Green tea growth environment intelligent monitoring method based on adaptive sensor network

Abstract

The invention discloses a green tea growth environment intelligent monitoring method based on an adaptive sensor network. The method comprises the following steps: S1, deploying a plurality of sensor nodes in a green tea planting area; s2, dynamically adjusting the working state and the data transmission path of each sensor node through a self-adaptive sensor network protocol; s3, performing primary processing and analysis on the acquired environment data; s4, performing deep analysis on the environment data by using cloud computing resources, and extracting key environment features; s5, using the prediction model to analyze the future trend of the green tea growth environment, and generating targeted management suggestions; s6, equipment in the tea garden is automatically controlled through an intelligent decision system, and the green tea growth environment is adjusted and optimized; and S7, providing a visual monitoring interface for the user on the cloud platform, and pushing optimization suggestions and alarm information. According to the method, an efficient and scientific optimization scheme can be provided in green tea growth environment monitoring, and remarkable technical values and economic benefits are brought to practical application.

Description

Technical Field

[0001] The present invention relates to the field of agriculture, and particularly to an intelligent monitoring method for the green tea growth environment based on an adaptive sensing network. Background Art

[0002] With the development of intelligent agriculture, the environmental monitoring of green tea planting has gradually become a key factor in improving tea yield and quality. Traditional environmental monitoring systems usually rely on manual monitoring and basic climate data collection. Although they can provide certain monitoring data, due to the limitations of human resources and technical means, they cannot capture the changes in the tea garden environment in real time and comprehensively, resulting in the inability to effectively respond to the fluctuations of key environmental factors such as temperature and humidity. Existing systems usually use a single sensor to collect data and lack the ability of automated and intelligent decision-making. They often require manual intervention to judge and adjust, increasing labor costs and management difficulties.

[0003] With the development of Internet of Things technology, although there are some intelligent monitoring systems that can collect data through sensors and transmit it to the cloud for analysis, the intelligence level of these systems is low and they cannot process complex data in real time. In the prior art, most systems rely on basic statistical methods and simple algorithms for environmental data analysis and lack in-depth mining and analysis of multi-dimensional environmental data. For example, existing systems cannot effectively identify the mutual relationships and long-term trends between different environmental parameters. Therefore, when facing sudden climate changes, the systems react slowly and it is difficult to provide timely and accurate management suggestions.

[0004] In addition, traditional sensor networks mostly rely on fixed deployment methods and limited network bandwidth, lack adaptability, and cannot adjust the working state or data transmission path according to environmental changes, resulting in low system efficiency. The calculation and data transmission burden of the system is heavy, which easily causes data delay and cannot feedback the system state in real time.

[0005] The defects of the prior art are that it cannot accurately capture the spatial and temporal changes of the environment, lacks the ability of in-depth analysis and intelligent decision-making, resulting in the system being unable to respond to environmental changes in real time and accurately. To solve this problem, the present invention provides an efficient and intelligent monitoring method for the green tea growth environment through an adaptive sensing network combined with advanced technologies such as deep learning and graph neural network, which can analyze multi-dimensional data in real time, accurately predict future environmental changes, so as to optimize management decisions and improve the management efficiency of the tea garden. Summary of the Invention

[0006] An object of the present invention is to propose an intelligent monitoring method for the green tea growth environment based on an adaptive sensing network. The present invention can provide an efficient and scientific optimization scheme in the intelligent monitoring of the green tea growth environment, bringing significant technical value and economic benefits to practical applications.

[0007] An intelligent monitoring method for the green tea growth environment based on an adaptive sensor network according to an embodiment of the present invention includes the following steps:

[0008] S1. Deploy a plurality of sensor nodes inside the green tea planting area for collecting environmental data;

[0009] S2. Through the adaptive sensor network protocol, dynamically adjust the working states and data transmission paths of each sensor node according to the power, network quality, and data load of the sensor nodes;

[0010] S3. Transmit the environmental data collected by each sensor node to the data processing center through wireless communication technology, and use edge computing devices to preliminarily process and analyze the collected environmental data;

[0011] S4. Upload the processed environmental data to the cloud platform, and use cloud computing resources to deeply analyze the environmental data, and extract key environmental features based on the K-means algorithm and data mining algorithm;

[0012] S5. According to the real-time environmental data and historical environmental data, use the prediction model to analyze the future trend of the green tea growth environment and generate targeted management suggestions;

[0013] S6. Through the intelligent decision-making system, automatically control the devices in the tea garden, including the irrigation system, temperature control equipment, and ventilation device, to adjust and optimize the green tea growth environment;

[0014] S7. Provide a visual monitoring interface for users on the cloud platform to display the environmental status of the tea garden, and push optimization suggestions and alarm information according to the analysis results.

