Garden environment monitoring data analysis method and system

Abstract

The invention provides an intelligent garden environment monitoring data analysis method and system. Sensors are deployed to collect temperature, humidity and other data in real time to form a data set. Analyzing an environment parameter change mode of each region by adopting a time sequence clustering algorithm, and determining characteristics and trends; and carrying out dimension reduction on the data by using a multi-dimensional scale transformation algorithm to a three-dimensional space for visualization, and generating an environment state map. And evaluating the plant fitness in combination with a plant physiological ecology knowledge base, and generating a scoring matrix to guide the preliminary layout scheme. And dynamically adjusting the layout by applying reinforcement learning, selecting an optimal scheme, and predicting a future environmental evolution path based on the optimal scheme and historical data. An automatic control system framework is constructed, a self-adaptive control algorithm is designed, intelligent regulation and control of plant layout are realized, garden management efficiency is improved, and ecological protection and sustainable development are promoted. Therefore, full-chain intelligent management from environment monitoring to plant layout optimization is realized, and the intelligent level of garden environment monitoring and management is remarkably improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of environmental science and ecology technology, and in particular, to a garden environment monitoring data analysis method and system. Background Art

[0002] With the acceleration of urbanization, the quality and management of garden environment have become increasingly important. Modern gardens need to be not only beautiful, but also have ecological functions to improve the quality of life and environmental health of urban residents. In order to achieve this goal, it is necessary to monitor key parameters such as temperature, humidity, light intensity and air quality index in the garden environment in real time. By deploying distributed sensor nodes at key locations in the garden environment, continuous environmental data streams can be obtained, providing a basis for subsequent analysis and decision-making. However, how to efficiently process these massive data and extract valuable information from them is a major challenge facing current garden management and optimization.

[0003] At present, most garden environment monitoring systems rely on traditional static data analysis methods, mainly focusing on monitoring environmental parameters at a single point or within a small range. Although some advanced systems have begun to introduce time series analysis to identify periodic and non-periodic change patterns, they generally lack in-depth analysis of interactions between different regions. In addition, although existing visualization tools can display the changing trends of some environmental parameters, they are still insufficient in multivariate vector dimensionality reduction processing and comprehensive display in three-dimensional space. For plant fitness assessment, existing methods are mostly based on fixed rules and empirical models, which are difficult to flexibly respond to complex and changeable actual environmental conditions.

[0004] There are multiple limitations in existing garden environment monitoring and analysis methods. First, traditional methods cannot fully capture the dynamic interactions between various areas in the garden environment, resulting in an insufficient understanding of the overall environmental characteristics. Second, the lack of effective dimensionality reduction and visualization methods makes it difficult for managers to intuitively grasp the changing trends of multivariate environmental parameters over time, affecting decision-making efficiency. Finally, the existing plant fitness assessment model is too static and fails to fully utilize the advantages of machine learning and optimization algorithms, and cannot dynamically adjust plant layout to adapt to future environmental changes. Therefore, there is an urgent need for a method that integrates multiple advanced technologies to improve the level of intelligence in garden environment monitoring and management and ensure the healthy and stable development of garden ecosystems. Summary of the invention

[0005] The embodiments of the present application provide a garden environment monitoring data analysis method and system to solve the problem that the monitoring and analysis methods in the prior art have multiple limitations.

[0006] In a first aspect, an embodiment of the present application provides a garden environment monitoring data analysis method, comprising: Optionally, using distributed sensor nodes deployed at various key locations in the garden environment, environmental parameters corresponding to the garden environment are monitored in real time to obtain an environmental data stream, wherein the environmental parameters include temperature, humidity, light intensity and air quality index; According to the environmental data stream, a time series clustering algorithm is used to identify the change pattern of environmental parameters in different areas divided by the garden environment, wherein the change pattern includes periodic change or non-periodic change, and the interaction of environmental parameters between adjacent or close areas is analyzed in combination with spatial association rule mining technology, and the environmental characteristics of the environmental parameters of each area and the change trend over time are determined according to the change pattern and the interaction; Based on the environmental characteristics of each area and its changing trend over time, a multidimensional scaling algorithm is applied to reduce the dimension of the multivariate vector corresponding to the environmental parameters of the environmental data stream at each time point, and visualize it in three-dimensional space to generate an environmental state map; Based on the environmental state map, using a random forest classifier and a Bayesian optimization algorithm, combined with a preset plant physiological ecology knowledge base, the adaptability of various plants in the garden environment to the current environmental conditions and the predicted future environmental conditions is evaluated, and a fitness scoring matrix is ​​generated; According to the preset ecological goals and the fitness score matrix, the optimal plant configuration combination is searched through the genetic algorithm and the particle swarm optimization algorithm to obtain a preliminary plant layout plan, and the plant layout plan is comprehensively evaluated according to the influence relationship between various types of plants and the environmental impact by implementing the network analysis method, and the preliminary plant layout plan is dynamically adjusted in combination with the adaptive strategy of reinforcement learning to determine the optimal plant layout plan; The layout of various types of plants in the garden environment is achieved based on the optimal plant layout plan.

[0007] Optionally, historical environmental data is obtained, and a long short-term memory network is used to construct a time series of environmental data according to the historical environmental data to predict the evolution path of environmental parameters in a future period of time to obtain a prediction result; Formulate a personalized maintenance plan using a decision tree regression model based on the prediction results and the optimal plant layout plan; Based on the personalized maintenance plan, an automatic control system framework is constructed, and an adaptive control algorithm is designed to automatically adjust the control parameters of the automatic control system framework according to the current environmental parameters obtained in real time, so as to realize intelligent regulation and control, so as to realize intelligent monitoring and management of the garden environment.

