Paddy field environment monitoring system and method based on intelligent sensor network

Abstract

The invention relates to a rice field environment monitoring system and method based on an intelligent sensor network. The system comprises a sensor module, a field data transmission module, a data calculation module, a network data transmission module, a cloud platform and a user terminal. The sensor module collects rice field environment parameters and rice field images; the field data transmission module transmits the rice field environment parameters and the rice field image to the data calculation module; the data calculation module obtains rice field environment monitoring data and disease and pest classification scores based on the rice field environment parameters and the rice field images; the network data transmission module transmits the data information to the cloud platform; the cloud platform stores and displays the data and analyzes disease and pest classification scores to obtain disease and pest early warning information; the user terminal carries out remote monitoring, disease and pest early warning and irrigation control. The system has the advantages of real-time data transmission, intelligent monitoring, low energy consumption and the like, and the scientificity, accuracy and working efficiency of rice planting management are greatly improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of paddy field monitoring, and particularly relates to a paddy field environment monitoring system and method based on an intelligent sensor network. Background Art

[0002] Traditional paddy field agricultural environment monitoring systems usually deploy temperature, humidity or moisture sensors, lacking comprehensive monitoring of multiple parameters such as light, CO2 concentration, water level, etc., and unable to comprehensively reflect the dynamics of the paddy field microenvironment. Secondly, traditional paddy field agricultural environment monitoring systems lack visual perception ability and are unable to timely identify and warn of the occurrence of pests and diseases. It still stays in manual visual observation, with low efficiency and easy to miss inspections. Traditional monitoring systems have the disadvantages of high latency and low timeliness in data transmission and processing, and cannot meet the characteristics of the complex and changeable microenvironment of current paddy field agriculture. At the same time, in terms of sustainability, traditional sensors usually use wired communications such as RS485, and the humid paddy fields are prone to corrosion of the lines, resulting in high maintenance costs. Summary of the Invention

[0003] To solve the problems existing in the prior art, the present invention provides a paddy field environment monitoring system and method based on an intelligent sensor network, aiming to achieve precise collection of paddy field environment parameters, intelligent management of pest and disease warning and irrigation.

[0004] To achieve the above object, the present invention provides the following solutions:

[0005] A paddy field environment monitoring system based on an intelligent sensor network includes a sensor module, a field data transmission module, a data calculation module, a network data transmission module, a cloud platform and a user terminal;

[0006] The sensor module is used to collect paddy field environment parameters and paddy field images;

[0007] The field data transmission module is used to transmit the paddy field environment parameters and the paddy field images to the data calculation module;

[0008] The data calculation module is used to obtain paddy field environment monitoring data based on the paddy field environment parameters and obtain pest and disease classification scores based on the paddy field images;

[0009] The network data transmission module is used to transmit the paddy field environment monitoring data and the pest and disease classification scores to the cloud platform;

[0010] The cloud platform is used to store and display the paddy field environment monitoring data and the pest and disease classification scores, and analyze the pest and disease classification scores to obtain pest and disease warning information;

[0011] The user terminal is used for remote monitoring, pest warning, and irrigation control based on the paddy field environment monitoring data and the pest warning information.

[0012] Preferably, the sensor module includes a light sensor, a water level sensor, a CO2 sensor, a temperature and humidity sensor, a soil moisture sensor, and a camera;

[0013] The light sensor, water level sensor, CO2 sensor, temperature and humidity sensor, and soil moisture sensor are used to collect the paddy field environment parameters;

[0014] The paddy field environment parameters include the light intensity, water level, CO2 concentration, temperature and humidity, and soil moisture of the paddy field;

[0015] The camera is used to collect the paddy field images.

[0016] Preferably, the field data transmission module uses W5500 and STM32 to transmit the paddy field environment parameters and the paddy field images to the data calculation module.