[0015] Optionally, the S1 includes the following steps:

[0016] S11. Deploy a plurality of sensor nodes inside the green tea planting area, and each sensor node includes a temperature sensor, a humidity sensor, a light sensor, a soil humidity sensor, and a carbon dioxide concentration sensor;

[0017] S12. The temperature sensor of each sensor node is used to collect the temperature data of the green tea growth environment in real time, the humidity sensor is used to collect the humidity data of the ambient air, the light sensor is used to measure the light intensity of the green tea planting area, the soil humidity sensor is used to monitor the water content in the soil, and the carbon dioxide concentration sensor is used to detect the carbon dioxide concentration in the air;

[0018] S13. Each sensor node communicates through a preset wireless communication protocol and transmits data to the control node, and the communication protocol includes LoRa and NB-IoT;

[0019] S14. The deployment method of the sensor nodes is to be evenly distributed within the tea garden area. According to the geographical and climatic conditions of the tea garden, multiple monitoring points are selected for data collection;

[0020] S15. The sensor nodes dynamically adjust their working states according to environmental changes and battery power. The sensor nodes operate in the low-power mode and enter the standby mode after collecting a certain amount of data, reducing the energy consumption of the system and extending the working time of the sensor nodes.

[0021] Optionally, the S2 includes the following steps:

[0022] S21. Each sensor node calculates the remaining battery power E in real time according to environmental changes and power information remaining ;

[0023] S22. When E remaining is lower than the preset threshold E threshold , the sensor node automatically enters the low-power mode and adjusts its data collection frequency. The collection frequency is calculated through an adaptive algorithm and set to:

[0024] f new = f initial ·α;

[0025] where f new is the new collection frequency, f initial is the initial collection frequency, α is the adjustment factor, and α < 1;

[0026] S23. The sensor node selects the optimal transmission path P optimal , and the selection of the transmission path is based on an adaptive algorithm, comprehensively considering the signal strength S signal between current nodes, the packet transmission delay D delay and the network load L load . The preference degree V path of the transmission path is calculated through the following formula:

[0027]

[0028] Select the path with the largest V path as the optimal transmission path;

[0029] S24. During the communication process between nodes, the transmission rate R transmission is dynamically adjusted. The transmission rate is adjusted through an adaptive algorithm according to the network load L load and the delay D delay and set to:

[0030]

[0031] Among them, R max is the maximum transmission rate, and β is the adjustment coefficient;

[0032] S25. The node obtains the time to enter the standby state through an adaptive algorithm according to the real-time measured environmental data and network quality information. The standby time T standby is calculated dynamically according to the current minimum standby time T min of the node, the network load L load and the signal strength S signal , and is set as:

[0033] T standby = T min + γ·L load + δ·S signal ;

[0034] Among them, γ and δ are adjustment coefficients.

[0035] Optionally, the S3 includes the following steps:

[0036] S31. Transmit the environmental data collected by each sensor node to the data processing center through wireless communication technology, and use the edge computing device to perform preliminary processing and analysis on the collected environmental data;

[0037] S32. In the edge computing device, perform preliminary cleaning and screening on the data, remove noise and incomplete data, and obtain the effective data set D valid , and the effective data set D valid includes the filtered environmental parameters P filtered , P filtered = {P1, P2,..., P n}, where P i is the effective data collected by each corresponding sensor;

[0038] S33. Perform preliminary statistical analysis on the effective data set D valid , including data average value, maximum value, minimum value, and standard deviation, in order to screen out important monitoring parameters in the preliminary stage;

[0039] S34. The edge computing device performs simplified processing on the environmental data and extracts the real-time change trend T trend :

[0040]

[0041] where P current is the currently collected environmental data, P previous is the previously collected environmental data, and Δt is the collection interval time;

[0042] S35. Upload the processed effective dataset D processed and the change trend T trend to the cloud platform through the adaptive network protocol for more in-depth data analysis and processing.

[0043] Optionally, the said S4 includes the following steps:

[0044] S41. Transmit the processed dataset D in the edge computing device processed to the cloud platform through the wireless communication protocol. D processed includes the processing results of temperature, humidity, light intensity, and soil humidity data;

[0045] S42. On the cloud platform, preprocess the uploaded environmental data using data cleaning and standardization methods. The standardized dataset is denoted as D standardized , D standardized in which each environmental parameter P normalized is standardized through the following formula:

[0046]

[0047] where P original is the original data, μ is the mean of the parameter, and σ is the standard deviation of the parameter;

[0048] S43. Use the K-means clustering algorithm to process the standardized dataset D standardized and extract the key environmental features F key , including temperature and humidity changes, soil humidity fluctuations, and periodic light intensity changes;

[0049] S44. Based on data mining techniques, use association rule analysis and time series analysis methods to explore the internal laws of environmental data and obtain the correlation matrix M correlation of environmental parameters. Each correlation coefficient r correlation in M ij represents the degree of correlation between environmental parameters P i and P j . The correlation matrix is calculated through the following formula:

[0050]

[0051] where cov(P i , P j ) is the covariance between environmental parameters P i and P j , and σ(P i ) and σ(P j ) are the standard deviations of the corresponding parameters.