[0008] Optionally, according to the preset ecological goal and the fitness score matrix, the optimal plant configuration combination is searched by genetic algorithm and particle swarm optimization algorithm to obtain a preliminary plant layout plan, and the plant layout plan is comprehensively evaluated according to the influence relationship between various types of plants and environmental impact by implementing network analysis method, and the preliminary plant layout plan is dynamically adjusted in combination with the adaptive strategy of reinforcement learning to determine the optimal plant layout plan, including: According to the specific requirements of the garden design, a series of ecological goals are set, including biodiversity, carbon absorption, and landscape aesthetics. A genetic algorithm and a particle swarm optimization algorithm are used to perform a global search process on potential plant configuration combinations with the fitness score matrix as input to obtain a preliminary plant configuration solution set that meets the ecological goals and has a high comprehensive fitness score, and the optimal solution is selected from the global search process as the preliminary plant layout solution; Using the implementation network analysis method, based on the influence relationship between various types of plants and the impact of various types of plants on the environment, the preliminary plant layout plan is comprehensively evaluated to ensure that the plant layout meets the ecological goals and obtain an intermediate plant layout plan; Combined with the adaptive strategy of reinforcement learning, the intermediate plant layout plan is dynamically adjusted. The intermediate plant layout plan is continuously and dynamically adjusted through a reward mechanism. After multiple rounds of iterations and optimization, the optimal plant layout plan is determined.

[0009] Optionally, the intermediate plant layout scheme is dynamically adjusted in combination with an adaptive strategy of reinforcement learning, and the intermediate plant layout scheme is continuously and dynamically adjusted through a reward mechanism, and after multiple rounds of iterations and optimizations, the optimal plant layout scheme is determined, including: Constructing a reinforcement learning environment, using the state of the current plant layout and environmental conditions in the intermediate plant layout scheme as the state space, defining adjustment operations including adding, removing or replacing plant species at specific locations as the action space, and establishing an environmental feedback system based on the environmental monitoring data and plant growth data received in real time; In the reinforcement learning environment, the state space and action space are quantified according to the ecological goal to generate a reward function, and appropriate penalty items are set for operations that violate the ecological goal or cause adverse effects to generate a reward mechanism; based on the new state information provided by the environmental feedback system, the reward value is calculated for each executed action to ensure that the reward mechanism can truly reflect the actual effect of the plant layout; Using a reinforcement learning algorithm, trial-and-error learning is performed on the states and actions in the reinforcement learning environment, and the reinforcement learning model is continuously updated through the reward mechanism to gradually improve the quality of the plant layout and obtain a preliminary adjustment strategy; After multiple rounds of iterations, the comprehensive evaluation process of the preliminary adjustment strategy is repeatedly performed. When several consecutive rounds of iterations fail to significantly improve the layout quality, the adjustment is stopped and the optimal plant layout solution is output.

[0010] Optionally, based on the environmental state map, a random forest classifier and a Bayesian optimization algorithm are used, combined with a preset plant physiological ecology knowledge base, to evaluate the adaptability of various plants in the garden environment to the current environmental conditions and the predicted future environmental conditions, and generate a fitness score matrix, including: Using principal component analysis technology, a multivariate vector obtained from the environmental state map is subjected to preliminary dimensionality reduction processing to obtain a compressed environmental parameter vector, and based on the environmental parameter vector, a deep learning-driven nonlinear dimensionality reduction processing is further performed using an autoencoder to obtain an optimized environmental parameter representation; Based on the gradient boosting decision tree algorithm, the optimized environmental parameter representation and the corresponding plant growth conditions are learned and processed to identify the key environmental features that affect the growth of specific plants. After the training is completed, the feature contribution analysis value of each feature in the model is calculated using an open source feature contribution analysis library. By analyzing the feature contribution analysis values, a plant growth prediction model with enhanced generalization ability is obtained; The Bayesian optimization algorithm is combined with Gaussian process regression to perform uncertainty modeling, and the hyperparameter settings of the plant growth prediction model are optimized. After multiple rounds of iterations, a set of hyperparameter configurations that makes the plant growth prediction model perform best on the validation set is determined, thereby obtaining an optimized plant growth prediction model with higher prediction accuracy and reduced resource consumption; According to the future environmental state map generated by the climate change model, the fitness of plants under future environmental conditions is predicted using a long short-term memory network, and the generated future environmental parameter time series prediction values ​​are input into the optimized plant growth prediction model to obtain future plant fitness prediction results; Based on the shortest path algorithm in graph theory, the fitness prediction results of all plants under current and predicted future environmental conditions are integrated, the interactions and competition relationships between plants are analyzed, and a fitness scoring matrix is ​​generated to assist in formulating a reasonable layout strategy.

[0011] Optionally, according to the environmental data stream, a time series clustering algorithm is used to identify the change pattern of environmental parameters in different areas divided by the garden environment, the change pattern includes periodic change or non-periodic change, and the interaction of environmental parameters between adjacent or close areas is analyzed in combination with spatial association rule mining technology, and the environmental characteristics and change trends of environmental parameters of each area over time are determined according to the change pattern and the interaction, including: Cleaning and preprocessing the environmental data stream to remove outliers and noise to ensure the quality and consistency of the environmental data, and at the same time classifying and arranging the environmental data by time and geographical location to obtain a preprocessed environmental data stream; According to the pre-processed environmental data stream, a time series clustering algorithm is used to perform cluster analysis on the environmental parameters in different areas divided by the garden environment, identify the change pattern of the environmental parameters, and generate a typical change law in each area; Based on the typical change rules in each area, combined with the spatial association rule mining technology, the interaction of environmental parameters between adjacent or close areas is analyzed to obtain the spatial association rules between the areas; According to the spatial association rules, the environmental characteristics of the environmental parameters of each area and their changing trends over time are comprehensively evaluated.

