Systems and methods for forecasting annual rice yields using mathematical models.

Abstract

OCR 09WP 22 / 01 / 2569 Systems and methods for forecasting annual rice yields using mathematical models, which include various units. Input units include mathematical modeling units, production forecasting units, and display units, among others. Data preparation, specifically the preparation of rice production data without existing rice production information, by preparing the data as follows: Verify the data on newly planted areas, which lack information on average yield per rai each year. Specify the average yield per rai data. Each year, using box plotting techniques and preparing weather data that includes rainfall data, Temperature and the National Diversity Index (NDVI) are prepared by finding the maximum, minimum, and average values ​​for each area. Each province for each month is included in every index, and weather data, including humidity data, is prepared. By taking measurements from telemetric stations with and without data, where measurements are taken from telemetric stations without data. Moisture content data was prepared by identifying representative points for each province and building a model to predict rice production data. Mathematical modeling unit using multiple linear regression analysis.

Claims

OCR 09WP 22 / 01 / 2569 1. A system for forecasting rice yields using a mathematical model that includes: Data input unit (100) which imports weather data (11) and rice production data (12). The mathematical modeling unit (300) is connected to the input unit (100) and brings the conditional data. Weather (11) and rice production data (12) were used to create a model for forecasting rice production (12). The production forecasting unit (400), which connects the production forecasting modeling unit (300), and Rice production forecast (12) using a mathematical model. The display unit (500) is connected to the production forecast unit (400) and displays the production forecast results. Rice (12) Its unique characteristic is the system for forecasting rice yields using the aforementioned mathematical model. Additionally, there is a data preparation unit (200) which is connected to the input unit (100) and the data preparation unit. (200) The above was carried out in preparing rice production data (12) and preparing weather data (11). Where the preparation of rice production data (12) with production data and without production data, where For projects without rice production data, prepare the data by verifying information on newly planted area that lacks production data. The average yield per rai each year is indicated using box plotting techniques. Check the average yield per acre for each year outside the adjacent values ​​(301) and The values ​​that are extreme outliers (302) are converted to adjacent values ​​(301). Instead, values ​​greater than the 100th percentile will be changed to the 100th percentile, and values ​​lower than that will be changed to the 100th percentile. The 0th percentile will be changed to the 0th percentile. Where the preparation of weather data (11) which includes rainfall data, temperature and The Normalized Difference Vegetation Index (NDVI) (903) was prepared with Finding the maximum, minimum, and average values ​​for each province in each month for all indices. Where the preparation of weather data (11) which includes humidity data is prepared by measurement. Values ​​from telemetry stations with and without data, where measurements are taken from telemetry stations without data. Humidity data was prepared by finding representative points for each province (501), which consisted of the center point. Determine the latitude (equation) and longitude (equation) center points of each province and find the humidity values. Representing the province from the nearest telemetry stations (502) at least 3 stations by finding the weighted average. The weight is based on the distance from the telemetric station. Where the weather data (11) that affects the rice production data (12) will be divided into: If regional data is unavailable, national data will be used instead. Where the mathematical modeling unit (300) built a model to predict rice production data (12). Multiple Linear Regression Analysis Method using the following factors: The maximum, minimum, and average values ​​of the humidity data from each telemetry station.

2. The system for forecasting annual rice yields using a mathematical model, based on Claim 1, where the data input unit... (100) Import rice production data (12) selected from the production of Jasmine rice, White rice, Pathum Thani 1, Aromatic rice. Province, colored rice, or glutinous rice.

3. A system for forecasting annual rice yield using a mathematical model based on claim 1 or 2, where the point values... The provincial representative (501) defines (x1, y1), (x2, y2), ..., (xn, yn) as the boundaries of the province. Find the vertices and the center of latitude (equation) and the center of longitude (equation) from... The relationship between provincial boundaries and area. (Equation) where (The equation) represents the center of latitude, or the y-axis. (The equation) represents the center of longitude, or the x-axis. A represents the area of ​​the province. n is the total number of vertices. and find the moisture value that is representative of the province from the nearest telemetric station (502) from the weighted average. Based on the distances to the nearest three telemetric stations, as shown in the equation: (Equation) where (The equation) is a weighted average based on the distances from the three nearest telemetric stations. da, db, and dc represent the distances between the centroid and telemetry stations a and b. and c xa, xp, xc are the measured values ​​(highest, lowest, or average for each month) of a telephone station. Measures a, b, and c 4. A system for forecasting rice yield using a mathematical model based on any one of claims 1 to 3. It is further comprised of a database (142) which is connected to the mathematical modeling unit (300) and Store yield prediction data in kilograms per rai and application interface documentation. (Application Program Interface: API) (143) which connects to the database (142) and retrieves the results. Predict the output stored in the database (142) and display it back to the map display application. (15) or display on the user's device through the identification of the latitude and longitude coordinates of the area.