[0052] Optionally, S5 includes the following steps:

[0053] S51. Receive the processed dataset D from the cloud platform standardized , including the real-time collected environmental data and historical record data;

[0054] S52. Combine the historical data and real-time data, and use the method of combining graph neural network and time series modeling to establish a prediction model M GNN , M GNN Process the mutual relationship between sensors through the graph structure, and capture the temporal features in the environmental data at the same time. The calculation process is as follows:

[0055]

[0056] Among them, represents the hidden state of node i at the k+1 layer, N(i) is the set of neighbor nodes of node i, c ij is the normalization coefficient, W (k) is the weight matrix of the kth layer, b (k) is the bias term of the kth layer, σ(.) is the sigmoid activation function, is the output of the k+1 layer;

[0057] S53. In the graph neural network, the mutual dependence of sensor nodes is modeled through the graph structure to generate the prediction result Y of the future environmental state of each node future ;

[0058] S54. On the basis of the output of the graph neural network, combine the multi-layer convolutional neural network to extract features from the time series data to obtain a more accurate prediction T of the future environmental trend future :

[0059]

[0060] Among them, x i is the feature of the i-th moment of the input environmental data, w i is the convolution kernel weight;

[0061] S55. Combine the spatial dependence of the graph neural network and the time features extracted by the convolutional neural network to generate an all-round prediction result Y of the green tea growth environment predict , and obtain the future environmental change trend T predict :

[0062]

[0063] Among them, Y current is the current environmental data, and Δt is the time span;

[0064] S56. Generate targeted management advice A based on the prediction results and trends of change recommendation , including optimizing irrigation strategies and adjusting temperature and humidity measures.

[0065] Optionally, S6 includes the following steps:

[0066] S61. Based on the future environmental trend prediction T GNN generated by the prediction model M future and the current environmental data Y current , adjust the environmental equipment in the tea garden in real time;

[0067] S62. According to the soil humidity P soil , air humidity P humidity , and temperature P temperature predict the trend of change, and use the following control algorithm to adjust the irrigation system:

[0068]

[0069] where I control is the irrigation intensity, k soil is the irrigation adjustment coefficient, P soil,target is the target soil humidity, and P soil is the current soil humidity;

[0070] S63. According to the predicted changes in temperature P temperature and humidity P humidity , adjust the working states of the temperature control equipment and the ventilation system, and use the fuzzy control algorithm to achieve the balanced adjustment of temperature and humidity:

[0071] U temperature = λ1·(P temperature-t - P temperature ) + λ2·(P humidity-t - P humidity );

[0072] where U temperature is the adjustment output of the temperature control equipment, λ1 and λ2 are control coefficients, P temperature-t is the target temperature, and P humidity-t is the target humidity;

[0073] S64. According to the predicted environmental trends and the current working states of the equipment, adjust the operation order and priorities of multiple equipment through an optimization algorithm, and use the priority scheduling model for equipment control:

[0074]

[0075] where C device is the selected equipment, and w i is the weight coefficient of the equipment, The contribution value of the i-th device to the future environmental trend;

[0076] S65. According to the system control feedback, adjust the control parameters in real time to optimize the operation of the irrigation, temperature control, and ventilation systems.

[0077] Optionally, the S7 includes the following steps:

[0078] S71. Create a real-time data display interface on the cloud platform. Through data visualization technology, compare the real-time environmental data of the sensor nodes with the historical data to generate a trend analysis graph T trend,display , The trend analysis graph shows the change of environmental data over time, and calculates the trend change rate according to the following formula:

[0079]

[0080] Where P current is the data at the current time point, P previous is the data at the previous time point, and Δt is the time difference;

[0081] S72. According to the prediction model M GNN and the prediction result T future , the system automatically generates and pushes optimization suggestions A recommendation,display , The suggestion content includes adjusting irrigation, temperature control, and ventilation operations:

[0082] A recommendation,display = f(T future ,P current ,P target );

[0083] Where T future is the predicted environmental trend, P target is the target environmental parameter, and f(.) is the mapping function;

[0084] S73. Display the alarm information A alert in the user interface. When a certain environmental data P current exceeds the set threshold P threshold , an alarm is automatically triggered:

[0085]

[0086] Where Alert indicates that the alarm is triggered, and No Alert indicates that the alarm is not triggered.

[0087] The beneficial effects of the present invention are:

[0088] (1) The present invention combines an adaptive sensing network and deep learning technology to achieve real-time collection and analysis of tea garden environmental data. Through multi-dimensional environmental data collection and in-depth analysis, the system can more comprehensively reflect the changes in the tea garden environment, thus effectively avoiding management errors caused by incomplete or lagged data collection in traditional monitoring systems.

[0089] (2) The present invention utilizes graph neural networks and advanced time series prediction algorithms to accurately capture the spatial dependencies and temporal variation patterns in environmental data. Compared with traditional statistical methods and simple algorithms, the deep learning model can deeply explore the complex relationships between environmental parameters, identify potential environmental change trends, and through accurate prediction of future environmental trends, the system can respond in advance, providing more reliable decision-making basis for management personnel, thereby greatly improving the early warning ability of the system and the ability to handle emergencies.

[0090] (3) The present invention also optimizes the data transmission path and sensor working state through an adaptive network protocol and an intelligent decision-making mechanism. Compared with the disadvantages of traditional sensor networks that cannot adapt to environmental changes, the system can dynamically adjust the working state according to the power of sensor nodes, network load, etc., improving the reliability of the network and the energy utilization efficiency. The adaptive adjustment method not only ensures the stable transmission of data, but also reduces the operating cost of the system and extends the service life of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0091] The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation to the present invention. In the drawings:

[0092] Figure 1 is a flowchart of an intelligent monitoring method for the green tea growth environment based on an adaptive sensing network proposed by the present invention;

[0093] Figure 2 is a flowchart of intelligent regulation of the green tea growth environment in an intelligent monitoring method for the green tea growth environment based on an adaptive sensing network proposed by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0094] Now, the present invention will be further described in detail with reference to the drawings. These drawings are all simplified schematic diagrams, only illustrating the basic structure of the present invention in a schematic manner, so they only show the components related to the present invention.

