Local microclimate refined forecasting system and forecasting method based on communication tower group

Abstract

The invention discloses a local microclimate refined forecasting system based on a communication tower group, and belongs to the technical field of weather forecasting, and the system comprises a data collection layer which is used for carrying out the continuous monitoring and collection of multi-dimensional meteorological data through a meteorological station disposed on a communication tower; the data transmission layer is used for transmitting the collected multi-dimensional meteorological data in real time through a communication tower; the data analysis and processing layer is used for carrying out space-time analysis on the transmitted multi-dimensional meteorological data in combination with a GIS technology, establishing a forecasting model of local microclimate, and carrying out model optimization to obtain an optimized forecasting model; and the forecasting service layer is used for forecasting meteorological conditions in a period of time in the future according to the optimized forecasting model, generating customized forecasting information and issuing the customized forecasting information to users through multiple channels. According to the method, communication tower resources can be effectively utilized, the modern information technology is combined, refined forecasting of local area weather changes is achieved, and the method has wide application prospects and social value.

Description

Technical Field

[0001] The present invention belongs to the technical field of meteorological forecasting, and in particular relates to a local microclimate refined forecasting system and forecasting method based on a communication tower group. Background Art

[0002] Traditional weather forecasts often rely on a large-scale network of weather stations, and the forecast accuracy is limited for local microclimate areas such as complex terrain and urban microenvironment. With the popularization of communication infrastructure, especially communication towers throughout urban and rural areas, new possibilities for high-density and high-precision meteorological monitoring have been provided. However, how to effectively integrate these resources to achieve real-time data collection, processing and analysis to improve the accuracy and timeliness of local microclimate forecasts has become the key to current technological development. Summary of the invention

[0003] The purpose of the present invention is to overcome the deficiencies of the prior art and to provide a local microclimate refined forecasting system and forecasting method based on a communication tower group.

[0004] The present invention integrates a micro-meteorological station on or deploys it near a communication tower to collect multi-dimensional meteorological data including temperature, humidity, wind speed, wind direction, air pressure and rainfall, and combines them with the currently known geographic information system (GIS) and artificial intelligence algorithms to achieve high-precision forecasts of future short- to medium-term weather changes in a specific small area.

[0005] The first object of the present invention is to provide a local microclimate fine-grained forecasting system based on a communication tower group, comprising: The data collection layer is used to continuously monitor and collect multi-dimensional meteorological data through meteorological stations deployed on or near communication towers; The data transmission layer is used to transmit the multi-dimensional meteorological data collected by the data collection layer to the data analysis and processing layer in real time through the communication tower; The data analysis and processing layer is used to combine GIS technology to conduct spatiotemporal analysis on the transmitted multi-dimensional meteorological data, establish a forecast model for local microclimate, and optimize the model to obtain an optimized forecast model; The forecast service layer is used to forecast the meteorological conditions in the future based on the optimized forecast model, generate customized forecast information, and publish it to users through various channels.

[0006] Preferably, during the data transmission process, the data transmission layer encrypts and compresses the multi-dimensional meteorological data collected by the data collection layer and transmits the data to the data analysis and processing layer in real time.

[0007] Preferably, the data analysis and processing layer includes: A data preprocessing module is used to preprocess the transmitted multi-dimensional meteorological data to obtain preprocessed data; GIS spatial analysis module, which combines GIS technology and geographic location information to perform spatiotemporal analysis on preprocessed data to obtain time series data sets; The forecast model building module is used to build a forecast model for local microclimate based on the time series data set; The model optimization module is used to automatically optimize the model parameters of the established forecast model to obtain an optimized forecast model.

[0008] Preferably, the local microclimate forecast model is constructed by the following method: S1, divide the time series dataset into a training set and multiple test sets; S2. Use the training set to train different machine learning models, and predict multiple test sets to obtain the average prediction error. S3. Compare the average prediction errors of different machine learning models and select the best performing model, which is the prediction model for the local microclimate.