[0012] Optionally, based on the typical change rules in each region, combined with the spatial association rule mining technology, the interaction of environmental parameters between adjacent or similar regions is analyzed to obtain the spatial association rules between the regions, including: According to the typical change rules in each area, a spatiotemporal correlation matrix is ​​constructed, wherein the spatiotemporal correlation matrix is ​​used to represent the environmental parameter values ​​and their changes in different areas at different time points; A suitable spatial association rule mining algorithm is selected to analyze the spatiotemporal association matrix, to mine the interaction of environmental parameters between adjacent or close areas, and to obtain spatial association rules.

[0013] In a second aspect, the present application embodiment provides a garden environment monitoring data analysis system, including: A monitoring module is used to use distributed sensor nodes deployed at various key locations in the garden environment to monitor the environmental parameters corresponding to the garden environment in real time to obtain an environmental data stream, wherein the environmental parameters include temperature, humidity, light intensity and air quality index; An analysis module, for identifying, based on the environmental data stream, a change pattern of environmental parameters in different areas divided by the garden environment using a time series clustering algorithm, wherein the change pattern includes periodic change or non-periodic change, and analyzing the interaction of environmental parameters between adjacent or close areas in combination with a spatial association rule mining technique, and determining the environmental characteristics of the environmental parameters of each area and their change trend over time based on the change pattern and the interaction; A dimension reduction module is used to apply a multidimensional scaling algorithm to reduce the dimension of the multivariate vector corresponding to the environmental parameters of the environmental data stream at each time point based on the environmental characteristics of each area and the change trend over time, and visualize it in three-dimensional space to generate an environmental state map; An evaluation module is used to evaluate the adaptability of various plants in the garden environment to the current environmental conditions and the predicted future environmental conditions based on the environmental state map, using a random forest classifier and a Bayesian optimization algorithm, and combining a preset plant physiological ecology knowledge base, and generate a fitness scoring matrix; A search module is used to search for the best plant configuration combination according to the preset ecological goal and the fitness score matrix through genetic algorithm and particle swarm optimization algorithm to obtain a preliminary plant layout plan, and to comprehensively evaluate the plant layout plan according to the influence relationship between various types of plants and environmental impact by implementing a network analysis method, and to dynamically adjust the preliminary plant layout plan in combination with a self-adaptive strategy of reinforcement learning to determine the optimal plant layout plan; An implementation module is used to implement the layout of various types of plants in the garden environment based on the optimal plant layout plan.

[0014] In a third aspect, an embodiment of the present application provides a computing device, comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a garden environment monitoring data analysis method as described in the first aspect above.

[0015] In a fourth aspect, an embodiment of the present application provides a computer storage medium storing a computer program, which, when executed by a computer, implements a garden environment monitoring data analysis method as described in the first aspect.

[0016] In an embodiment of the present application, distributed sensor nodes deployed at various key locations in a garden environment are used to monitor the environmental parameters corresponding to the garden environment in real time to obtain an environmental data stream, wherein the environmental parameters include temperature, humidity, light intensity, and air quality index; based on the environmental data stream, a time series clustering algorithm is used to identify the change pattern of the environmental parameters in different areas divided by the garden environment, wherein the change pattern includes periodic change or non-periodic change, and the interaction of environmental parameters between adjacent or similar areas is analyzed in combination with spatial association rule mining technology, and the environmental characteristics of the environmental parameters of each area and the change trend over time are determined based on the change pattern and the interaction; based on the environmental characteristics of each area and the change trend over time, a multidimensional scaling algorithm is applied to the multivariate vector corresponding to the environmental parameters of the environmental data stream at each time point. The data are processed by dimensionality reduction and visualized in three-dimensional space to generate an environmental state map; based on the environmental state map, the adaptability of various plants in the garden environment to the current environmental conditions and the predicted future environmental conditions is evaluated by using a random forest classifier and a Bayesian optimization algorithm, in combination with a preset plant physiological ecology knowledge base, and a fitness scoring matrix is ​​generated; according to the preset ecological goals and the fitness scoring matrix, the optimal plant configuration combination is searched by a genetic algorithm and a particle swarm optimization algorithm to obtain a preliminary plant layout plan, and the plant layout plan is comprehensively evaluated according to the influence relationship between various plants and the environmental impact by implementing a network analysis method, and the preliminary plant layout plan is dynamically adjusted in combination with an adaptive strategy of reinforcement learning to determine the optimal plant layout plan; based on the optimal plant layout plan, the layout of various plants in the garden environment is realized.