[0095] Reference Figure 1 - Figure 2 , an intelligent monitoring method for the green tea growth environment based on an adaptive sensing network, includes the following steps:

[0096] S1. Deploy multiple sensor nodes in the green tea planting area to collect environmental data;

[0097] S2. Dynamically adjust the working states and data transmission paths of each sensor node according to the power, network quality, and data load of the sensor nodes through the adaptive sensing network protocol;

[0098] S3. Transmit the environmental data collected by each sensor node to the data processing center through wireless communication technology, and use edge computing devices to perform preliminary processing and analysis on the collected environmental data;

[0099] S4. Upload the processed environmental data to the cloud platform, and use cloud computing resources to perform in-depth analysis on the environmental data, and extract key environmental features based on the K-means algorithm and data mining algorithm;

[0100] S5. Analyze the future trends of the green tea growth environment based on real-time environmental data and historical environmental data, and generate targeted management suggestions;

[0101] S6. Automatically control the devices in the tea garden, including irrigation systems, temperature control devices, and ventilation devices, through an intelligent decision-making system to adjust and optimize the green tea growth environment;

[0102] S7. Provide a visual monitoring interface for users on the cloud platform to display the environmental status of the tea garden, and push optimization suggestions and alarm information according to the analysis results.

[0103] In this embodiment, S1 includes the following steps:

[0104] S11. Deploy multiple sensor nodes inside the green tea planting area, and each sensor node includes a temperature sensor, a humidity sensor, a light sensor, a soil humidity sensor, and a carbon dioxide concentration sensor;

[0105] S12. The temperature sensor of each sensor node is used to collect the temperature data of the green tea growth environment in real time, the humidity sensor is used to collect the humidity data of the ambient air, the light sensor is used to measure the light intensity of the green tea planting area, the soil humidity sensor is used to monitor the water content in the soil, and the carbon dioxide concentration sensor is used to detect the carbon dioxide concentration in the air;

[0106] S13. Each sensor node communicates through a preset wireless communication protocol and transmits data to the control node. The communication protocols include LoRa and NB-IoT;

[0107] S14. The sensor nodes are arranged in a uniform distribution within the tea garden area, and multiple monitoring points are selected for data collection according to the geographical and climatic conditions of the tea garden;

[0108] S15. The sensor node dynamically adjusts its working state according to environmental changes and battery power. The sensor node operates in the low-power mode and enters the standby mode after collecting a certain amount of data, reducing the energy consumption of the system and extending the working time of the sensor node.

[0109] In this embodiment, S2 includes the following steps:

[0110] S21. Each sensor node calculates the remaining battery power E in real time according to environmental changes and power information remaining ;

[0111] S22. When E remaining is lower than the preset threshold E threshold , the sensor node automatically enters the low-power mode and adjusts its data collection frequency. The collection frequency is calculated by an adaptive algorithm and set as:

[0112] f new = f initial ·α;

[0113] where f new is the new collection frequency, f initial is the initial collection frequency, and α is the adjustment factor, with α < 1;

[0114] S23. The sensor node selects the optimal transmission path P optimal according to network quality and data transmission load. The selection of the transmission path is based on an adaptive algorithm, comprehensively considering the signal strength S signal between current nodes, the data packet transmission delay D delay and the network load L load . The preference V path of the transmission path is calculated through the following formula:

[0115]

[0116] Select the path with the maximum V path as the optimal transmission path;

[0117] S24. During the communication process between nodes, the transmission rate R transmission is dynamically adjusted. The transmission rate is adjusted by an adaptive algorithm according to the network load L load and the delay D delay and set as:

[0118]

[0119] where R max is the maximum transmission rate and β is the adjustment coefficient;

[0120] S25. The node obtains the time to enter the standby state through an adaptive algorithm based on the real-time measured environmental data and network quality information, and the standby time is T standby According to the current minimum standby time T of the node min , network load L load and signal strength S signal dynamically calculate and set it as:

[0121] T standby = T min + γ·L load + δ·S signal ;

[0122] where γ and δ are adjustment coefficients.

[0123] In this embodiment, S3 includes the following steps:

[0124] S31. Transmit the environmental data collected by each sensor node to the data processing center through wireless communication technology, and use the edge computing device to perform preliminary processing and analysis on the collected environmental data;

[0125] S32. In the edge computing device, perform preliminary data cleaning and screening to remove noise and incomplete data, and obtain the effective data set D valid , the effective data set D valid includes the filtered environmental parameters P filtered , P filtered = {P1, P2,..., P n}), where P i is the effective data collected by each corresponding sensor;

[0126] S33. Perform preliminary statistical analysis on the effective data set D valid through the edge computing device, including data average value, maximum value, minimum value, and standard deviation, in order to screen out important monitoring parameters in the preliminary stage;

[0127] S34. The edge computing device simplifies the environmental data and extracts the real-time change trend T trend :

[0128]

[0129] where P current is the currently collected environmental data, P previous is the previously collected environmental data, and Δt is the collection interval time;

[0130] S35. Transmit the processed effective data set D processed and the change trend T trendUploaded to the cloud platform through the adaptive network protocol for more in-depth data analysis and processing.