[0009] Preferably, step S2 comprises the following steps: S21, using the training set to train different machine learning models respectively to obtain a first training model; S22, using the first training model to predict the first test set to obtain a first prediction result; S23, calculating the difference between the first prediction result and the actual observation value of the first test set to obtain a first prediction error; S24, expanding the training set, training the machine learning model again, and predicting the second test set to obtain a second prediction error; S25, repeating steps S5-S8 until all remaining test sets have been traversed to obtain the Nth prediction error; S26. Calculate the average values ​​of the first prediction error, the second prediction error, ..., and the Nth prediction error to obtain the prediction error average value.

[0010] Preferably, step S24 comprises the following steps: S241, expanding the training set forward, adding the data of the first test set, to obtain an extended training set; S242, retraining the machine learning model using the extended training set to obtain a second training model; S243, using the second training model to predict the second test set to obtain a second prediction result; S244. Calculate the difference between the second prediction result and the actual observation value of the second test set to obtain a second prediction error.

[0011] Preferably, the forecast service layer includes: The forecast module is used to forecast the meteorological conditions in the future based on the optimized forecast model; The forecast information generation module is used to generate customized forecast information based on the forecasted meteorological conditions according to the needs of different industries; The information release module is used to release forecast information to users in a timely manner through multiple channels.

[0012] The second object of the present invention is to provide a local microclimate forecasting method of the above-mentioned local microclimate refined forecasting system based on a communication tower group, comprising the following steps: Step 1: Continuously monitor and collect multi-dimensional meteorological data through meteorological stations deployed on or near communication towers; Step 2: The multi-dimensional meteorological data collected by the data collection layer is transmitted to the data analysis and processing layer in real time through the communication tower; Step 3: Combine GIS technology to conduct spatiotemporal analysis on the transmitted multi-dimensional meteorological data, establish a forecast model for local microclimate, and optimize the model to obtain an optimized forecast model; Step 4: Forecast the weather conditions for a period of time in the future based on the optimized forecast model, generate customized forecast information, and release it to users through various channels.

[0013] Preferably, in step 3, the local microclimate forecast model is constructed by the following method: Step 31, dividing the time series data set into a training set and multiple test sets; Step 32: Use the training set to train different machine learning models respectively, and predict multiple test sets respectively to obtain the average prediction error; Step 33: Compare the average prediction errors of different machine learning models and select the best performing model, which is the prediction model for the local microclimate.

[0014] Preferably, step 32 comprises the following steps: Step 321: Use the training set to train different machine learning models respectively to obtain a first training model; Step 322: Use the first training model to predict the first test set to obtain a first prediction result; Step 323: Calculate the difference between the first prediction result and the actual observation value of the first test set to obtain a first prediction error; Step 324: Expand the training set, train the machine learning model again, and predict the second test set to obtain a second prediction error; Step 325, repeat steps S5-S8 until all remaining test sets are traversed and the Nth prediction error is obtained; Step 326: Calculate the average values ​​of the first prediction error, the second prediction error, ..., and the Nth prediction error to obtain the prediction error average value.

[0015] Compared with the prior art, the present invention has the following beneficial effects: (1) The present invention can ensure the accuracy and continuity of data collection through the deployment of micro-meteorological stations, and can achieve fast and stable data transmission through the combined use of micro-meteorological stations and communication towers; (2) The present invention efficiently processes the collected multi-dimensional meteorological data and extracts microclimate characteristics through a machine learning model to construct a forecast model, and can automatically adjust the model parameters of the forecast model to improve the adaptability and accuracy of the forecast; (3) The forecast accuracy of the present invention is higher than that of traditional forecast methods, which significantly improves the temporal and spatial resolution and accuracy of local microclimate forecasts, provides more accurate and timely information support for urban planning, agricultural production, disaster warning and other fields, reduces economic losses caused by sudden weather changes, and enhances society's ability to cope with climate change; (4) The present invention utilizes existing communication tower resources, reducing the infrastructure investment required for building new meteorological stations. In addition, since data transmission uses the network resources of the communication tower, the cost of data transmission is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The architecture diagram, core solution and application field diagram of the local microclimate refined forecasting system provided in Example 1 of the present invention; Figure 2 This is a flow chart of the local microclimate forecasting method provided in Example 2 of the present invention. DETAILED DESCRIPTION