[0017] In the present application example, distributed sensor nodes deployed at key locations in the garden environment are used to monitor environmental parameters such as temperature, humidity, light intensity and air quality index in real time, ensuring the accuracy and timeliness of the data. The sensor network can cover the entire garden area, providing multi-dimensional, high-resolution environmental data streams, providing a solid foundation for subsequent analysis. The time series clustering algorithm is used to identify the change patterns of environmental parameters in different regions, including periodic changes (such as day and night temperature differences, seasonal changes) and non-periodic changes (such as sudden weather events, human activities), thereby revealing the typical change patterns in each region. Combined with spatial association rule mining technology, the interaction of environmental parameters between adjacent or similar regions is analyzed to more comprehensively understand the complex relationship between environmental parameters. This not only enhances the cognition of the environmental characteristics of each region, but also reveals the dynamic interaction between regions. Based on the environmental characteristics of each region and the changing trend over time, the multi-dimensional scaling algorithm is applied to reduce the dimensionality of the multivariate vectors of the environmental data stream at each time point, simplify the data structure and retain key information. Generate an environmental state map to intuitively display complex multivariate data in three-dimensional space to help managers better understand and make decisions. The random forest classifier is used to evaluate the fitness of plants under current environmental conditions, and the Bayesian optimization algorithm is used to predict the fitness under future environmental conditions. Combined with the preset plant physiological ecology knowledge base, a fitness scoring matrix is ​​generated to provide a scientific basis. According to the fitness scoring matrix, the adaptation of various types of plants under current and future environmental conditions is evaluated to ensure that the layout plan meets the ecological goals. The optimal plant configuration combination is searched by genetic algorithm and particle swarm optimization algorithm to obtain a preliminary plant layout plan. These algorithms can find the optimal solution in a large number of configurations and improve the quality of the layout plan. According to the influence relationship between various types of plants and the environmental impact, the preliminary plant layout plan is comprehensively evaluated to ensure the rationality and feasibility of the plan. Combined with the adaptive strategy of reinforcement learning, the preliminary plant layout plan is dynamically adjusted to cope with changes and uncertainties in practical applications and determine the optimal plant layout plan.

[0018] The technical solution of this application has the following beneficial effects: By deploying distributed sensor nodes at key locations in the garden, environmental parameters such as temperature, humidity, light intensity and air quality index are monitored in real time, ensuring high accuracy and timeliness of the data. This provides a solid foundation for subsequent analysis. The sensor network can cover the entire garden area and provide multi-dimensional, high-resolution environmental data streams, making environmental monitoring more comprehensive and detailed, helping to discover potential problems and respond in a timely manner. The time series clustering algorithm is used to identify the change patterns of environmental parameters in different regions, including periodic changes (such as day and night temperature differences, seasonal changes) and non-periodic changes (such as sudden weather events, human activities), thereby revealing the typical change patterns in each region. This enhances the understanding of dynamic environmental changes. Combined with spatial association rule mining technology, the interaction between environmental parameters between adjacent or similar regions is analyzed to more comprehensively understand the complex relationship between environmental parameters and reveal the dynamic interaction between regions. This not only improves the understanding of the environmental characteristics of each region, but also provides a basis for refined management. Based on the environmental characteristics of each region and the trend of change over time, the multidimensional scaling algorithm is applied to reduce the dimensionality of the environmental data stream, simplify the data structure and retain key information. The generated environmental state map intuitively displays complex multivariate data in three-dimensional space, helping managers to better understand and make decisions. The random forest classifier is used to evaluate the fitness of plants under current environmental conditions, and the Bayesian optimization algorithm is used to predict the fitness under future environmental conditions. Combined with the preset plant physiological ecology knowledge base, a fitness score matrix is ​​generated to provide a scientific basis for plant configuration. According to the fitness score matrix, the genetic algorithm and particle swarm optimization algorithm are used to search for the best plant configuration combination to obtain a preliminary plant layout plan. These algorithms can find the optimal solution in a large number of configurations, improve the quality of the layout plan, and ensure that the plan meets the ecological goals. According to the influence relationship between various types of plants and the environmental impact, the preliminary plant layout plan is comprehensively evaluated to ensure the rationality and feasibility of the plan. This step ensures that the final plan is not only reasonable in theory, but also feasible in practical applications. Combined with the adaptive strategy of reinforcement learning, the preliminary plant layout plan is dynamically adjusted to cope with changes and uncertainties in practical applications. This method improves the flexibility and adaptability of the plan and ensures long-term effectiveness. Through intelligent management and optimized layout, not only the efficiency of garden environment monitoring and management is improved, but also the protection and sustainable development of the ecological environment are promoted. This helps to create a healthier, more beautiful and sustainable garden environment. This method not only improves the intelligence level of garden environment monitoring and management, but also provides strong support for ecological environment protection and sustainable development.

[0019] Furthermore, based on the environmental state map, the random forest classifier and Bayesian optimization algorithm are used, combined with the preset plant physiological ecology knowledge base, to evaluate the adaptability of various plants in the garden environment to the current environmental conditions and the adaptability to the predicted future environmental conditions, and generate a fitness score matrix. The combined use of the random forest classifier and the Bayesian optimization algorithm provides a scientific evaluation method for plant fitness, which improves the accuracy and reliability of the evaluation. This method not only takes into account the current environmental conditions, but also predictively evaluates the impact of future environmental changes. Each plant has its own unique physiological and ecological needs. Combined with the plant physiological and ecological knowledge base, personalized evaluation can be performed for different types of plants, making the evaluation results more in line with the actual situation and helping to formulate a more reasonable plant configuration plan. Predicting the fitness of plants under future environmental conditions enables managers to take measures before environmental changes, such as adjusting plant layout or improving environmental conditions, thereby enhancing the resistance and sustainability of the garden. The generated fitness score matrix provides managers with intuitive data support, helping them understand which plants are most suitable for current and future environmental conditions, so as to make more informed planting decisions and optimize resource allocation. By scientifically evaluating and predicting plant fitness, we can significantly improve the efficiency and quality of garden management, promote ecological environmental protection and sustainable development, and ultimately achieve a healthy, beautiful, and functionally diverse garden environment.