[0131] In this embodiment, S4 includes the following steps:

[0132] S41. Transmit the processed dataset D in the edge computing device to the cloud platform through the wireless communication protocol. D processed transmits, and D processed includes the processing results of temperature, humidity, light intensity, and soil humidity data;

[0133] S42. On the cloud platform, preprocess the uploaded environmental data using data cleaning and standardization methods. The standardized dataset is denoted as D standardized , and D standardized each environmental parameter P normalized is standardized through the following formula:

[0134]

[0135] where P original is the original data, μ is the mean of the parameter, and σ is the standard deviation of the parameter;

[0136] S43. Use the K-means clustering algorithm to process the standardized dataset D standardized and extract the key environmental features F key , including temperature and humidity changes, soil humidity fluctuations, and periodic changes in light intensity;

[0137] S44. Based on data mining techniques, use association rule analysis and time series analysis methods to explore the internal laws of environmental data and obtain the correlation matrix M correlation of environmental parameters. Each correlation coefficient r correlation in M ij represents the degree of correlation between environmental parameters P i and P j . The correlation matrix is calculated through the following formula:

[0138]

[0139] where cov(P i , P j ) is the covariance of environmental parameters P i and P j , and σ(P i ) and σ(P j ) are the standard deviations of the corresponding parameters.

[0140] In this embodiment, S5 includes the following steps:

[0141] S51. Receive the processed dataset D from the cloud platform standardized , including the real-time collected environmental data and historical record data;

[0142] S52. Combine the historical data and real-time data, and use the method of combining graph neural network with time series modeling to establish the prediction model M GNN , M GNN Process the mutual relationship between sensors through the graph structure, and capture the temporal features in the environmental data at the same time. The calculation process is as follows:

[0143]

[0144] Among them, represents the hidden state of node i at the k+1 layer, N(i) is the set of neighbor nodes of node i, c ij is the normalization coefficient, W (k) is the weight matrix of the kth layer, b (k) is the bias term of the kth layer, σ(.) is the sigmoid activation function, is the output of the k+1 layer;

[0145] S53. In the graph neural network, the mutual dependence of sensor nodes is modeled through the graph structure to generate the prediction result Y of the future environmental state of each node future ;

[0146] S54. On the basis of the output of the graph neural network, combine the multi-layer convolutional neural network to extract features from the time series data to obtain a more accurate prediction T of the future environmental trend future :

[0147]

[0148] Among them, x i is the feature of the i-th moment of the input environmental data, w i is the convolutional kernel weight;

[0149] S55. Combine the spatial dependence of the graph neural network and the time features extracted by the convolutional neural network to generate an all-round prediction result Y of the green tea growth environment predict , and obtain the future environmental change trend T predict :

[0150]

[0151] Among them, Y current is the current environmental data, and Δt is the time span;

[0152] S56. Generate targeted management suggestions A according to the prediction results and change trends recommendation, including optimized irrigation strategies and measures for adjusting temperature and humidity.

[0153] In this embodiment, S6 includes the following steps:

[0154] S61. Based on the future environmental trend prediction T GNN generated by the prediction model M future and the current environmental data Y current , adjust the environmental equipment in the tea garden in real time;

[0155] S62. According to the predicted change trends of soil humidity P soil , air humidity P humidity , and temperature P temperature , adjust the irrigation system using the following control algorithm:

[0156]

[0157] where I control is the irrigation intensity, k soil is the irrigation adjustment coefficient, P soil,target is the target soil humidity, and P soil is the current soil humidity;

[0158] S63. According to the predicted changes in temperature P temperature and humidity P humidity , adjust the working states of the temperature control equipment and the ventilation system, and use the fuzzy control algorithm to achieve the balanced adjustment of temperature and humidity:

[0159] U temperature = λ1·(P temperature-t - P temperature ) + λ2·(P humidity-t - P humidity );

[0160] where U temperature is the adjustment output of the temperature control equipment, λ1 and λ2 are control coefficients, P temperature-t is the target temperature, and P humidity-t is the target humidity;

[0161] S64. According to the predicted environmental trend and the current working states of the equipment, adjust the operation sequence and priority of multiple equipment through an optimization algorithm, and use the priority scheduling model for equipment control:

[0162]

[0163] where C device is the selected equipment, w i is the weight coefficient of the equipment, is the contribution value of the i-th equipment to the future environmental trend;

[0164] S65. Adjust the control parameters in real time according to the system control feedback to optimize the operation of the irrigation, temperature control, and ventilation systems.

[0165] In this embodiment, S7 includes the following steps:

[0166] S71. Create a real-time data display interface on the cloud platform. Through data visualization technology, compare the real-time environmental data of the sensor nodes with the historical data to generate a trend analysis chart T trend,display , which shows the change of environmental data over time, and calculate the trend change rate according to the following formula:

[0167]

[0168] where P current is the data at the current time point, P previous is the data at the previous time point, and Δt is the time difference;

[0169] S72. According to the prediction model M GNN and the prediction result T future , the system automatically generates and pushes optimization suggestions A recommendation,display , and the suggestion content includes adjusting irrigation, temperature control, and ventilation operations:

[0170] A recommendation,display = f(T future , P current , P target );

[0171] where T future is the predicted environmental trend, P target is the target environmental parameter, and f(.) is the mapping function;

[0172] S73. Display the alarm information A alert in the user interface. When a certain environmental data P current exceeds the set threshold P threshold , an alarm is automatically triggered:

[0173]

[0174] where Alert indicates that an alarm is triggered, and No Alert indicates that no alarm is triggered.