[0017] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Example 1

[0018] The embodiment of the present invention provides a local microclimate fine-grained forecasting system based on a communication tower group, which specifically includes: The data collection layer is used to continuously monitor and collect multi-dimensional meteorological data through meteorological stations deployed on or near communication towers. Specifically, meteorological stations are deployed on communication towers or 1-3 kilometers away from communication towers. The weather station in the embodiment of the present invention specifically uses an existing micro weather station, which is equipped with high-precision sensors for measuring key meteorological parameters such as temperature, humidity, wind speed and direction, air pressure, rainfall, solar radiation, etc. Considering the particularity of the environment in different regions, the weather station should have the characteristics of dustproof, waterproof, high temperature / low temperature resistance, etc., to ensure stable operation and long-term operation in various environments.

[0019] When the embodiment of the present invention makes a specific selection of a micro-meteorological station, it is necessary to select a weather station equipped with low-power consumption and high-precision sensors, such as a temperature and humidity integrated sensor, an ultrasonic wind speed and direction meter, an infrared radiation sensor, an ultraviolet radiation sensor, a particle sensor, a barometer, and a rain gauge. It can collect data such as temperature and humidity, wind speed, wind direction, surface temperature of objects, ultraviolet radiation intensity, concentration of particles in the air, air pressure, and rainfall. When laying out, factors such as terrain and building shielding are considered to ensure the representativeness and comprehensiveness of the data. If necessary, multi-point monitoring at vertical heights is added to capture meteorological changes at different heights.

[0020] The data transmission layer is used to transmit the multi-dimensional meteorological data collected by the data collection layer to the data analysis and processing layer in real time through the communication tower; In the embodiment of the present invention, the communication tower itself serves as a communication relay point, and the built-in data transmission module (such as 4G / 5G, Wi-Fi or LoRaWAN) can directly transmit the data of each micro-meteorological station to the data analysis and processing layer in real time; During the data transmission process, the data transmission layer uses the existing AES algorithm to encrypt the multi-dimensional meteorological data collected by the data collection layer to ensure information security; and uses the existing differential coding data compression algorithm to compress the encrypted data and transmit it to the data analysis and processing layer in real time. Data compression processing can reduce the transmission bandwidth requirements and improve data transmission efficiency.

[0021] The data analysis and processing layer is used to combine GIS technology to conduct spatiotemporal analysis on the transmitted multi-dimensional meteorological data, establish a forecast model for local microclimate, and optimize the model to obtain an optimized forecast model; The data analysis and processing layer in the embodiment of the present invention specifically includes: The data preprocessing module is used to preprocess the transmitted multi-dimensional meteorological data to obtain preprocessed data; specifically, it performs quality control on the transmitted multi-dimensional meteorological data, removes outliers, fills in missing values, and performs standardization to ensure the accuracy of subsequent analysis; GIS spatial analysis module, which combines GIS technology and geographic location information to perform spatiotemporal analysis on preprocessed data to obtain time series data sets; The forecast model building module is used to build a forecast model for local microclimate based on the time series data set; In an embodiment of the present invention, a training model is constructed by applying machine learning models such as deep learning, random forest, and support vector machine, and the training model is used to identify complex meteorological laws. The training design of the model includes steps such as data cleaning, feature engineering, and model optimization, specifically including: using historical meteorological data and ground-truth data for preprocessing, including removing outliers, filling missing values, and data standardization; feature selection to identify the meteorological parameters most relevant to the forecast; applying time series cross-validation technology to train and test the model during the training process to avoid overfitting, and maintaining the time order of the data to ensure the rationality of the model training and testing.