[0017] Preferably, the data calculation module includes an edge computing unit, an AI recognition unit, and an intelligent irrigation scheduling unit;

[0018] The edge computing unit is used to process the paddy field environment parameters and obtain the paddy field environment monitoring data;

[0019] The AI recognition unit is used to process the paddy field images and obtain the pest classification scores;

[0020] The intelligent irrigation scheduling unit is used to obtain the final irrigation water volume based on the paddy field environment monitoring data.

[0021] Preferably, the edge computing unit is used to process the paddy field environment parameters and obtain the paddy field environment monitoring data, including:

[0022]

[0023] where y k is the filtered output at time k, z k-i is the measured value of the paddy field environment parameters at time k - i, and N is the sliding window size.

[0024] Preferably, the AI recognition unit is used to process the paddy field images and obtain the pest classification scores, including:

[0025] logits = MobileNetV2(I tensor )

[0026] where logits is the classification score, and Itensor is a PyTorch tensor of the paddy field image.

[0027] Preferably, the intelligent irrigation scheduling unit is used to obtain the final irrigation water volume based on the paddy field environment monitoring data, including:

[0028]

[0029] where WA is the final irrigation water volume, w is the irrigation water volume value, and μ WA (w) is the membership degree after aggregation of the paddy field environment monitoring data.

[0030] Preferably, the network data transmission module uses FL-EM7688 as the main control unit, combines with a 4G Internet access module, and transmits the paddy field environment monitoring data, the pest classification score, and the final irrigation water volume to the cloud platform through the HTTP protocol.

[0031] Preferably, the cloud platform includes an AI early warning system, which analyzes the pest classification score to obtain pest early warning information.

[0032] The present invention also provides a paddy field environment monitoring method based on an intelligent sensor network, which is implemented by applying the foregoing paddy field environment monitoring system based on an intelligent sensor network, and includes:

[0033] S1. Collect paddy field environment parameters and paddy field images;

[0034] S2. Obtain paddy field environment monitoring data based on the paddy field environment parameters, and obtain pest classification scores based on the paddy field images;

[0035] S3. Transmit the paddy field environment monitoring data and the pest classification scores to the cloud platform;

[0036] S4. Store and display the paddy field environment monitoring data and the pest classification scores through the cloud platform, and analyze the pest classification scores to obtain pest early warning information;

[0037] S5. Perform remote monitoring, pest early warning, and irrigation control through a user terminal.

[0038] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0039] 1. Multi-parameter monitoring: Compared with traditional systems that only monitor temperature, humidity, or soil moisture, this system integrates multiple environmental parameters such as light, CO2, and water level, providing more comprehensive dynamic data of the paddy field microenvironment;

[0040] 2. Intelligent pest detection: Integrating cameras and AI image recognition, it realizes automatic monitoring and early warning of pests, reduces manual dependence, and significantly improves monitoring efficiency and accuracy;

[0041] 3. High-efficiency data transmission: The wireless heterogeneous network (LoRaWAN / NB-IoT + 4G) is adopted to avoid the corrosion problem of traditional wired communication in humid environments and reduce maintenance costs;

[0042] 4. Edge computing ability: Local terminal devices can perform intelligent analysis, reduce dependence on cloud computing, improve data processing efficiency, and reduce latency;

[0043] 5. Strong real-time data acquisition: The intelligent sensor network can continuously and automatically monitor environmental parameters, reducing manual intervention;

[0044] 6. Convenient remote management: Supports access from mobile and PC terminals, enabling remote monitoring anytime and anywhere;

[0045] 7. Low power consumption and long battery life: Optimized communication protocols and low-power sensing technologies are adopted to extend the device operation time. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0047] Figure 1 It is a schematic diagram of the system framework of Embodiment 1 of the present invention;

[0048] Figure 2 It is a daily change monitoring diagram of the temperature of the surface layer (dashed line) and the temperature at 50 cm height (solid line) of paddy field soil in Embodiment 1 of the present invention; among them, a is the irrigation treatment of 0 - 1 cm; b is the irrigation treatment of 4 - 6 cm; c is the irrigation treatment of 8 - 10 cm;

[0049] Figure 3 It is a flowchart of the method of Embodiment 2 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the protection scope of the present invention.