[0175] Example:

[0176] In this embodiment, a green tea plantation located in Xihu District, Zhejiang Province is used as the application scenario. The climate in this area is warm and humid, suitable for the growth of green tea. The soil is fertile and maintains an appropriate humidity throughout the year. With the increasing production of tea leaves, the deficiencies of the traditional manual management mode have gradually emerged. Especially in environmental monitoring and management, due to the lag of manual work and the one-sidedness of monitoring data, the growth of tea trees is often adversely affected. Therefore, the intelligent monitoring method for the green tea growth environment based on the adaptive sensor network proposed by the present invention aims to solve these problems. By real-time monitoring, data analysis and intelligent decision-making, it optimizes the tea garden management and improves the yield and quality of green tea.

[0177] In this green tea plantation, the implementer deployed multiple sensor nodes to collect key environmental data in real time, including parameters such as temperature, humidity, light intensity, soil humidity, carbon dioxide concentration, etc. These sensors upload the data to the edge computing device through wireless communication for preliminary data processing, and then transmit the processed data to the cloud platform. The cloud platform conducts in-depth analysis of the data through deep learning models and graph neural networks.

[0178] To verify the effectiveness of the present invention, the implementer conducted an experiment for three months from June to August 2024. During the experiment, the implementer verified the advantages of the intelligent monitoring system by comparing the green tea growth data obtained by using the traditional manual management method and the method of the present invention.

[0179] At the beginning of the experiment, the implementer collected detailed data on the tea garden environment and recorded the initial environmental parameters and the growth situation of tea trees. Subsequently, the implementer compared the fluctuations of environmental parameters such as temperature, humidity, and soil humidity in the tea garden under the traditional management method with the environmental data automatically adjusted by the intelligent monitoring system based on the present invention. The experimental results showed that the tea garden using the intelligent monitoring system had significant improvements in many aspects.

[0180] First of all, the control of temperature and humidity showed obvious advantages. Under the traditional management, the temperature in the tea garden once reached above 32°C. However, due to the lag of manual operation, the temperature control equipment could not be started in time, resulting in the phenomenon of dry tea tree leaves and growth stagnation. After using the intelligent monitoring system of the present invention, the temperature in the tea garden was effectively controlled between 28°C and 30°C, and the humidity was maintained within the appropriate range of 70%-80%, ensuring the best growth conditions for tea trees.

[0181] Secondly, in terms of irrigation management, the traditional method relies on manual inspection of soil humidity and usually conducts irrigation once every one or two days. After using the intelligent monitoring system, the soil humidity data can be monitored in real time, and the system automatically judges whether the soil humidity is lower than the set threshold and starts irrigation in time to ensure that the soil humidity is maintained between 40%-60%.

[0182] In terms of the yield and quality of tea leaves, the use of the intelligent monitoring system has also brought significant improvements. According to the experimental data, during the three-month experimental period, the tea gardens using the intelligent system had approximately 20% higher yields compared to those managed by traditional methods. Specifically, the tea leaves in the tea gardens were more plump, the growth rate of tea buds accelerated, and the quality improved. According to the analysis data after tea picking, the tea leaves under the intelligent monitoring system contained higher levels of tea polyphenols and amino acids, which significantly helped improve the quality of green tea.

[0183] By comparing with the traditional method, the specific data is shown in Table 1 below:

[0184] Table 1 Comparison of environmental parameters between the present invention and the traditional method

[0185]

[0186] Throughout the embodiments, the implementer not only solved the problems of the traditional green tea planting environment monitoring system being unable to respond to environmental changes in real time and having low management efficiency through the method of the present invention, but also improved the intelligent level of tea garden management, achieving precise prediction and automatic adjustment based on environmental data.

[0187] The present invention combines advanced technologies such as adaptive sensor networks, deep learning, and graph neural networks to conduct real-time monitoring and precise analysis of multi-dimensional data of the tea garden environment. Compared with the traditional single-sensor data acquisition method, the present invention can collect environmental data more comprehensively and combine spatial and temporal information to make precise predictions of future environmental changes, having significant advantages in controlling key environmental parameters such as temperature, humidity, and soil moisture. Through precise prediction, the system can automatically adjust the irrigation, temperature control, and ventilation systems, reducing delays in manual operations and improving the response speed and accuracy of the system.

[0188] The present invention introduces the fusion of deep learning models and graph neural networks in tea garden management, using a method that combines spatial dependence and time series prediction to improve the accuracy of predicting environmental change trends. By adaptively adjusting the working modes and data transmission paths of each sensor, the intelligent monitoring system greatly improves the irrigation efficiency and environmental regulation efficiency of the tea garden, reduces water resource waste, and at the same time ensures the best environmental conditions for the growth of tea trees.

