Intelligent indoor plant environment maintenance and growth monitoring guarantee system

Abstract

An intelligent indoor plant environment maintenance and growth monitoring guarantee system is characterized in that a sensing layer is used as an infrastructure layer and is responsible for acquiring data from an environment or a physical world and realizing data acquisition, equipment monitoring and information transmission; the service layer is responsible for processing, storing, integrating and analyzing the data transmitted by the sensing layer and providing various service interfaces; and the application layer and the final user interaction interface layer are responsible for displaying and utilizing data and calculation results provided by the service layer, providing various specific application functions for users, and covering user operation interfaces, data presentation and logic processes related to services. The plant growth environment is monitored in real time and intelligently adjusted, it is ensured that plants grow under the optimal growth condition, the plant growth quality and ornamental value are improved, through intelligent adjustment and data analysis, manual intervention and maintenance cost are reduced, maintenance efficiency is improved, a visual and easy-to-use user interface and personalized maintenance suggestions are provided, and the method is suitable for popularization and application. And the user experience and satisfaction are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of plant environment, and in particular to an intelligent indoor plant environment maintenance and growth monitoring and guarantee system. Background Art

[0002] Indoor plants are widely used in decoration, air purification and psychological relaxation. However, due to the complexity of the indoor environment and the limitations of manual management, the growth and maintenance of plants often face some challenges. Traditional plant maintenance methods rely on manual judgment and experience, and it is difficult to precisely control the growth environment of plants, resulting in unstable plant growth states or even death. Therefore, it is very necessary to develop an intelligent indoor plant environment maintenance and growth monitoring and guarantee system. Summary of the Invention

[0003] In order to solve the problems existing in the above-mentioned prior art, the present invention provides an intelligent indoor plant environment maintenance and growth monitoring and guarantee system.

[0004] The technical solution adopted by the present invention to solve its technical problems is as follows: An intelligent indoor plant environment maintenance and growth monitoring and guarantee system, which realizes real-time monitoring and intelligent adjustment of the plant growth environment through integrating sensors, intelligent control modules, data analysis algorithms and visual interaction interfaces, so as to guarantee the healthy growth of plants; it includes a perception layer, a service layer and an application layer; wherein: The perception layer, as the infrastructure layer, is responsible for obtaining data from the environment or the physical world, realizing data collection, device monitoring and information transmission; the perception layer includes a sensor module, an automatic adjustment device and a video monitoring camera; the sensor module includes a temperature and humidity sensor, a light sensor, a soil sensor, a pH sensor, and a water quality environment sensor. The sensor module monitors the growth environment data of plants in real time, mainly including temperature, humidity, light intensity, soil humidity, pH value, conductivity, and turbidity; the automatic adjustment device performs relevant environment adjustment actions according to the monitored data and analysis results, and automatically adjusts the growth environment of indoor plants; the video monitoring camera is used to monitor the real-time picture of plant growth; The service layer is responsible for collecting, acquiring, processing, storing, integrating, and analyzing the data transmitted from the perception layer, and providing various service interfaces. The service layer includes an intelligent control module for automatically adjusting environmental parameters according to sensor data, a data storage and management module, a data analysis algorithm module, and an edge computing gateway. The intelligent control module automatically adjusts the environmental parameters such as temperature, humidity, and light for the plant growth environment according to the data provided by the sensor module. The data storage and management module is responsible for storing sensor data, analysis results, user configurations, and operation logs to ensure long-term data preservation and traceability. The data storage and management module also includes the historical records of sensor data, maintenance suggestions, and the historical records of user adjustment operations. The data analysis algorithm module preprocesses, filters, and analyzes the raw data collected by the sensors. Based on the analysis results, it predicts the plant growth trend, evaluates the health status of the current growth environment, and generates maintenance suggestions. The data analysis algorithm module comes with AI algorithms that process sensor data through the algorithms to detect in real time whether the plant growth environment is in the best state, predict the demand changes of the plants, and make intelligent decisions. The edge computing gateway is responsible for the communication between different modules, including the uploading of sensor data and the sending of control instructions. It supports high-precision real-time data acquisition of multi-source heterogeneous sensors with different functions and different communication protocols. The communication protocols include ModbusRTU, ModbusTCP, Wi-Fi, 4G / 5G, NB-IOT, Bluetooth, Zigbee, and MQTT, providing edge-side computing power support for the environmental automatic adjustment algorithm. The application layer is the end-user interaction interface layer, which is responsible for presenting and utilizing the data and calculation results provided by the service layer, and providing various specific application functions for users, covering the user operation interface, data presentation, and business-related logical processes. The user interaction module provides a user-friendly interface for users to view the plant growth environment data, status, and maintenance suggestions in real time. The user interaction module can manually adjust the environmental parameters, automatically adjust the environmental parameters by AI, view historical data, and receive maintenance reminders. Users can remotely access the system through mobile phones or computer devices to monitor the plant growth environment in real time and perform necessary controls. Users can view real-time data, plant status, and maintenance suggestions through the mobile phone APP or the PC side, and manually adjust the environmental settings as needed. The system supports remote control, allowing users to intervene at any time and place.