[0022] In the embodiment of the present invention, the prediction model of the local microclimate is constructed by the following method: S1. Divide the time series data set into a training set and multiple test sets; specifically select the first m observations of the time series data as the initial training set. The choice of m depends on the amount of data you want to use in model training; S2. Use the training set to train different machine learning models, and predict multiple test sets to obtain the average prediction error. Specifically, the following steps are included: S21, using the training set to train different machine learning models respectively to obtain a first training model; S22, using the first training model to predict the first test set (the data in this test set is a section of data immediately following the training set) to obtain a first prediction result; S23, calculating the difference between the first prediction result and the actual observation value of the first test set to obtain a first prediction error; S24, expanding the training set, training the machine learning model again, and predicting the second test set to obtain a second prediction error, specifically comprising the following steps: S241, the training set is extended forward, and the data of the first test set is added to obtain an extended training set, and the data length of the extended training set becomes longer; S242, retraining the machine learning model using the extended training set to obtain a second training model; S243, using the second training model to predict the second test set (the data in this test set is a section of data immediately following the first test set) to obtain a second prediction result; S244, calculating the difference between the second prediction result and the actual observation value of the second test set to obtain a second prediction error; S25, repeating steps S5-S8 until all remaining test sets have been traversed to obtain the Nth prediction error; S26, calculating the average of the first prediction error, the second prediction error, ..., and the Nth prediction error to obtain the average prediction error. The above prediction errors are calculated using indicators such as root mean square error (RMSE) and mean absolute error (MAE); S3. Compare the average prediction errors of different machine learning models and select the best performing model, which is the prediction model for the local microclimate.

[0023] The model optimization module is used to automatically optimize the model parameters of the established forecast model to obtain an optimized forecast model. The model optimization module in this embodiment has an online self-learning and optimization mechanism. It regularly retrains the model through the latest meteorological data collected to optimize the model parameters to reflect the latest climate patterns; the model parameters can be automatically adjusted through the adaptive algorithm to improve the prediction accuracy. The specific adaptive algorithm used is the algorithm gradient descent method, which continuously updates the weights and biases to minimize the prediction error and improve the prediction accuracy.

[0024] The forecast service layer is used to forecast the meteorological conditions in the future based on the optimized forecast model, generate customized forecast information, and publish it to users through various channels.

[0025] In the embodiment of the present invention, the forecast service layer includes: The forecast module is used to forecast the meteorological conditions in the future according to the optimized forecast model; specifically, it combines the short-term climate prediction theory with statistical methods and uses the optimized forecast model to forecast the meteorological conditions in the future, including but not limited to hourly, daily and weekly forecasts; The forecast information generation module is used to generate customized forecast information based on the forecasted meteorological conditions according to the needs of different industries; specifically develop a user-friendly interface to provide customized forecast services according to the needs of different industries (such as agriculture, transportation, and energy). For example, it provides soil moisture and evaporation forecasts for agricultural production, and fog, rain, and snow warning information for urban transportation.

[0026] The information release module is used to release forecast information to users in a timely manner through multiple channels. Specifically, forecast information can be released in a timely manner through mobile apps, websites, text messages, social media and other channels to ensure that the information is widely covered and instantly accessible. Example 2