[0020] Furthermore, according to the preset ecological goals and fitness score matrix, the optimal plant configuration combination is searched by genetic algorithm and particle swarm optimization algorithm to obtain the preliminary plant layout plan, and the influence relationship and environmental impact between various plants are comprehensively evaluated by implementing network analysis method. The preliminary plant layout plan is dynamically adjusted by combining the adaptive strategy of reinforcement learning to determine the optimal plant layout plan. The genetic algorithm and particle swarm optimization algorithm can find the optimal solution among a large number of possible configurations, improve the quality of the plant layout plan, and ensure that the plan is both reasonable and efficient. By implementing the network analysis method to comprehensively evaluate the interaction between plants and the environmental impact, the decision-making process is more scientific and rigorous, reducing blindness and subjectivity. The introduction of reinforcement learning enables the plan to have self-adjustment ability, can quickly respond to environmental changes or the emergence of new information, and increase the flexibility and adaptability of the layout plan. The determination of the optimal plant layout plan not only improves the beauty and functionality of the garden, but also promotes the protection and sustainable development of the ecological environment, and helps to create a healthy and stable ecosystem. Scientific and reasonable plant configuration reduces unnecessary resource waste, improves resource utilization efficiency, and reduces maintenance costs. The dynamic adjustment mechanism ensures the effectiveness and stability of the layout plan in the long term, and it can remain in good condition even in the face of environmental changes or emergencies.

[0021] These and other aspects of the present application will become more clearly understood in the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0023] Figure 1 A flow chart of a garden environment monitoring data analysis method provided in an embodiment of the present application; Figure 2 A structural diagram of a garden environment monitoring data analysis system provided in an embodiment of the present application; Figure 3 A schematic diagram of the structure of a computing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0024] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0025] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0026] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0027] Figure 1 A flow chart of a garden environment monitoring data analysis method is provided for an embodiment of the present application, such as Figure 1 As shown, the method includes: 101. Using distributed sensor nodes deployed at various key locations in the garden environment, real-time monitoring of environmental parameters corresponding to the garden environment is performed to obtain environmental data streams, wherein the environmental parameters include temperature, humidity, light intensity, and air quality index; In this embodiment, distributed sensor nodes are deployed at key locations in the garden environment to perform high-frequency real-time monitoring of environmental parameters such as temperature, humidity, light intensity, and air quality index. These sensor nodes form a network that can continuously collect environmental data streams. The environmental data stream includes not only the above basic parameters, but may also be extended to other related factors such as soil moisture and wind speed. These data are crucial for understanding the microclimate conditions in different areas of the garden.

[0028] In the embodiment of the present application, in a large urban park, the management department deploys a wireless sensor network (WSN) to achieve comprehensive monitoring of the entire garden environment. Each sensor node regularly sends the collected data to the central server to ensure the timeliness and accuracy of the data. This step provides a solid foundation for subsequent data analysis and allows managers to respond quickly to any abnormal situations.

[0029] Here is a specific example: For example, in a large public green space project in a certain city, the team installed hundreds of sensor nodes distributed in different areas of the park, such as the lakeside, wooded areas, lawns and playgrounds. These sensors record environmental data once a minute and transmit it to the cloud platform through wireless communication technology. Administrators can view this data in real time through mobile phone applications or computer terminals, thereby achieving all-weather monitoring of the garden environment.

[0030] 102. Based on the environmental data stream, a time series clustering algorithm is used to identify the change pattern of environmental parameters in different areas divided by the garden environment, wherein the change pattern includes periodic change or non-periodic change, and the interaction of environmental parameters between adjacent or close areas is analyzed in combination with spatial association rule mining technology, and the environmental characteristics and change trend of environmental parameters of each area over time are determined according to the change pattern and the interaction; In this embodiment, the environmental data stream obtained from step 101 is processed using a time series clustering algorithm to identify the change patterns of environmental parameters in each area of ​​the garden, including periodic changes (such as day and night temperature differences, seasonal changes) and non-periodic changes (such as sudden weather events, human activities). Combined with spatial association rule mining technology, the interaction of environmental parameters between adjacent or close areas is further analyzed to determine the environmental characteristics of each area and its change trend over time.

[0031] In the embodiment of the present application, based on the collected environmental data stream, a time series clustering algorithm is used to group data with similar change patterns to reveal the typical change patterns in different regions. Subsequently, the interaction between regions is analyzed through spatial association rule mining technology to help understand the complex relationship between environmental parameters and provide deeper insights.

[0032] Here is a specific example: Continuing with the previous example, the data analysis team used a time series clustering algorithm to discover some interesting patterns, such as areas near the lake cooling down faster at night, while areas with dense trees showed a strong temperature buffering effect. In addition, spatial association rule mining showed that humidity changes in certain areas were closely related to irrigation activities in neighboring areas. This information can help develop more precise management and maintenance strategies, such as adjusting irrigation plans to suit the needs of specific areas.

[0033] 103. Based on the environmental characteristics of each area and the changing trend over time, a multidimensional scaling algorithm is applied to reduce the dimension of the multivariate vector corresponding to the environmental parameters of the environmental data stream at each time point, and visualize it in three-dimensional space to generate an environmental state map; In this embodiment, the multidimensional scaling algorithm (MDS) is a dimensionality reduction method that can convert high-dimensional environmental parameter data into low-dimensional representations while maintaining the distance relationship between the original data, usually visually displayed in three-dimensional space. This method simplifies the data structure while retaining key information, and the generated environmental status map can help managers intuitively understand and analyze complex environmental data.

[0034] In the embodiment of the present application, the MDS algorithm is applied to reduce the dimension of the multivariate vector in the environmental data stream according to the environmental characteristics of each area and the trend of change over time. Then, the reduced-dimensional data is mapped into three-dimensional space to create an environmental state map. This map is not only an intuitive display tool for data, but also provides visual assistance for subsequent decision support.