[0051] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below with reference to the drawings and specific embodiments.

[0052] Example 1

[0053] As Figure 1 shown, the present invention provides a paddy field environment monitoring system based on an intelligent sensor network, including: a sensor module, a field data transmission module, a data calculation module, a network data transmission module, a cloud platform, and a user terminal;

[0054] The sensor module is used to collect paddy field environment parameters and paddy field images;

[0055] The field data transmission module is used to transmit the paddy field environment parameters and paddy field images to the data calculation module;

[0056] The data calculation module is used to obtain paddy field environment monitoring data based on the paddy field environment parameters, obtain pest classification scores based on the paddy field images, and obtain the final irrigation water volume based on the paddy field environment monitoring data;

[0057] The network data transmission module is used to transmit the paddy field environment monitoring data, pest classification scores, and final irrigation water volume to the cloud platform;

[0058] The cloud platform is used to store and display the paddy field environment monitoring data, pest classification scores, and final irrigation water volume, and analyze the pest classification scores to obtain pest warning information;

[0059] The user terminal is used to perform remote monitoring, pest warning, and irrigation control based on the paddy field environment monitoring data, pest warning information, and final irrigation water volume.

[0060] The sensor module includes a light sensor, a water level sensor, a CO2 sensor, a temperature and humidity sensor, a soil moisture sensor, and a camera;

[0061] Among them, the temperature and humidity sensor adopts a high-precision model with an accuracy of ±0.3°C / ±2%RH to monitor the air temperature and humidity in real time; the measurement range of the light sensor is 0.1 to 40,000 lux; the measurement accuracy of the soil moisture sensor reaches ±2%; the CO2 sensor adopts NDIR technology with a measurement accuracy of ±50 ppm; the water level sensor adopts radar measurement with a resolution of 0.1 mm to 1 mm.

[0062] The light sensor, water level sensor, CO2 sensor, temperature and humidity sensor, and soil moisture sensor are used to collect paddy field environment parameters;

[0063] The paddy field environment parameters include the light intensity, water level, CO2 concentration, temperature and humidity, and soil moisture of the paddy field;

[0064] The camera is used to collect paddy field images.

[0065] The field data transmission module uses W5500 + STM32 to transmit paddy field environmental parameters and paddy field images to the data calculation module.

[0066] The data calculation module includes an edge computing unit, an AI recognition unit, and an intelligent irrigation scheduling unit;

[0067] The edge computing unit is used to process paddy field environmental parameters and obtain paddy field environmental monitoring data;

[0068] The AI recognition unit is used to process paddy field images and obtain pest classification scores;

[0069] The intelligent irrigation scheduling unit is used to obtain the final irrigation water volume based on paddy field environmental monitoring data.

[0070] The edge computing unit is used to process paddy field environmental parameters and obtain paddy field environmental monitoring data, specifically as follows:

[0071] The sliding average is used to process the temperature and humidity, water level, light intensity, and CO2 concentration in the paddy field environmental parameters. By averaging the recent several measurement values, the data is smoothed to reduce the influence of noise. The sliding average (also known as moving average or mean filtering) is a time series smoothing method that estimates the current state by calculating the average of the recent N measurement values:

[0072]

[0073] where y k is the filtered output (smoothed value) at time k, z k-i is the measured value of the paddy field environmental parameter at time k - i, and N is the size of the sliding window. For example, N = 5 means taking the average of the recent 5 measurement values. The sliding average needs to store the recent N measurement values and is implemented using a circular buffer.