[0005] Compared with the prior art, the advantages of the present invention are as follows: Improve the quality of plant growth: By monitoring and intelligently adjusting the plant growth environment in real time, ensure that the plants grow under the best growth conditions, and improve the quality and ornamental value of plant growth.

[0006] Reduce maintenance costs: Through intelligent adjustment and data analysis, reduce manual intervention and maintenance costs, and improve maintenance efficiency.

[0007] Enhance user experience: Provide an intuitive and easy-to-use user interface and personalized maintenance suggestions to enhance user experience and satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 It is a schematic structural diagram of the perception layer, service layer and application layer of the present invention; Figure 2 It is a schematic structural diagram of the functional components of the present invention; Figure 3 It is a schematic diagram of the method logic structure of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0009] The exemplary embodiments disclosed by the present invention will be described in more detail below with reference to the accompanying drawings.

[0010] An intelligent indoor plant environment maintenance and growth monitoring and guarantee system realizes real-time monitoring and intelligent adjustment of the plant growth environment by integrating sensors, intelligent control modules, data analysis algorithms and visual interaction interfaces, so as to guarantee the healthy growth of plants; it includes a perception layer, a service layer and an application layer.

[0011] As the infrastructure layer, the perception layer is responsible for obtaining data from the environment or the physical world, realizing data collection, device monitoring, and information transmission; the perception layer includes monitoring temperature and humidity sensors, light sensors, soil sensors, pH sensors, water quality environment sensors, an automatic adjustment device for adjusting temperature, humidity and light environment parameters, and a video monitoring camera for monitoring the real-time picture of plant growth.

[0012] The service layer is responsible for processing, storing, integrating, analyzing the data transmitted from the perception layer, and providing various service interfaces; the service layer includes an intelligent control module for automatically adjusting environmental parameters according to sensor data, a data storage and management module, and a data analysis algorithm module.

[0013] The application layer is the end-user interaction interface layer, which is responsible for displaying and utilizing the data and calculation results provided by the service layer, and providing various specific application functions for users, covering the user operation interface, data presentation, and business-related logic processes.

[0014] The intelligent indoor plant environment maintenance and growth monitoring and guarantee system also includes real-time display of plant environment data, display of plant growth status, maintenance suggestions, manual adjustment strategies, and AI automatic adjustment strategies.

[0015] The intelligent indoor plant environment maintenance and growth monitoring and guarantee system includes the following functional components: User Interaction Module: The user interaction module provides a user-friendly interface for users to view the growth environment data, status, and maintenance suggestions of plants in real time. The user interaction module enables manual adjustment of environmental parameters, AI automatic adjustment of environmental parameters, viewing of historical data, and receiving of maintenance reminders. Users can remotely access the system through mobile or computer devices to monitor the plant growth environment in real time and perform necessary controls.

[0016] Intelligent Control Module: The intelligent control module automatically adjusts environmental parameters such as temperature, humidity, and light for the plant growth environment based on the data provided by the sensor module. For example, it adjusts heaters, cooling devices, humidifiers, and dehumidifier equipment through temperature control devices, and controls the brightness and spectrum of lighting devices. According to the water requirements of plants, it automatically supplies appropriate amounts of water and nutrient solution through liquid supply devices to ensure the growth of plants.

[0017] Data Storage and Management Module: The data storage and management module is responsible for storing sensor data, analysis results, user configurations, and operation logs to ensure long-term data preservation and traceability. The data storage and management module also includes historical records of sensor data, maintenance suggestions, and historical user adjustment operations.

[0018] Data Analysis Algorithm Module: The data analysis algorithm module preprocesses, filters, and analyzes the raw data collected by sensors. Based on the analysis results, it predicts the plant growth trend, evaluates the health status of the current growth environment, and generates maintenance suggestions. The data analysis algorithm module comes with AI algorithms that process sensor data to detect in real time whether the plant growth environment is in the optimal state, predict changes in plant requirements, and make intelligent decisions.