[0035] Here is a specific example: Continuing with the above example, the team used the MDS algorithm to convert the multivariate environmental data collected daily into three-dimensional graphs and displayed them through a specially developed software interface. Administrators can easily browse these graphs and quickly identify which areas have abnormal conditions or need special attention. For example, when it is observed that the air quality index in a certain area is continuously high, other related parameters in the graph can be used to infer the cause and take corresponding measures, such as increasing green area or adjusting traffic flow.

[0036] 104. Based on the environmental state map, using a random forest classifier and a Bayesian optimization algorithm, combined with a preset plant physiological ecology knowledge base, evaluate the adaptability of various plants in the garden environment to the current environmental conditions and the predicted future environmental conditions, and generate a fitness score matrix; In this embodiment, the random forest classifier is a machine learning model that is suitable for processing multivariate inputs and has strong noise resistance, and is used to evaluate the fitness of various types of plants under the current environment. The Bayesian optimization algorithm is used to predict the fitness of plants under future environmental conditions, and the optimal model configuration is selected by optimizing hyperparameters to improve the prediction accuracy. Combined with the preset plant physiological ecology knowledge base, it ensures that the evaluation process fully considers the specific environmental conditions required for plant growth and its response mechanism.

[0037] In the embodiment of the present application, the random forest classifier is a machine learning model that is suitable for processing multivariate inputs and has strong noise resistance, and is used to evaluate the fitness of various types of plants in the current environment. The Bayesian optimization algorithm is used to predict the fitness of plants under future environmental conditions, and the optimal model configuration is selected by optimizing hyperparameters to improve the prediction accuracy. Combined with the preset plant physiological ecology knowledge base, it ensures that the evaluation process fully considers the specific environmental conditions required for plant growth and its response mechanism.

[0038] Here is a specific example: In the aforementioned park project, the team used a random forest classifier to assess the fitness of existing plant species under current environmental conditions, and used a Bayesian optimization algorithm to predict the impact of climate change on plant fitness in the region over the next decade. The results showed that some drought-tolerant plants perform well under future drought conditions, while other moisture-loving plants may face challenges. This information is integrated into a fitness scoring matrix to guide subsequent plant selection and layout planning, ensuring that the planting plan can adapt to environmental changes in the long term.

[0039] 105. According to the preset ecological goals and the fitness score matrix, the optimal plant configuration combination is searched by genetic algorithm and particle swarm optimization algorithm to obtain a preliminary plant layout plan, and the plant layout plan is comprehensively evaluated according to the influence relationship between various types of plants and environmental impact by implementing network analysis method, and the preliminary plant layout plan is dynamically adjusted in combination with the adaptive strategy of reinforcement learning to determine the optimal plant layout plan; In this embodiment, genetic algorithm (GA) and particle swarm optimization algorithm (PSO) are two commonly used global optimization algorithms, which are used to find the optimal solution among a large number of possible solutions. By setting specific ecological goals (such as biodiversity, landscape aesthetics, water resource utilization efficiency, etc.), combined with the fitness scoring matrix, these algorithms can efficiently search for the best plant configuration combination and obtain a preliminary plant layout plan. The implementation of the network analysis method takes into account the interaction between plants and environmental impacts to ensure that the plan is reasonable and feasible. The adaptive strategy of reinforcement learning enables the plan to be dynamically adjusted to cope with changes and uncertainties in practical applications.

[0040] In the embodiment of the present application, according to the preset ecological goals and fitness score matrix, GA and PSO are used to search for the best plant configuration combination to generate a preliminary plant layout plan. Then, the interaction between plants and environmental impact are evaluated by implementing network analysis method to ensure the rationality of the plan. Finally, combined with the adaptive strategy of reinforcement learning, the preliminary layout plan is dynamically adjusted to determine the optimal plant layout plan.

[0041] Here is a specific example: The park management team used GA and PSO algorithms to explore a variety of plant configuration combinations and finally found the best solution that meets ecological goals. By implementing the network analysis method, the team verified that the solution not only improved biodiversity, but also enhanced the landscape aesthetics and resource utilization efficiency. In order to cope with possible environmental changes in the future, the team introduced a reinforcement learning strategy so that the plant layout plan can be automatically adjusted according to actual conditions to ensure long-term effectiveness and stability. For example, when the lighting conditions in a certain area change, the system will recommend replacing plant species suitable for the new lighting conditions.

[0042] 106. Based on the optimal plant layout plan, various types of plants in the garden environment are arranged.

[0043] In this embodiment, based on the optimal plant layout plan obtained in the first five steps, specific planting and management measures are implemented to achieve the layout of various types of plants in the garden environment. This process not only takes into account the current adaptability of the plants, but also foresees the possibility of future environmental changes, ensuring the health, stability and sustainable development of the garden ecosystem.

[0044] In the embodiment of the present application, according to the determined optimal plant layout plan, the team began to implement a specific planting plan. During this process, the team closely monitored the growth of the plants and adjusted management measures such as irrigation, fertilization, and pest control as needed. Through continuous feedback and optimization, the effect of the plant layout plan is maximized.

[0045] Here is a specific example: In the final stage of the park project, the team carried out a large-scale tree planting activity according to the optimal plant layout plan. They selected plants suitable for local climate and soil conditions and arranged reasonable planting density and spacing. After that, the team regularly checked the growth of plants and adjusted maintenance measures as needed. Over time, the park gradually formed a beautiful and eco-friendly environment, attracting more tourists and wildlife, and demonstrating a successful case of smart garden management. For example, some areas in the park have reduced dependence on external water sources by planting highly adaptable local plants, achieving the goal of water conservation and environmental protection.