[0074] The AI recognition unit is used to process paddy field images and obtain pest classification scores, specifically as follows:

[0075] The existing MobileNetV2 model is used for image recognition. An RGB camera is used to collect paddy field images with a resolution of 1280x720 and a frame rate of 1 frame / min. The existing standard rice pest dataset is used for model training. The paddy field images are adjusted to the input size of MobileNetV2 (224×224):

[0076]

[0077] where I represents the paddy field image, and I resized represents the paddy field image after size transformation.

[0078] To improve the training effect and stability of the model, the data is normalized:

[0079]

[0080] where μ = [0.485, 0.456, 0.406] (RGB channel mean), σ = [0.229, 0.224, 0.225] (RGB channel standard deviation). The generalization ability of the model is improved by random flipping, rotation, and brightness adjustment, enabling the model to learn the same pest and disease characteristics from images with different transformations:

[0081] I augmented = augment(I normalized )

[0082] I normalized is the data after normalization, and augment() represents the data augmentation operation. Convert these processed images into PyTorch tensors:

[0083]

[0084] Input the preprocessed images into the MobileNetV2 model to output classification scores (logits):

[0085] logits = MobileNetV2(I tensor )

[0086] Convert logits into probabilities through the Softmax function:

[0087]

[0088] P[c] is the predicted probability for class c.

[0089] Calculate the loss, and use the cross-entropy loss function to calculate the error between the predicted value and the true label:

[0090]

[0091] y c : true label, c: number of classes (4 classes).

[0092] Calculate the gradient of the loss with respect to the model parameters:

[0093]

[0094] θ: model parameters

[0095] Use the Adam optimizer to update the model parameters:

[0096]

[0097] η: Learning rate (0.001).

[0098] After each round of training, use the validation set to evaluate the model performance and calculate the accuracy rate:

[0099]

[0100] The intelligent irrigation scheduling unit is used to obtain the final irrigation water volume based on the paddy field environment monitoring data, as follows:

[0101] Adopt the intelligent irrigation scheduling algorithm, and use the triangular membership function to convert the input variables (soil moisture (SM), temperature (T), humidity (H)) into membership degrees. In the whole set of algorithms, three fuzzy sets are set for the three input variables respectively: Low, Medium, High. At the same time, the output variable, that is, the irrigation water volume, is also divided into three fuzzy sets: None, Low, High.

[0102] Convert the input variables into membership degrees and use the triangular membership function:

[0103]

[0104] In the formula, x: input value. a, b, c: three vertices of the triangular membership function.

[0105] Set the vertex values as shown in Table 1 according to experience:

[0106] Table 1

[0107] Level (a, b, c) Soil Moisture (SM) Low (0,0,40) Medium (20,50,80) High (60,100,100) Temperature (T) Low (0,0,20) Medium (10,25,40) High (30,50,50) Humidity (H) Low (20,20,40) Medium (30,55,80) High (70,90,90) Irrigation Water Amount (WA) None (0,0,20) Less (10,30,50) More (40,100,100)

[0108] Define the fuzzy rules (a total of 3×3×3 = 27 rules) as shown in Table 2:

[0109] Table 2

[0110]

[0111]

[0112] Fuzzy inference, calculate the rule activation degree: For each rule, use the minimum operation to calculate the degree:

[0113] α i = mon(μ SM (x), μ T (y), μ H (z))

[0114] α i : Activation degree of the i-th rule. μ SM , μ T , μ H: Membership degrees of soil moisture, temperature, and humidity.

[0115] Use the maximum operation to aggregate and convert the fuzzy output into the specific irrigation water volume using the centroid method:

[0116]

[0117] Where, WA is the final irrigation water volume (liters per minute), w is the irrigation water volume value (0 - 100), and μ WA (w) is the membership degree after aggregating the paddy field environment monitoring data.

[0118] Irrigation time (Duration):

[0119]

[0120] Q is the pump flow rate (10 liters per minute).