[0189] The above is only the preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, making equivalent substitutions or changes, should be covered within the protection scope of the present invention.

Claims

1. An intelligent monitoring method for green tea growth environment based on an adaptive sensor network, characterized in that: The steps include: S1. Deploy multiple sensor nodes in the green tea planting area to collect environmental data; S2, through the adaptive sensor network protocol, dynamically adjust the working state and data transmission path of each sensor node according to the power of the sensor node, network quality and data load; S3, transmitting the environmental data collected by each sensor node to the data processing center through wireless communication technology, and using edge computing equipment to perform preliminary processing and analysis on the collected environmental data; S4, upload the processed environmental data to the cloud platform, and use cloud computing resources to conduct in-depth analysis of the environmental data, and extract key environmental features based on the K-means algorithm and data mining algorithm; S5. Based on real-time and historical environmental data, use the prediction model to analyze the future trend of the green tea growing environment and generate targeted management suggestions; S6. Through the intelligent decision-making system, the equipment in the tea garden is automatically controlled, including the irrigation system, temperature control equipment, and ventilation devices, to adjust and optimize the green tea growing environment; S7. Provide users with a visual monitoring interface on the cloud platform to display the environmental status of the tea garden and push optimization suggestions and alarm information based on the analysis results.

2. The method for intelligent monitoring of green tea growth environment based on adaptive sensor network according to claim 1, characterized in that: The S1 comprises the following steps: S11. Deploy multiple sensor nodes in the green tea planting area, each sensor node including a temperature sensor, a humidity sensor, a light sensor, a soil moisture sensor, and a carbon dioxide concentration sensor; S12, the temperature sensor of each sensor node is used to collect temperature data of the green tea growing environment in real time, the humidity sensor is used to collect humidity data of the ambient air, the light sensor is used to measure the light intensity of the green tea planting area, the soil moisture sensor is used to monitor the moisture content in the soil, and the carbon dioxide concentration sensor is used to detect the carbon dioxide concentration in the air; S13, each sensor node communicates through a preset wireless communication protocol and transmits data to the control node. The communication protocol includes LoRa and NB-IoT; S14, the layout of sensor nodes is evenly distributed in the tea garden area, and multiple monitoring points are selected for data collection according to the geographical and climatic conditions of the tea garden; S15. The sensor node dynamically adjusts its working state according to environmental changes and battery power. The sensor node works in a low-power mode and enters a standby mode after collecting a certain amount of data, thereby reducing the energy consumption of the system and extending the working time of the sensor node.

3. The method for intelligent monitoring of green tea growth environment based on adaptive sensor network according to claim 1, characterized in that: The S2 comprises the following steps: S21, each sensor node calculates the remaining battery power E in real time based on environmental changes and power information remaining ; S22, when E remaining Below the preset threshold E threshold When the sensor node automatically enters the low power consumption mode and adjusts its data collection frequency, the collection frequency is calculated by the adaptive algorithm and is set as: f new =f initial ·a; Among them, f new is the new acquisition frequency, f initial is the initial acquisition frequency, α is the adjustment factor, α<1; S23, sensor nodes select the optimal transmission path P according to network quality and data transmission load optimal The transmission path is selected based on an adaptive algorithm, taking into account the signal strength S between the current nodes. signal , Data packet transmission delay D delay and network load L load , the preferred degree V of the transmission path is calculated by the following formula path : Select V path The largest path is taken as the optimal transmission path; S24, in the process of communication between nodes, the transmission rate R is used transmission Dynamically adjust the transmission rate, the transmission rate is based on the network load L load and delay D delay Through adaptive algorithm adjustment, it is set to: Among them, R max is the maximum transmission rate, β is the adjustment coefficient; S25, the node obtains the time to enter the standby state through an adaptive algorithm based on the real-time measured environmental data and network quality information. The standby time T standby According to the current minimum standby time T of the node min 、Network load L load and signal strength S signal Dynamic calculation, set to: T standby =T min +γ·L load +δ·S signal ; Among them, γ and δ are adjustment coefficients.

4. The method for intelligently monitoring the green tea growth environment based on an adaptive sensor network according to claim 1, characterized in that: The S3 comprises the following steps: S31, transmitting the environmental data collected by each sensor node to the data processing center through wireless communication technology, and using edge computing equipment to perform preliminary processing and analysis on the collected environmental data; S32. In the edge computing device, the data is preliminarily cleaned and screened to remove noise and incomplete data to obtain a valid data set D. valid , effective data set D valid Including filtered environmental parameters P filtered , P filtered = {P1,P2,…,P n }, where P i The valid data collected for each corresponding sensor; S33, using edge computing devices to calculate the effective data set D valid Conduct preliminary statistical analysis, including data mean, maximum, minimum, and standard deviation, in order to screen out important monitoring parameters at the preliminary stage; S34, edge computing equipment simplifies environmental data and extracts real-time change trends T trend : Among them, P current is the currently collected environmental data, P previous is the environmental data collected last time, and Δt is the collection interval; S35, the processed valid data set D processed and the changing trend T trend Upload to the cloud platform via adaptive network protocol for more in-depth data analysis and processing.