[0027] The embodiment of the present invention provides a local microclimate forecasting method of the above-mentioned local microclimate refined forecasting system based on a communication tower group, which specifically includes the following steps: Step 1: Continuously monitor and collect multi-dimensional meteorological data through meteorological stations deployed on or near communication towers; Step 2: The multi-dimensional meteorological data collected by the data collection layer is transmitted to the data analysis and processing layer in real time through the communication tower; Step 3: Combine GIS technology to conduct spatiotemporal analysis on the transmitted multi-dimensional meteorological data, establish a forecast model for local microclimate, and optimize the model to obtain an optimized forecast model; In this embodiment, the local microclimate forecast model is constructed by the following method: Step 31, dividing the time series data set into a training set and multiple test sets; Step 32: Use the training set to train different machine learning models respectively, and predict multiple test sets respectively to obtain the average prediction error, which specifically includes the following steps: Step 321: Use the training set to train different machine learning models respectively to obtain a first training model; Step 322: Use the first training model to predict the first test set to obtain a first prediction result; Step 323: Calculate the difference between the first prediction result and the actual observation value of the first test set to obtain a first prediction error; Step 324: Expand the training set, train the machine learning model again, and predict the second test set to obtain a second prediction error; Step 325, repeat steps S5-S8 until all remaining test sets are traversed and the Nth prediction error is obtained; Step 326: Calculate the average values ​​of the first prediction error, the second prediction error, ..., and the Nth prediction error to obtain the prediction error average value.

[0028] Step 33: Compare the average prediction errors of different machine learning models and select the best performing model, which is the prediction model for the local microclimate.

[0029] Step 4: Forecast the weather conditions for a period of time in the future based on the optimized forecast model, generate customized forecast information, and release it to users through various channels.

[0030] The forecast system provided by the embodiment of the present invention utilizes the communication tower resources throughout urban and rural areas. Since the communication towers are arranged at a high density and are generally distributed at relatively high points with a ground distance of about 30-50m, higher weather forecast accuracy can be obtained in theory. Moreover, the communication capability of the communication tower itself is utilized during data transmission to ensure the freshness of the data and reduce delays, so that the forecast model can timely reflect the latest meteorological conditions. At the same time, advanced data processing technology and machine learning models are used to make data analysis highly accurate, and the model has strong robustness and generalization capabilities. Therefore, the forecast accuracy of the forecast system provided by the present invention will be significantly higher than that of traditional forecast methods. The system of the present invention can be applied to the fields of urban heat island effect monitoring, sudden weather warning in mountainous areas, and refined management of agricultural production, effectively improving disaster prevention and resource utilization efficiency.

[0031] In summary, the forecasting system provided by the present invention can significantly improve the temporal and spatial resolution and accuracy of local microclimate forecasts, provide more accurate and timely information support for urban planning, agricultural production, disaster warning and other fields, reduce economic losses caused by sudden weather changes, and enhance society's ability to respond to climate change.

[0032] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

Claims

1. A local microclimate fine-grained forecasting system based on a communication tower group, characterized in that: include: The data collection layer is used to continuously monitor and collect multi-dimensional meteorological data through meteorological stations deployed on or near communication towers; The data transmission layer is used to transmit the multi-dimensional meteorological data collected by the data collection layer to the data analysis and processing layer in real time through the communication tower; The data analysis and processing layer is used to combine GIS technology to conduct spatiotemporal analysis on the transmitted multi-dimensional meteorological data, establish a forecast model for local microclimate, and optimize the model to obtain an optimized forecast model; The forecast service layer is used to forecast the meteorological conditions in the future based on the optimized forecast model, generate customized forecast information, and publish it to users through various channels.

2. The local microclimate refined forecasting system based on communication tower groups according to claim 1 is characterized in that: During the data transmission process, the data transmission layer encrypts and compresses the multi-dimensional meteorological data collected by the data collection layer and transmits them to the data analysis and processing layer in real time.

3. The local microclimate refined forecasting system based on communication tower groups according to claim 1 is characterized in that: The data analysis and processing layer includes: A data preprocessing module is used to preprocess the transmitted multi-dimensional meteorological data to obtain preprocessed data; GIS spatial analysis module, which combines GIS technology and geographic location information to perform spatiotemporal analysis on preprocessed data to obtain time series data sets; The forecast model building module is used to build a forecast model for local microclimate based on the time series data set; The model optimization module is used to automatically optimize the model parameters of the established forecast model to obtain an optimized forecast model.