[0019] Edge Computing Gateway: The edge computing gateway is responsible for communication between different modules, including uploading sensor data and issuing control instructions. It supports high-precision real-time data collection from multi-source heterogeneous sensors with different functions and communication protocols. The communication protocols include ModbusRTU, ModbusTCP, Wi-Fi, 4G / 5G, NB-IOT, Bluetooth, Zigbee, MQTT, providing edge-side computing support for the environmental automatic adjustment algorithm.

[0020] Sensor Module: The sensor module monitors the growth environment data of plants in real time, mainly including temperature, humidity, light intensity, soil humidity, pH value, conductivity, and turbidity. The sensor module includes environmental monitoring sensors (temperature and humidity sensors, light sensors, CO 2 sensors) and water quality and soil sensors (pH sensors, conductivity sensors, dissolved oxygen sensors, soil temperature and humidity sensors).

[0021] Automatic adjustment device: The automatic adjustment device performs relevant environmental adjustment actions based on the monitored data and analysis results, and automatically adjusts the growth environment of indoor plants. For example: ceiling fans, jet fans, electric windows, mist forest devices, liquid supply devices, central / master and sub-control of fresh air systems, central / master and sub-control of air conditioning systems, heaters, cooling devices, humidifiers, dehumidifiers.

[0022] Intelligent indoor plant environment maintenance and growth monitoring and guarantee system, including the following: Data collection: The sensor module continuously monitors various parameters (such as temperature, humidity, light) in the plant growth environment and transmits the real-time data to the data analysis algorithm module.

[0023] Data analysis: The data analysis algorithm module receives the sensor data, processes and analyzes it, provides maintenance suggestions, and evaluates the health status of the plant growth environment.

[0024] Environmental adjustment: The intelligent control module automatically adjusts the environmental parameters (such as adjusting temperature, humidity, light or water supply) according to the analysis results to ensure that the plants are in the best growth environment.

[0025] User interaction: Users can view real-time data, plant status and maintenance suggestions through the mobile phone APP or PC, and manually adjust the environmental settings as needed. The system supports remote control, allowing users to intervene anytime and anywhere.

[0026] Data storage: All data is stored in the data storage and management module, enabling users to view historical data and operation log information to help understand the growth process of plants.

[0027] Data collection and data analysis include: Data collection: The system monitors the temperature, humidity, and light parameters of indoor plants through the sensor component and collects the real-time data into the database.

[0028] Data processing: The system preprocesses the collected data through the data processing module, including data cleaning and outlier removal operations to ensure the accuracy and reliability of the data.

[0029] Handling of missing data values: Missing values in temperature and humidity data may occur due to sensor failures or communication problems. The method for handling missing values is selected according to the characteristics of the data: Interpolation of time series data: Linear interpolation: When the missing data points are in the time series, the linear interpolation method is used to fill in the missing values. That is, assuming the data points are linear, the missing values are filled according to the linear relationship between the valid data points before and after.

[0030] Spline Interpolation: For more complex time series data, spline interpolation (such as cubic interpolation) can fill in missing values more precisely, especially for non-linear changes.

[0031] Forward Filling / Backward Filling: When the data changes little in a short period, fill in the missing values with the previous or next valid data.

[0032] Statistical Interpolation: For missing values within a specific time period, use the mean or median of that time period to fill. Note that this method is applicable to the situation where the data changes relatively smoothly.

[0033] Duplicate Data Processing: Duplicate data is generated due to multiple acquisitions or data merging, and it needs to be removed by deleting duplicate records.

[0034] Merging Duplicate Values: For each sensor, merge the duplicate data at the same timestamp and take the mean or median.

[0035] Format Standardization: Ensure that all data has consistent units and formats. Common format standardization operations include: ensuring that temperature uses the same unit (such as Celsius); ensuring that the humidity data range is within 0 - 100%; ensuring that the timestamp format is unified.

[0036] Time Series Alignment: In the case of multiple sensors, there are slight deviations in the data acquisition time. For comparative analysis or further processing, the data needs to be aligned to a unified time frequency (such as hourly, daily).

[0037] Outlier Detection and Removal: Outliers are caused by various reasons, such as sensor failures and drastic environmental changes. For environmental acquisition data in a large space over a long time, outlier detection should combine the time series characteristics and spatial characteristics of the data.