5. The method for intelligently monitoring the green tea growth environment based on an adaptive sensor network according to claim 1, characterized in that: The S4 comprises the following steps: S41, transmitting the processed data set D in the edge computing device through the wireless communication protocol processed Transmit to the cloud platform, D processed Including the processing results of temperature, humidity, light intensity, and soil moisture data; S42. On the cloud platform, the uploaded environmental data is preprocessed using data cleaning and standardization methods. The standardized data set is recorded as D standardized , D standardized Each environmental parameter P normalized Normalization is performed using the following formula: Among them, P original is the original data, μ is the mean of the parameter, and σ is the standard deviation of the parameter; S43, using K-means clustering algorithm to standardize the data set D standardized Processing is performed to extract key environmental features F key , including temperature and humidity changes, soil moisture fluctuations, and periodic changes in light intensity; S44. Based on data mining technology, association rule analysis and time series analysis methods are used to explore the inherent laws of environmental data and obtain the correlation matrix M between environmental parameters. correlation , M correlation Each correlation coefficient r ij Represents environmental parameter P i With P j The correlation matrix is ​​calculated by the following formula: Among them, cov(P i ,P j ) is the environmental parameter P i With P j The covariance of i ) and σ(P j ) is the standard deviation of the corresponding parameter.

6. The method for intelligent monitoring of green tea growth environment based on adaptive sensor network according to claim 1, characterized in that: The S5 comprises the following steps: S51. Receive processed data set D from the cloud platform standardized , including real-time collected environmental data and historical record data; S52. Combine historical data and real-time data to establish a prediction model M using graph neural network combined with time series modeling GNN , M GNN The relationship between sensors is processed through the graph structure, while capturing the temporal characteristics of environmental data. The calculation process is: in, represents the hidden state of node i in the k+1th layer, N(i) is the set of neighbor nodes of node i, c ij is the normalization coefficient, W (k) is the weight matrix of the kth layer, b (k) is the bias term of the kth layer, σ(.) is the sigmoid activation function, is the output of the k+1th layer; S53. In the graph neural network, the interdependence of sensor nodes is modeled through the graph structure to generate the prediction result Y of the future environmental state of each node. future ; S54. Based on the output of the graph neural network, a multi-layer convolutional neural network is used to extract features from time series data to obtain more accurate predictions of future environmental trends. future : Among them, x i is the feature of the input environment data at the i-th moment, w i is the convolution kernel weight; S55. Combine the spatial dependency of the graph neural network and the temporal features extracted by the convolutional neural network to generate a comprehensive prediction result Y for the green tea growth environment predict , and obtain the future environmental change trend T predict : Among them, Y current is the current environmental data, Δt is the time span; S56. Generate targeted management suggestions based on forecast results and change trends. recommendation , including optimizing irrigation strategies and regulating temperature and humidity measures.

7. The method for intelligently monitoring the green tea growth environment based on an adaptive sensor network according to claim 1, characterized in that: The S6 comprises the following steps: S61. Based on the prediction model M GNN The generated future environmental trend forecast T future and current environmental data Y current , to adjust the environmental equipment in the tea garden in real time; S62, according to soil moisture P soil , air humidity P humidity , Temperature P temperature Predict the changing trend and use the following control algorithms to adjust the irrigation system: Among them, I control is the irrigation intensity, k soil is the irrigation adjustment coefficient, P soil,target is the target soil moisture, P soil is the current soil moisture; S63, according to the temperature P temperature and humidity P humidity The predicted changes are used to adjust the working status of the temperature control equipment and ventilation system, and the fuzzy control algorithm is used to achieve the balance of temperature and humidity: U temperature =λ1·(P temperature-t -P temperature )+λ2·(P humidity-t -P humidity ); Among them, U temperature is the regulation output of the temperature control device, λ1 and λ2 are control coefficients, P temperature-t is the target temperature, P humidity-t is the target humidity; S64. According to the predicted environmental trends and the current working status of the equipment, the operation order and priority of multiple equipment are adjusted through the optimization algorithm, and the equipment is controlled using the priority scheduling model: Among them, C device For the selected device, w i is the weight coefficient of the device, is the contribution value of the i-th device to the future environmental trend; S65. According to the system control feedback, the control parameters are adjusted in real time to optimize the operation of the irrigation, temperature control and ventilation systems.

8. The method for intelligently monitoring the green tea growth environment based on an adaptive sensor network according to claim 1, characterized in that: The S7 comprises the following steps: S71. Create a real-time data display interface on the cloud platform, compare the real-time environmental data of the sensor node with the historical data through data visualization technology, and generate a trend analysis chart T trend,display , the trend analysis chart shows how environmental data changes over time, and calculates the trend change rate according to the following formula: Among them, P current is the data at the current time point, P previous is the data of the previous time point, Δt is the time difference; S72, according to the prediction model M GNN And the prediction result T future , the system automatically generates and pushes optimization suggestions A recommendation,display , recommendations include adjusting irrigation, temperature control and ventilation operations: A recommendation,display =f(T future ,P current ,P target ); Among them, T future is the predicted environmental trend, P target is the target environment parameter, f(.) is the mapping function; S73. Displaying alarm information A in the user interface alert , when a certain environmental data P current Exceeding the set threshold P threshold The alarm is automatically triggered when: Among them, Alert means that the alarm is triggered, and No Alert means that the alarm is not triggered.

Images