[0121] The network data transmission module uses FL-EM7688 as the main control unit, combines with a 4G Internet access module, and transmits the paddy field environment monitoring data, pest classification scores, and final irrigation water volume to the cloud platform through the HTTP protocol.

[0122] The cloud platform uses a MySQL / NoSQL database to store the paddy field environment monitoring data, and performs data visualization display through a Web console and a mobile APP. Users can remotely view the paddy field environment monitoring data, pest classification scores, and select to manually or automatically control the irrigation system. The cloud platform also includes an AI early warning system, which combines long-term data analysis to predict possible pest risks and notifies users through text messages, APP push, etc.

[0123] The user terminal supports multi-platform access, including a mobile APP and a PC console. Users can receive alarm notifications, view real-time data, and remotely manage sensor nodes and the irrigation system through the user terminal.

[0124] The present invention provides scientific decision-making support for paddy field irrigation management by real-time scheduling irrigation through a wireless intelligent sensor array composed of a light sensor, a water level sensor, a CO2 sensor, a temperature and humidity sensor, and a soil moisture sensor, based on soil moisture and temperature data.

[0125] The present invention can adapt to different types of paddy field environments, realizes rapid deployment through wireless transmission, avoids the problem of corrosion of wired communication lines due to a humid environment, and the system has high scalability. Other types of sensors, such as a wind speed and direction sensor, a pH sensor, etc., can be added according to requirements to meet more refined agricultural environment monitoring needs.

[0126] Figure 2Shows the monitoring results of the daily temperature changes of the paddy field soil surface layer (dashed line) and at a height of 50 cm (solid line) under different irrigation treatments. It can be seen from the figure that the temperature change trends of the three groups of treatments during the day are generally the same, that is, the temperature at night is relatively low, and during the day, as the surrounding environmental temperature begins to rise with solar radiation and reaches the highest value at around 12:00 noon. However, when the water layer in the paddy field is relatively shallow, the temperature difference between the soil surface layer and the height of 50 cm above is greater than that of the other two groups of treatments, and the diurnal temperature difference at 50 cm is also much larger than that of the other two groups of treatments. The data collection and transmission have achieved the ideal effect of the expected monitoring environment.

[0127] Example Two

[0128] As Figure 3 shown, the present invention also provides a paddy field environment monitoring method based on an intelligent sensor network. Applying the system described in Example One, it includes the following steps:

[0129] S1. Collect paddy field environment parameters and paddy field images;

[0130] S2. Obtain paddy field environment monitoring data based on the paddy field environment parameters, obtain pest and disease classification scores based on the paddy field images, and determine the final irrigation water volume based on the paddy field environment monitoring data;

[0131] S3. Transmit the paddy field environment monitoring data, the pest and disease classification scores, and the final irrigation water volume to the cloud platform;

[0132] S4. Store and display the paddy field environment monitoring data, the pest and disease classification scores, and the final irrigation water volume through the cloud platform, and analyze the pest and disease classification scores to obtain pest and disease early warning information;

[0133] S5. Conduct remote monitoring, pest and disease early warning, and irrigation control through the user terminal.

[0134] The above-described embodiments are only descriptions of the preferred embodiments of the present invention and do not limit the scope of the present invention. Without departing from the design spirit of the present invention, various deformations and improvements made by those of ordinary skill in the art to the technical solutions of the present invention shall fall within the protection scope determined by the claims of the present invention.

Claims

1. A paddy field environment monitoring system based on an intelligent sensor network, characterized in that It includes a sensor module, a field data transmission module, a data calculation module, a network data transmission module, a cloud platform and a user terminal; The sensor module is used to collect paddy field environment parameters and paddy field images; The field data transmission module is used to transmit the paddy field environment parameters and the paddy field images to the data calculation module; The data calculation module is used to obtain paddy field environment monitoring data based on the paddy field environment parameters, and obtain pest classification scores based on the paddy field images; The network data transmission module is used to transmit the paddy field environment monitoring data and the pest classification scores to the cloud platform; The cloud platform is used to store and display the paddy field environment monitoring data and the pest classification scores, and analyze the pest classification scores to obtain pest warning information; The user terminal is used to perform remote monitoring, pest warning and irrigation control based on the paddy field environment monitoring data and the pest warning information.