[0038] Intelligent Indoor Plant Environment Maintenance and Growth Monitoring Assurance System, the data analysis algorithm module includes: Outlier Detection Based on Statistical Methods: The Z-Score method is suitable for detecting points where the data deviates too much from the mean within a certain time period. Calculate the Z-Score on the data of each sensor.

[0039] Outlier Detection Based on Quartiles is Suitable for Detecting Extreme Values: For time series data, using conventional statistical methods is not precise enough, so time series anomaly detection methods are used.

[0040] Intelligent Indoor Plant Environment Maintenance and Growth Monitoring Assurance System, time series anomaly detection methods include: Moving Average Method: Calculate the average value of the data within a time period using a sliding window and then compare it with the actual value.

[0041] Seasonal decomposition method: Use a seasonal decomposition method (such as STL) to decompose the temperature and humidity data into trend, seasonal, and residual components. Detect anomalies by examining the deviation of the residuals.

[0042] Spatial outlier detection method: Since the data involves multiple sensors (multiple locations), the non-uniformity of the spatial distribution needs to be considered.

[0043] Spatial correlation analysis method: Detect outliers by calculating the correlation between each sensor. If the temperature and humidity values of a certain point are significantly different from those of other points, it is an outlier.

[0044] K-means clustering method: Use a clustering algorithm (such as K-means) to group the spatial data and find the outliers that do not belong to any group. For obvious outliers, directly delete the relevant data rows by deleting the outliers. For some small outliers, handle the outliers through interpolation or fill them with neighboring values. For spatially abnormal points, use a spatial interpolation method (such as Kriging interpolation) to smooth the outliers.

[0045] After data preprocessing, the system extracts features related to plant growth from the data through a feature extraction algorithm, such as the change trend of temperature requirement, the change trend of humidity requirement, and the change trend of light intensity requirement; among them: Extraction of the change trend of temperature requirement: Temperature has a significant impact on plant growth. The change trend of temperature is extracted through the following algorithm: Moving average, the moving average is a commonly used smoothing method for removing noise in the data and revealing the long-term trend. For a temperature time series T(t), calculate the moving average at each time point: ; where n is the size of the moving window. is the average temperature; the trend is reflected by the slope or rate of change of the moving average sequence.

[0046] Linear regression, linear regression is used to fit the time series data and estimate the linear trend of temperature change over time. Use a linear regression model to fit the temperature data: ; where is the slope of the temperature change over time, indicating the rate of temperature change. The slope reflects the trend of temperature rising or falling. The intercept : reflects the starting temperature of the time series.

[0047] Trend decomposition, use the STL or HP filtering method to decompose the temperature time series into trend, seasonal, and residual components; ; the obtained by decomposition is the long-term trend of temperature. The trend component Reflect the long-term trend of temperature changes. Seasonal component : Reflect the periodic fluctuations of temperature.

[0048] Extraction of the changing trend of humidity demand. Humidity has a great impact on the water supply and evaporation process of plants. The method for extracting the humidity change trend is similar to that of temperature, but special attention should be paid to the seasonal fluctuations and diurnal variations of humidity.

[0049] Moving standard deviation. The moving standard deviation is also an effective trend feature, which can reveal the amplitude of humidity fluctuations. Calculate the standard deviation of the humidity time series H(t) over the moving window n: ; where is the mean of the humidity sequence. The moving standard deviation of the humidity change amplitude reflects the change amplitude of humidity. The humidity change trend is extracted by calculating the change rate or slope of the humidity sequence to obtain the long-term trend.

[0050] Exponential smoothing. The exponential smoothing method assigns higher weights to recent data and is suitable for capturing the trend of humidity changing over time. ; where is the smoothing coefficient, usually taking values between 0 and 1. The smoothed humidity value reflects the short-term trend of humidity. The humidity trend change is extracted by the change of the smoothed sequence to obtain the long-term trend of humidity.

[0051] Extraction of the changing trend of light intensity demand. Light intensity directly affects the photosynthesis of plants, so its change trend is equally important. The change of light usually has obvious day-night periodicity. Light intensity data usually has strong periodicity, and the periodic trend of light is extracted through fast Fourier transform (FFT) or periodic analysis. Perform fast Fourier transform on the light intensity L(t) to obtain the frequency spectrum diagram: ; is the frequency. The main frequency component reflects the periodic change of light intensity (such as the day-night cycle). The light cycle obtains the periodic fluctuations of light intensity through frequency analysis.