[0046] Through the above six steps, this scheme not only improves the intelligent level of garden environment monitoring and management, but also promotes the protection and sustainable development of the ecological environment. Specifically, real-time monitoring ensures the accuracy and timeliness of data; change pattern recognition and spatial correlation analysis reveal the complex relationship between environmental parameters; multidimensional scaling and visualization show an intuitive map of environmental status; plant fitness assessment provides a scientific basis; the search for the best plant configuration combination improves the quality of the layout plan; and the optimal plant layout finally achieved promotes the healthy development of the garden ecosystem. Overall, this methodology not only enhances the efficiency and quality of garden management, but also lays a solid foundation for creating a healthier, more beautiful and sustainable garden environment. This method is not only applicable to urban park and green space management, but can also be applied to agriculture, forestry and other fields involving ecological protection and resource management.

[0047] Optionally, historical environmental data is obtained, and a long short-term memory network is used to construct a time series of environmental data based on the historical environmental data to predict the evolution path of environmental parameters in a future period of time to obtain prediction results; a personalized maintenance plan is formulated using a decision tree regression model based on the prediction results and the optimal plant layout plan; based on the personalized maintenance plan, an automatic control system framework is constructed, and an adaptive control algorithm is designed to automatically adjust the control parameters of the automatic control system framework according to the current environmental parameters obtained in real time, so as to realize intelligent regulation and control, so as to realize intelligent monitoring and management of the garden environment.

[0048] In this embodiment, the historical environmental data includes records of key environmental parameters such as temperature, humidity, light intensity, and air quality index over a period of time in the past. These data are used to train a long short-term memory network (LSTM), a recursive neural network (RNN) specifically designed to process time series data. LSTM can capture long-term dependencies in the data and is very suitable for predicting changes in environmental parameters over a period of time in the future. Personalized maintenance plans refer to maintenance strategies customized for different regions and plant species based on the prediction results and the existing plant layout, such as irrigation frequency, fertilizer application, etc. The decision tree regression model is a machine learning method that predicts continuous value outputs by creating a series of rules, which is used here to generate specific maintenance recommendations. The automatic control system framework refers to a system that combines hardware and software, which can automatically adjust control parameters according to preset rules or models to ensure that environmental conditions are maintained within an ideal range.

[0049] In the embodiment of the present application, the historical environmental data of each sensor node in the garden environment is first collected and sorted, and the LSTM model is trained using these data to build a time series model that can accurately predict the evolution path of future environmental parameters. Based on this prediction result, and in combination with the previously determined optimal plant layout plan, a decision tree regression model is used to formulate a personalized maintenance plan for each area. Then, an automatic control system framework is constructed, which includes multiple sensors and actuators (such as irrigation systems, shading equipment, etc.), and the control parameters are continuously optimized through an adaptive control algorithm to respond to environmental changes monitored in real time. This not only improves the accuracy of maintenance work, but also achieves effective utilization of resources and reduces maintenance costs.

[0050] Here is a specific example: In a large urban park project, the team wanted to further improve its intelligent management level. To this end, they decided to introduce an LSTM model to enhance the ability to predict future environmental changes. First, the team collected detailed environmental data from the past two years, including temperature, humidity, light intensity, and air quality index. This data was used to train the LSTM model to enable it to predict the trend of environmental parameters in each area in the next month. For example, the model predicts that rainfall will decrease during a certain period of spring, resulting in a decrease in soil moisture. Based on this prediction, the team used a decision tree regression model to develop a personalized maintenance plan. For areas that are expected to become drier, the frequency of opening the irrigation system is increased; for areas with plenty of sunlight, the working mode of the shading equipment is adjusted to protect sensitive plants from excessive exposure. In addition, the team also built an automatic control system framework that integrates multiple sensors (such as soil moisture sensors, temperature sensors) and actuators (such as automatic irrigation systems, electric awnings). Through an adaptive control algorithm, this system can dynamically adjust control parameters such as irrigation volume and shading degree according to real-time monitoring data, thereby ensuring that environmental conditions are always in the most favorable state for plant growth. Ultimately, this intelligent monitoring and management system significantly improved the ecological environment quality of the park, reduced water waste, and enabled managers to respond to various environmental changes more efficiently. Visitors enjoyed a more comfortable and beautiful tour experience, and the park became an important ecological demonstration site in the city.

[0051] Optionally, in order to further improve the optimization degree and adaptability of the plant configuration combination, in some embodiments, the step 105 searches for the best plant configuration combination according to the preset ecological goal and the fitness score matrix through the genetic algorithm and the particle swarm optimization algorithm to obtain a preliminary plant layout plan, and implements the network analysis method to comprehensively evaluate the plant layout plan according to the influence relationship between various types of plants and the environmental impact, and dynamically adjusts the preliminary plant layout plan in combination with the adaptive strategy of reinforcement learning to determine the optimal plant layout plan, including: According to the specific requirements of garden design, a series of ecological goals are set, including biodiversity, carbon absorption, and landscape aesthetics. Genetic algorithms and particle swarm optimization algorithms are used to take the fitness score matrix as input to perform a global search process on potential plant configuration combinations, so as to obtain a set of preliminary plant configuration schemes that meet the ecological goals and have a high comprehensive fitness score. The optimal solution is selected from the global search process as the preliminary plant layout scheme; the implementation network analysis method is used to conduct a comprehensive evaluation of the preliminary plant layout scheme based on the influence relationship between various types of plants and the impact of various types of plants on the environment to ensure that the plant layout meets the ecological goals and obtain an intermediate plant layout scheme; combined with the adaptive strategy of reinforcement learning, the intermediate plant layout scheme is dynamically adjusted, and the intermediate plant layout scheme that is continuously dynamically adjusted through the reward mechanism is determined after multiple rounds of iteration and optimization to determine the optimal plant layout scheme.