4. The local microclimate refined forecasting system based on communication tower groups according to claim 3 is characterized in that: The local microclimate forecast model is constructed by the following method: S1, divide the time series dataset into a training set and multiple test sets; S2. Use the training set to train different machine learning models, and predict multiple test sets to obtain the average prediction error. S3. Compare the average prediction errors of different machine learning models and select the best performing model, which is the prediction model for the local microclimate.

5. The local microclimate refined forecasting system based on communication tower groups according to claim 4 is characterized in that: Step S2 comprises the following steps: S21, using the training set to train different machine learning models respectively to obtain a first training model; S22, using the first training model to predict the first test set to obtain a first prediction result; S23, calculating the difference between the first prediction result and the actual observation value of the first test set to obtain a first prediction error; S24, expanding the training set, training the machine learning model again, and predicting the second test set to obtain a second prediction error; S25, repeating steps S5-S8 until all remaining test sets have been traversed to obtain the Nth prediction error; S26. Calculate the average values ​​of the first prediction error, the second prediction error, ..., and the Nth prediction error to obtain the prediction error average value.

6. The local microclimate refined forecasting system based on communication tower groups according to claim 5 is characterized in that: Step S24 comprises the following steps: S241, expanding the training set forward, adding the data of the first test set, to obtain an extended training set; S242, retraining the machine learning model using the extended training set to obtain a second training model; S243, using the second training model to predict the second test set to obtain a second prediction result; S244. Calculate the difference between the second prediction result and the actual observation value of the second test set to obtain a second prediction error.

7. The local microclimate refined forecasting system based on communication tower groups according to claim 5 is characterized in that: The forecast service layer includes: The forecast module is used to forecast the meteorological conditions in the future based on the optimized forecast model; The forecast information generation module is used to generate customized forecast information based on the forecasted meteorological conditions according to the needs of different industries; The information release module is used to release forecast information to users in a timely manner through multiple channels.

8. A method for fine-grained forecasting of local microclimate based on a communication tower group, characterized in that: The following steps are involved: Step 1: Continuously monitor and collect multi-dimensional meteorological data through meteorological stations deployed on or near communication towers; Step 2: The multi-dimensional meteorological data collected by the data collection layer is transmitted to the data analysis and processing layer in real time through the communication tower; Step 3: Combine GIS technology to conduct spatiotemporal analysis on the transmitted multi-dimensional meteorological data, establish a forecast model for local microclimate, and optimize the model to obtain an optimized forecast model; Step 4: Forecast the weather conditions for a period of time in the future based on the optimized forecast model, generate customized forecast information, and release it to users through various channels.

9. The local microclimate forecasting method according to claim 8, characterized in that: In step 3, the local microclimate forecast model is constructed by the following method: Step 31, dividing the time series data set into a training set and multiple test sets; Step 32: Use the training set to train different machine learning models respectively, and predict multiple test sets respectively to obtain the average prediction error; Step 33: Compare the average prediction errors of different machine learning models and select the best performing model, which is the prediction model for the local microclimate.

10. The local microclimate forecasting method according to claim 9, characterized in that: Step 32 includes the following steps: Step 321: Use the training set to train different machine learning models respectively to obtain a first training model; Step 322: Use the first training model to predict the first test set to obtain a first prediction result; Step 323: Calculate the difference between the first prediction result and the actual observation value of the first test set to obtain a first prediction error; Step 324: Expand the training set, train the machine learning model again, and predict the second test set to obtain a second prediction error; Step 325, repeat steps S5-S8 until all remaining test sets are traversed and the Nth prediction error is obtained; Step 326: Calculate the average values ​​of the first prediction error, the second prediction error, ..., and the Nth prediction error to obtain the prediction error average value.

Images