[0052] Waveform fitting. Wavelet transform is a tool suitable for multi-scale analysis, which can reveal the mutation points and periodic fluctuations in the change of light intensity. Use wavelet transform on the light intensity L(t) to obtain the time-frequency characteristics at different scales. The light fluctuations extract the light change characteristics at different scales. The light mutations identify the mutations and abnormal fluctuations of light intensity.

[0053] Data modeling. The system uses machine learning algorithms to model the extracted features and build a prediction model. Use long short-term memory network (LSTM) for predicting the temperature, humidity and light demand in the plant production environment, which involves dealing with the characteristics of time series data. The following are the steps to realize the prediction of temperature, humidity and light demand in the plant production environment through LSTM.

[0054] Normalization: Since LSTM is sensitive to data scale, it is necessary to normalize the input data. The system uses historical data as the training set to train the established prediction model. By learning the plant growth trends in the historical data, the model is enabled to have prediction capabilities. During the training process of the LSTM model, the following aspects are adjusted to improve the model's performance: Adjust hyperparameters including the number of units in the LSTM layer, learning rate, and batch size. Increase more layers or units: If the model performance is poor, increase the number of LSTM layers or the number of units in each layer to capture more complex temporal dependencies. Regularization: Use Dropout layer or L2 regularization to avoid overfitting. Data augmentation: Increase the training data through data augmentation methods to enhance the robustness of the model.

[0055] Model evaluation: The system uses the validation set or other methods to evaluate the trained model. By evaluating the accuracy and generalization ability of the model, the optimal model is selected. The process of using the validation set for evaluation is as follows: Divide the original dataset into three parts: the training set (Trainingset), the validation set (Validationset), and the test set (Testset). The division ratio is: training set: 60% - 80%; validation set: 10% - 20%; test set: 10% - 20%. Use the training set data to train the prediction model. Use the validation set to evaluate the trained model. Calculate the following metrics to measure the model's performance: Mean Squared Error (MSE) measures the average of the squared errors between the predicted values and the true values. Root Mean Squared Error (RMSE) is the square root of MSE, which can more intuitively represent the error. Mean Absolute Error (MAE) is the average of the absolute errors between the predicted values and the true values. R² (Coefficient of Determination): Measures how well the model fits the data. The closer the value is to 1, the better the fitting effect. Adjust the hyperparameters of the model according to the evaluation results of the validation set to improve the model performance. Use the Grid Search method for hyperparameter optimization. Specify a discrete value range for each hyperparameter, and then the grid search will traverse all combinations, calculate the model performance of each combination, and finally return the best hyperparameter combination. After multiple validations, select the model with the best performance on the validation set and conduct the final test.

[0056] Trend prediction: After the model training is completed, the system uses the real-time monitored data as input and predicts the plant growth trend through the prediction model. According to the prediction results, the system automatically adjusts environmental parameters such as temperature and humidity to optimize the plant growth environment.

[0057] Result display: The system displays the prediction results to the user in the form of charts or other formats, and the user can understand the plant growth trend in real time through the data display interface.

[0058] Through the above algorithm implementation process, the intelligent indoor plant environment maintenance and growth monitoring and guarantee system can predict the growth trend of plants according to the results of data analysis, and realize the automatic adjustment of environmental parameters, so as to provide an optimized growth environment and promote the healthy growth of plants.

[0059] The implementation method of the intelligent indoor plant environment maintenance and growth monitoring and guarantee system includes the following steps: Install the sensor components: Install the temperature, humidity, light, soil, and water quality sensor components in the indoor plant growth area; Install the automatic adjustment device: Deploy the environmental adjustment devices (such as: ceiling fans, jet fans, electric windows, mist forest devices, liquid supply devices, heaters, cooling devices, humidifiers, dehumidifiers); Connect the edge computing gateway: Connect the sensor components and the braking adjustment device to the edge computing gateway to achieve data transmission and processing; Set the threshold parameters: Set the corresponding temperature, humidity, and light threshold parameters according to the growth requirements of the plants; Monitor and adjust: The control module conducts environmental monitoring and automatically adjusts the corresponding parameters according to the real-time data obtained by the sensor components to meet the growth requirements of the plants; Provide water and nutrients: The liquid supply device automatically supplies an appropriate amount of water and nutrient solution according to the needs of the plants; Adjust the light: The lighting device automatically adjusts the light intensity and spectrum according to the light requirements of the plants; Data display: The data display interfaces for interaction on the PC side, Web side, and mobile phone side provide real-time growth environment monitoring data and plant growth status display, so that users can understand the growth situation of the plants.

