step 4, which will not be elaborated here.

[0076] In addition, it should also be noted that for the data collected by the temperature sensor with the running reliability evaluation value R j (t) < R(t) min , since the running reliability of its sensing device does not meet the set threshold, it indicates that the confidence level of the data it obtains is relatively low and the probability of error increases. Therefore, it is necessary to screen out the data collected by this temperature sensor to improve the accuracy of the collected data.

[0077] Step 7: Update the initial set X according to the data screening result, and construct the temperature data iteration set X' of the agricultural meteorological data at time t.

[0078] Step 9: Use the normalized value of the running reliability evaluation value of the k-th data in the iteration set X' corresponding to the temperature sensor as the data confidence level θ k , and when the running reliability evaluation value of the temperature sensor is larger, it indicates that the confidence level of the accuracy of the data it collects is higher. Among them, the calculation formula of the data confidence level θ k is: In the formula, R k (t) represents the running reliability evaluation value of the k-th data in the iteration set X' corresponding to the temperature sensor.

[0079] Step 8: Perform weighted processing on all data at time t in the iteration set X' to obtain weighted data as the sampling value of the temperature data at time t within the monitoring period T in the target area. Among them, the weighted data is calculated by the formula: In the formula, x ik represents the k-th data at time t in the iteration set X' i Note that the deviation threshold

[0080] set in Step 4 and the reliability threshold R(t) set in Step 6 min determine the amount of temperature data at time t in the temperature data iteration set X'. When the set deviation threshold is too small or the reliability threshold R(t) min is too large, it may cause the amount of temperature data at time t in the temperature data iteration set X' to be 0. Therefore, in the actual operation process, the set thresholds need to be adjusted according to the actual situation to ensure that the temperature data iteration set X' is not an empty set.

[0081] For other types of agricultural meteorological data, including humidity, precipitation, wind speed, and light intensity, the processing flow is the same as Steps 1 - 9 above and will not be elaborated here. The technical solution proposed by the present invention, by adopting redundant sensor technology and setting multiple sensors of the same type to collect corresponding types of agricultural meteorological data, can effectively reduce data errors caused by single sensor failures; in addition, by comparing and processing the data collected by the same sensor, filtering out the collected data that may be in an abnormal state at a single moment, and evaluating the operational reliability of the sensor, filtering out the data collected by sensors with insufficient operational reliability; then, by performing weighted processing on the remaining data, the spatial and temporal resolution of the collected agricultural meteorological data can be improved, further reducing data errors caused by equipment aging, thereby improving the accuracy of the collected data, providing good preconditions for the practical applications of agricultural meteorological data such as agricultural production, agricultural scientific research, resource utilization, disaster prevention, and agricultural policy formulation.

[0082] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for processing agricultural meteorological data, characterized in that, Including: Step 1: Collect the agricultural meteorological time-series data of the target area within the monitoring period T, and construct the initial set of the i-th type of agricultural meteorological data using the collected time-series data wherein represents the i-th type of agricultural meteorological data collected by the m-th sensing device at time t; t ∈ T; Step 2: Fit the time-series data collected by the j-th sensing device to obtain the fitted value of the i-th type of agricultural meteorological data collected by the j-th sensing device at time t Step 3: Calculate the initial set X i In, the i-th type of agricultural meteorological data collected by the j-th sensing device at time t Data deviation degree Set the data deviation degree threshold Screen out the initial set X i In, the data deviation degree Of the data; Step 4: Use the Weibull distribution model to describe the changing trend of the operating reliability of the j-th sensing device within the monitoring period T, and obtain the operating reliability evaluation value R of the j-th sensing device at time t j (t), and set the reliability threshold R(t) of the sensing device min , and screen out the initial set X i Among them, the operating reliability evaluation value R j (t) ≥ R(t) min of the sensing device; Step 5: Update the initial set X according to the data screening results i , and construct the iterative set X' of the i-th type of agricultural meteorological data at time t i . Using the iterative set X' i , take the normalized value of the operation reliability evaluation value of the sensing device corresponding to the k-th data in X' as the data confidence θ k . Perform weighted processing on all the data at time t in the iterative set X' i to obtain the weighted data as the sampling value of the i-th type of agricultural meteorological data at time t in the target area within the monitoring period T 2. The agricultural meteorological data processing method according to claim 1, characterized in that Fitted value The calculation method is as follows: In the formula, represents the i-th type of agricultural meteorological data collected by the j-th sensing device at the t-r moment within the monitoring period T; is a constant coefficient, and 3. A method for processing agricultural meteorological data according to claim 1, characterized in that, Data deviation degree The calculation formula is as follows:

4. The agricultural meteorological data processing method according to claim 1, wherein The expression of the Weibull distribution model for describing the reliability change trend of the j-th sensing device at time t is: In the formula, e is the base of the natural logarithm; η is the characteristic life, representing the average life of the sensing device; β is the shape parameter, used to describe the change trend of the failure rate.

5. A method for processing agricultural meteorological data according to claim 1, characterized in that Data confidence level θ k The calculation formula is as follows: where R k (t) represents the operational reliability evaluation value of the k-th data corresponding to the sensing device in the iterative set X' i .

6. The agricultural meteorological data processing method according to claim 1, characterized in that Weighted data The calculation formula is as follows: where x ik represents the k-th data in the iterative set X' i .

7. A method for processing agricultural meteorological data according to claim 1, characterized in that, The agricultural meteorological data includes at least one of temperature, humidity, precipitation, wind speed, and light intensity.

8. An agricultural meteorological data processing system, characterized in that, The system is used to implement the steps of an agricultural meteorological data processing method described in any one of claims 1-7, including: A data acquisition module is used to collect the agricultural meteorological time-series data of the target area within the monitoring period T, and construct the initial set of the i-th type of agricultural meteorological data by using the collected time-series data. Among them, It is expressed as the i-th type of agricultural meteorological data collected by the m-th sensing device at time t; t ∈ T; the agricultural meteorological data includes at least one of temperature, humidity, precipitation, wind speed, and light intensity. A data fitting module, which is used to fit the time series data collected by the j-th sensing device to obtain the fitting value of the i-th type of agricultural meteorological data collected by the j-th sensing device at time t where j = 1, 2,..., m; The first data screening module is used to calculate the initial set X i In the data deviation degree of the i-th type of agricultural meteorological data collected by the j-th sensing device at time t And set the data deviation degree threshold Screen out the initial set X In, the data deviation degree i Of the data ; The second data screening module is used to describe the changing trend of the operation reliability of the j-th sensing device within the monitoring period T by using the Weibull distribution model, and obtain the operation reliability evaluation value R of the j-th sensing device at time t j (t), and set the reliability threshold R(t) of the sensing device min , and screen out the initial set X i in which the operation reliability evaluation value R j (t) ≥ R(t) min of the sensing device; A dataset update module for updating the initial set X according to the data screening results i , and constructing the iterative set X' of the i-th type of agricultural meteorological data at time t i ; A data processing module is used to use the normalization value of the running reliability evaluation value of the sensing device corresponding to the k-th data in the iterative set X' i as the data confidence θ, and perform weighted processing on all the data at time t in the iterative set X' k to obtain the weighted data i as the sampling value of the i-th type of agricultural meteorological data at time t within the monitoring period T in the target area. ​ 9. The agricultural meteorological data processing system according to claim 8, wherein The system includes a memory, a processor, and a computer program stored on the memory and executable on the processor. Among them, when the processor executes the program, it can implement the steps of an agricultural meteorological data processing method described in any one of claims 1-7.

Images