5. A system for forecasting annual rice yields using a mathematical model based on any one of claims 1 through 4. It is further comprised of a production comparison unit, which is connected to the production forecasting unit (400) and leads. The predicted output is compared to the actual output of the previous year, selecting those with lower output than the previous year. 25 kg / rai, yield less than last year; 0-25 kg / rai, yield the same as last year; yield more than last year; 0- 25 kg / rai, yield is 25 kg / rai higher than last year.

6. Methods for forecasting annual rice yields using mathematical models that include: The data import procedure, which imports weather data (11) and rice production data (12) through the unit. Import data (100) The process of creating a mathematical model which takes weather data (11) and rice production data (12). Let's create a model for forecasting rice production (12) through the mathematical modeling unit (300). The production forecasting procedure, which forecasts the rice production (12) using a mathematical model through units. Production forecast (400) Display procedure, which shows the rice production forecast (12) through the display unit (500). What makes it unique is the method of forecasting rice yields using the aforementioned mathematical model. Additionally, there is a data preparation step, which is connected to the data import and preparation steps. The above information was used to prepare rice production data (12) and weather data (11). Where the preparation of rice production data (12) with production data and without production data, where For projects without rice production data, prepare the data by verifying information on newly planted area that lacks production data. The average yield per rai each year is indicated using box plotting techniques. Check the average yield per acre for each year outside the adjacent values ​​(301) and The values ​​that are extreme outliers (302) are converted to adjacent values ​​(301). Instead, values ​​greater than the 100th percentile will be changed to the 100th percentile, and values ​​lower than that will be changed to the 100th percentile. The 0th percentile will be changed to the 0th percentile. Where the preparation of weather data (11) which includes rainfall data, temperature and The Normalized Difference Vegetation Index (NDVI) (903) was prepared with Finding the maximum, minimum, and average values ​​for each province in each month for all indices. Where the preparation of weather data (11) which includes humidity data is prepared by measurement. Values ​​from telemetry stations with and without data, where measurements are taken from telemetry stations without data. Humidity data was prepared by finding representative points for each province (501), which consisted of the center point. Determine the latitude (equation) and longitude (equation) center points of each province and find the humidity values. Representing the province from the nearest telemetry stations (502) at least 3 stations by finding the weighted average. The weight is based on the distance from the telemetric station. Where the weather data (11) that affects the rice production data (12) will be divided into: If regional data is unavailable, national data will be used instead. Where the mathematical modeling process creates a model to predict rice production data (12) with Multiple Linear Regression Analysis Method using the following factors: The maximum, minimum, and average values ​​of the humidity data from each telemetry station.

7. Methods for forecasting annual rice yields using mathematical models, pursuant to Claim 6, which outlines the import procedures. Input data for rice production (12) selected from the production of Jasmine rice, White rice, Pathum Thani 1, fragrant rice. Province, colored rice, or glutinous rice.

8. Methods for forecasting annual rice yield using mathematical models based on claims 6 or 7, where the point values... The provincial representative (501) defines (x1, y1), (x2, y2), ..., (xn, yn) as the boundaries of the province. Find the vertices and the center of latitude (equation) and the center of longitude (equation) from... The relationship between provincial boundaries and area. (Equation) where (The equation) represents the center of latitude, or the y-axis. (The equation) represents the center of longitude, or the x-axis. A represents the area of ​​the province. n is the total number of vertices. and find the moisture value that is representative of the province from the nearest telemetric station (502) from the weighted average. Based on the distances to the nearest three telemetric stations, as shown in the equation: (Equation) where (The equation) is a weighted average based on the distances from the three nearest telemetric stations. da, db, and dc represent the distances between the centroid and telemetry stations a and b. and c xa, xb, xc are the measured values ​​(highest, lowest, or average for each month) of the telephone station. Measures a, b, and c 9. Develop a mathematical model to forecast rice yield based on one of claims 6 to 8. This is further comprised of steps to establish a database containing yield prediction data in kilograms per rai, and... Procedures for providing Application Program Interface (API) documentation. (143) which retrieves the output prediction results stored in the database (142) and displays them to Map display application (15) or display on another system (16) on the user's device via Identifying the latitude and longitude coordinates of an area.

10. Develop a mathematical model to forecast rice yield based on one of claims 6 through 9. It also includes a yield comparison step, which compares the predicted output with the actual output. The actual yield from the previous year, selected from those with a yield 25 kg / rai lower than the previous year, or those with a yield lower than the previous year. 0-25 kg / rai, yield same as last year, yield 0-25 kg / rai higher than last year, yield 25 kg / rai higher than last year. kg / rai

Images