[0060] The above detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention.

Claims

1. An intelligent indoor plant environment maintenance and growth monitoring system, characterized by: By integrating sensors, intelligent control modules, data analysis algorithms and visual interactive interfaces, real-time monitoring and intelligent adjustment of the plant growth environment can be achieved to ensure the healthy growth of plants. It includes a perception layer, a service layer and an application layer. Among them: As an infrastructure layer, the perception layer is responsible for acquiring data from the environment or the physical world to realize data collection, equipment monitoring, and information transmission. The perception layer includes sensor modules, automatic adjustment devices, and video surveillance cameras. The sensor modules include temperature and humidity monitoring sensors, light sensors, soil sensors, pH sensors, and water quality environment sensors. The sensor modules monitor the plant growth environment data in real time, mainly including temperature, humidity, light intensity, soil moisture, pH value, conductivity, and turbidity. The automatic adjustment device performs relevant environmental adjustment actions based on the monitored data and analysis results, and automatically adjusts the growth environment of indoor plants. Video surveillance cameras are used to monitor real-time images of plant growth. The service layer is responsible for collecting, acquiring, processing, storing, integrating, and analyzing the data transmitted from the perception layer, and provides various service interfaces; the service layer includes an intelligent control module that automatically adjusts environmental parameters according to sensor data, a data storage and management module, a data analysis algorithm module, and an edge computing gateway; the intelligent control module automatically adjusts the temperature, humidity, and light plant growth environment parameters according to the data provided by the sensor module; the data storage and management module is responsible for storing sensor data, analysis results, user configuration, and operation logs to ensure long-term storage and traceability of data; the data storage and management module also includes sensor data history records, maintenance recommendations, and user adjustment operation history; the data analysis algorithm module preprocesses and filters the raw data collected by the sensor , analysis; based on the analysis results, predict plant growth trends, evaluate the health status of the current growth environment, and generate maintenance recommendations; the data analysis algorithm module comes with an AI algorithm, which processes sensor data through the algorithm to detect in real time whether the plant's growth environment is in the best state, predict changes in plant needs and make intelligent decisions; the edge computing gateway is responsible for communication between different modules, including uploading sensor data and issuing control instructions; it supports high-precision real-time data acquisition of multiple heterogeneous sensors with different functions and different communication protocols. The communication protocols include ModbusRTU, ModbusTCP, Wi-Fi, 4G / 5G, NB-IOT, Bluetooth, Zigbee, and MQTT, providing edge computing support for the automatic environmental adjustment algorithm; The final user interaction interface layer of the application layer is responsible for displaying and utilizing the data and calculation results provided by the service layer, providing users with various specific application functions, covering the user operation interface, data presentation, and business-related logical processes; the user interaction module provides a user-friendly interface, and users can view the plant's growth environment data, status, and maintenance suggestions in real time; the user interaction module manually adjusts environmental parameters, AI automatically adjusts environmental parameters, views historical data, and receives maintenance reminders; users can remotely access the system through mobile phones or computer devices to monitor the plant growth environment in real time and perform necessary controls; users view real-time data, plant status, and maintenance suggestions through mobile phone APP or PC, and manually adjust environmental settings as needed; the system supports remote control, and users can intervene anytime, anywhere.

2. The intelligent indoor plant environment maintenance and growth monitoring system according to claim 1 is characterized by: In the service layer: Data collection: The system monitors the temperature, humidity, and light parameters of indoor plants through sensor components and collects real-time data into the database; Data collection: The sensor module continuously monitors various parameters in the plant growth environment and transmits real-time data to the data analysis algorithm module; Data processing: The system pre-processes the collected data through the data processing module, including data cleaning and outlier removal operations to ensure the accuracy and reliability of the data; Data storage: All data are stored in the data storage and management module, and users can view historical data and operation log information to help understand the growth process of plants; Data integration: including data missing value processing, time series data interpolation, spline interpolation, forward filling / backward filling, statistical interpolation, duplicate data processing, merging duplicate values, format standardization, time series alignment, outlier detection and removal; Data Analysis: The data analysis algorithm module receives sensor data, processes and analyzes it, provides maintenance suggestions, and evaluates the health status of the plant growth environment.