2. The paddy field environment monitoring system based on an intelligent sensor network according to claim 1, characterized in that The sensor module includes a light sensor, a water level sensor, a CO2 sensor, a temperature and humidity sensor, a soil moisture sensor and a camera; The light sensor, the water level sensor, the CO2 sensor, the temperature and humidity sensor and the soil moisture sensor are used to collect the paddy field environment parameters; The paddy field environment parameters include the light intensity, water level, CO2 concentration, temperature and humidity, and soil moisture of the paddy field; The camera is used to collect the paddy field images.

3. The paddy field environment monitoring system based on the intelligent sensor network according to claim 1, characterized in that, The field data transmission module uses W5500 and STM32 to transmit the paddy field environment parameters and the paddy field images to the data calculation module.

4. The paddy field environment monitoring system based on an intelligent sensor network according to claim 1, characterized in that The data calculation module includes an edge computing unit, an AI recognition unit and an intelligent irrigation scheduling unit; The edge computing unit is used to process the paddy field environment parameters and obtain the paddy field environment monitoring data; The AI recognition unit is used to process the paddy field images and obtain the pest classification scores; The intelligent irrigation scheduling unit is used to obtain the final irrigation water volume based on the paddy field environment monitoring data.

5. The paddy field environment monitoring system based on an intelligent sensor network according to claim 4, characterized in that, The edge computing unit is used to process the paddy field environment parameters to obtain the paddy field environment monitoring data, including: where y k is the filtered output at time k, z k-i is the measured value of the paddy field environmental parameters at time k - i, and N is the size of the sliding window.

6. The paddy field environment monitoring system based on an intelligent sensor network according to claim 4, characterized in that, The AI recognition unit is used to process the paddy field images to obtain the pest classification scores, including: logits = MobileNetV2(I tensor ) where logits are classification scores, and I tensor is a PyTorch tensor of the paddy field image.

7. The paddy field environment monitoring system based on an intelligent sensor network according to claim 4, characterized in that, The intelligent irrigation scheduling unit is used to obtain the final irrigation water volume based on the paddy field environment monitoring data, including: Among them, WA is the final irrigation water volume, w is the value of the irrigation water volume, and μ WA (w) is the membership degree after aggregating the paddy field environmental monitoring data.

8. The paddy field environment monitoring system based on an intelligent sensor network according to claim 4, characterized in that The network data transmission module uses FL-EM7688 as the main control unit, combines with a 4G Internet access module, and transmits the paddy field environment monitoring data, the pest classification scores and the final irrigation water volume to the cloud platform through the HTTP protocol.

9. The paddy field environment monitoring system based on an intelligent sensor network according to claim 1, characterized in that, The cloud platform shown includes an AI warning system, and analyzes the pest classification scores to obtain pest warning information.

10. A rice paddy environment monitoring method based on an intelligent sensor network, characterized in that, Implemented by using the paddy field environment monitoring system based on an intelligent sensor network according to any one of claims 1-9, the method includes: S1. Collect paddy field environment parameters and paddy field images; S2. Obtain paddy field environmental monitoring data based on the paddy field environmental parameters, and obtain pest and disease classification scores based on the paddy field images; S3. Transmit the paddy field environmental monitoring data and the pest and disease classification scores to the cloud platform; S4. Store and display the paddy field environmental monitoring data and the pest and disease classification scores through the cloud platform, and analyze the pest and disease classification scores to obtain pest and disease early warning information; S5. Conduct remote monitoring, pest and disease early warning, and irrigation control through the user terminal.

Images