3. The intelligent indoor plant environment maintenance and growth monitoring system according to claim 2 is characterized by: Data integration includes: Data missing value processing: Temperature and humidity data may be missing due to sensor failure or communication problems. The method for processing missing values ​​is selected according to the characteristics of the data: Interpolation of time series data: Linear interpolation: When missing data points are in the time series, linear interpolation is used to fill the missing values; that is, assuming that the data points are linear, the missing values ​​are filled according to the linear relationship between the previous and subsequent valid data points; Spline interpolation: For more complex time series data, spline interpolation can fill in missing values ​​more accurately, especially for nonlinear changes; Forward filling / backward filling: When the data changes little in a short period of time, the missing values ​​are filled with the previous or next valid data; Statistical interpolation: For missing values ​​in a specific time period, use the mean or median of the time period to fill in the missing values; note that this method is suitable for situations where data changes are relatively stable; Duplicate data processing: Duplicate data is generated due to multiple collections or data merging, and needs to be removed. Delete duplicate records; Merge duplicate values: For each sensor, merge duplicate data at the same timestamp and take the mean or median; Format standardization: Ensure that all data units and formats are consistent; common format standardization operations include: ensuring that the temperature uses the same unit; ensuring that the humidity data range is within 0-100%; ensuring that the timestamp format is unified; Timing alignment: In the case of multiple sensors, there is a slight deviation in the data collection time; in order to conduct comparative analysis or further processing, the data needs to be aligned to a unified time frequency; Outlier detection and removal: Outliers are caused by a variety of reasons, such as sensor failure and drastic environmental changes. For large-scale, long-term environmental data collection, outlier detection must combine the temporal and spatial characteristics of the data.

4. The intelligent indoor plant environment maintenance and growth monitoring system according to claim 1 is characterized by: In the service layer: The data analysis algorithm modules include: Outlier detection based on statistical methods: The Z-Score method is suitable for detecting points where the data deviates too much from the mean within a certain period of time; the Z-Score is calculated on the data of each sensor; Quartile-based outlier detection is suitable for detecting extreme values: For time series data, conventional statistical methods are not accurate enough, so time series anomaly detection methods are used.

5. The intelligent indoor plant environment maintenance and growth monitoring system according to claim 4 is characterized by: Time series anomaly detection methods include: Moving average method: Use a sliding window to calculate the average value of data within a time period and then compare it with the actual value; Seasonal decomposition method: Use seasonal decomposition method to decompose temperature and humidity data into trend, seasonal and residual parts; detect anomalies by checking the deviation of residuals; Spatial outlier detection method: Since the data involves multiple sensors, the unevenness of spatial distribution needs to be considered: Spatial correlation analysis method: detect abnormal points by calculating the correlation between each sensor; if the temperature and humidity values ​​of a point are significantly different from those of another point, it is an abnormal value; K-means clustering method: Use clustering algorithm to group spatial data and find outliers that do not fit into any group; for obvious outliers, directly delete the related data rows by deleting the outliers; for some small outliers, use interpolation method to process the outliers by interpolation or filling with neighboring values; for abnormal points in space, use spatial interpolation method to smooth out the outliers.

6. The intelligent indoor plant environment maintenance and growth monitoring system according to claim 1 is characterized by: After data preprocessing, the service layer uses feature extraction algorithms to extract plant growth-related features from the data, such as temperature demand change trends, humidity demand change trends, and light intensity demand change trends; among them: Temperature demand trend extraction: Temperature has a significant impact on plant growth, and the temperature trend is extracted using the following algorithm: Sliding average, sliding average is a commonly used smoothing method to remove noise in data and reveal long-term trends. For a temperature time series T(t), calculate the sliding average of each time point: ; Where n is the size of the sliding window; is the average temperature; the trend is reflected by the slope or rate of change of the sliding average series; Linear regression is used to fit time series data and estimate the linear trend of temperature over time. The linear regression model is used to fit the temperature data: ;in, It is the slope of temperature change over time, indicating the rate of temperature change; slope Reflects the trend of temperature increase or decrease; intercept : The starting temperature of the time series; Trend decomposition, using STL or HP filtering methods to decompose the temperature time series into trend, seasonality and residual components; ; Decomposed is the long-term trend of temperature; the trend part Reflects the long-term trend of temperature change; seasonal component : reflects the periodic fluctuation of temperature; Extraction of humidity demand change trend. Humidity has a great influence on the water supply and evaporation process of plants. The extraction method of humidity change trend is similar to that of temperature, but special attention should be paid to seasonal fluctuations and diurnal changes in humidity. Sliding standard deviation, sliding standard deviation is also an effective trend feature, which can reveal the amplitude of humidity fluctuations; calculate the standard deviation of the humidity time series H(t) on the sliding window n: ;in, is the mean of the humidity series; the sliding standard deviation of humidity change reflects the change in humidity; the humidity change trend extracts the long-term trend by calculating the rate of change or slope of the humidity series; Exponential smoothing, the exponential smoothing method gives higher weight to recent data and is suitable for capturing the trend of humidity changes over time; ;in, is the smoothing coefficient, which usually takes a value between 0 and 1; the smoothed humidity value reflects the short-term trend of humidity; the change of humidity trend extracts the long-term trend of humidity by changing the smoothed sequence; Extraction of light intensity demand change trend: Light intensity directly affects plant photosynthesis, so its change trend is equally important. Changes in light usually have obvious day-night periodicity. Light intensity data usually has strong periodicity, and the periodic trend of light can be extracted through fast Fourier transform or periodic analysis. Perform fast Fourier transform on light intensity L(t) to obtain the spectrum: ; is the frequency; the main frequency component reflects the periodic change of light intensity; the periodic fluctuation of light intensity is obtained through frequency analysis of the light cycle; Waveform fitting, wavelet transform is a tool suitable for multi-scale analysis, which can reveal the mutation points and periodic fluctuations in light intensity changes; wavelet transform is used on light intensity L(t) to obtain time-frequency characteristics at different scales; light fluctuations extract the characteristics of light changes at different scales; sudden changes in light identify sudden changes and abnormal fluctuations in light intensity; Data modeling: The system uses machine learning algorithms to model the extracted features and build a prediction model. It uses long short-term memory networks to predict the temperature, humidity, and light requirements of plant production environments, which involves the characteristics of processing time series data. The following are the steps to achieve prediction of temperature, humidity, and light requirements of plant production environments through LSTM. Normalization: Because LSTM is sensitive to data scale, the input data needs to be normalized. The system uses historical data as a training set to train the established prediction model, and by learning the plant growth trends in historical data, the model has prediction capabilities. During the LSTM model training process, the following aspects are adjusted to improve the performance of the model: hyperparameters including the number of units, learning rate, and batch size of the LSTM layer are adjusted. More layers or units are added: If the model performs poorly, the number of LSTM layers or the number of units per layer is increased to capture more complex temporal dependencies. Regularization uses a Dropout layer or L2 regularization to avoid overfitting. Data enhancement uses data augmentation methods to increase training data and improve the robustness of the model. Model evaluation, the system uses the validation set or method to evaluate the trained model, and selects the optimal model by evaluating the accuracy and generalization ability of the model; the process of using the validation set evaluation is as follows: divide the original data set into three parts: training set, validation set and test set; the division ratio is: training set: 60%80%; validation set: 10%20%; test set: 10%20%; use the training set data to train the prediction model; use the validation set to evaluate the trained model; measure the performance of the model by calculating the following indicators: mean square error measures the average of the square of the error between the predicted value and the true value; root mean square error It is the square root of MSE, which can more intuitively express the error; mean absolute error is the average of the absolute errors between the predicted value and the true value; R² is a measure of how well the model fits the data. The closer the value is to 1, the better the fit is; adjust the model's hyperparameters based on the evaluation results of the validation set to improve model performance; use the grid search method for hyperparameter optimization; specify a discrete value range for each hyperparameter, and then the grid search will traverse all combinations, calculate the model performance of each combination, and finally return the best hyperparameter combination; after multiple verifications, select the model that performs best on the validation set and conduct the final test; Trend prediction: After model training is completed, the system uses real-time monitored data as input to predict plant growth trends through the prediction model. Based on the prediction results, the system automatically adjusts environmental parameters such as temperature and humidity to optimize the plant growth environment. Results display: the system displays the prediction results to the user in the form of charts or graphs, and the user can understand the growth trend of the plant in real time through the data display interface; Through the above algorithm implementation process, the intelligent indoor plant environment maintenance and growth monitoring guarantee system can predict the growth trend of plants according to the results of data analysis, and automatically adjust the environmental parameters, thereby providing an optimized growth environment and promoting the healthy growth of plants.

7. The intelligent indoor plant environment maintenance and growth monitoring system according to claim 1 is characterized by: The user interaction module also includes real-time data display of plant environment, plant growth status display, maintenance suggestions, manual adjustment strategies, and AI automatic adjustment strategies.

Images