[0052] In this embodiment, the specific requirements of the garden design set multiple ecological goals, such as biodiversity, carbon absorption, and landscape aesthetics. These goals not only take into account the richness of plant species and the health of the ecological environment, but also include visual beauty and functionality. The fitness score matrix is ​​based on the previous evaluation, which reflects the adaptability of different plants under current and future environmental conditions. Genetic algorithm (GA) and particle swarm optimization algorithm (PSO) are two intelligent optimization algorithms used to find the global optimal solution. GA simulates the natural selection process and continuously evolves individuals in the population through operations such as crossover and mutation; PSO imitates the foraging behavior of bird groups and finds the optimal position through cooperation between individuals. The implementation of network analysis is used to evaluate the interactions between plants and their impact on the environment to ensure that the plant layout meets the ecological goals. Reinforcement learning is a machine learning method that guides the system to make optimal decisions through a reward mechanism, so that the plant layout can dynamically adapt to changes.

[0053] In the embodiment of the present application, a series of ecological goals are first set, and then the GA and PSO algorithms are used to comprehensively search for potential plant configuration combinations with a fitness score matrix as input to generate a set of preliminary plant configuration solutions that meet the ecological goals and have a high comprehensive fitness score. The optimal solution is selected from the global search process as the preliminary plant layout plan. Next, the preliminary plant layout plan is comprehensively evaluated using the implementation network analysis method to ensure that it meets the ecological goals, thereby obtaining an intermediate plant layout plan. Finally, combined with the adaptive strategy of reinforcement learning, the intermediate plant layout plan is dynamically adjusted, and after multiple rounds of iteration and optimization, the optimal plant layout plan is finally determined.

[0054] Here is a specific example: In an urban park renovation project, the team hopes to create a green space that is both beautiful and environmentally friendly. To this end, they set three main ecological goals: increase biodiversity, increase carbon absorption, and enhance landscape aesthetics. To achieve these goals, the team collected a large amount of fitness score data on existing plants and integrated it into a fitness score matrix. Subsequently, the team used GA and PSO algorithms to conduct an extensive search for possible plant configuration combinations. For example, in a certain area, the team found a local tree species that can provide shade and promote soil health, which is matched with a variety of shade-tolerant ground cover plants to form a highly adaptable plant community. After the preliminary plant layout plan was formed, the team applied the implementation network analysis method to conduct an in-depth evaluation of it. This step revealed some unexpected relationships, such as some plants that are very suitable for the area individually, but when they coexist with other plants, they may cause competition or inhibit each other's growth. Therefore, the team adjusted the positions of some plants to ensure that each plant can obtain sufficient resources and support, while promoting the stability and health of the entire system. Finally, the team introduced a reinforcement learning adaptive strategy to further optimize the plant layout. The key to this stage is to establish an effective reward mechanism to encourage the system to make reasonable adjustments in the face of environmental changes. For example, when an area has experienced continuous drought, the system will automatically reduce the number of plants with high water requirements and introduce more drought-resistant varieties. In this way, the plant layout in the park not only maintains long-term stability, but also can respond quickly in the face of climate change and other external factors. After multiple rounds of iterations and optimization, the team finally determined an optimal plant layout plan that achieved all the predetermined ecological goals and provided citizens with a beautiful and sustainable public green space.

[0055] Optionally, in order to further improve the accuracy and generalization ability of plant fitness assessment, in some embodiments, the step 104 is based on the environmental state map, using a random forest classifier and a Bayesian optimization algorithm, and combining a preset plant physiological ecology knowledge base to assess the fitness of various plants in the garden environment to the current environmental conditions and the future predicted environmental conditions, and generate a fitness score matrix, including: The principal component analysis technique is used to perform preliminary dimensionality reduction processing on the multivariate vector obtained from the environmental state map to obtain a compressed environmental parameter vector. Based on the environmental parameter vector, an autoencoder is further used to perform deep learning-driven nonlinear dimensionality reduction processing to obtain an optimized environmental parameter representation; based on the gradient boosting decision tree algorithm, the optimized environmental parameter representation and the corresponding plant growth conditions are learned and processed to identify key environmental features that affect the growth of specific plants. After the training is completed, the open source feature contribution analysis library is used to calculate the feature contribution analysis value of each feature in the model. By analyzing the feature contribution analysis value, a plant growth prediction model with enhanced generalization ability is obtained; the Bayesian optimization algorithm is combined with Gaussian process regression for uncertainty modeling, and the plant growth prediction is carried out. The hyperparameter settings of the model are optimized, and after multiple rounds of iterations, a set of hyperparameter configurations that make the plant growth prediction model perform best on the validation set is determined, thereby obtaining an optimized plant growth prediction model with higher prediction accuracy and reduced resource consumption; based on the future environmental state map generated by the climate change model, the fitness of plants under future environmental conditions is predicted using a long short-term memory network, and the generated future environmental parameter time series prediction values ​​are input into the optimized plant growth prediction model to obtain future plant fitness prediction results; based on the shortest path algorithm in graph theory, the fitness prediction results of all plants under current and predicted future environmental conditions are integrated, the interactions and competition relationships between plants are analyzed, and a fitness scoring matrix is ​​generated to assist in formulating a reasonable layout strategy.

[0056] In this embodiment, the environmental data stream includes real-time monitoring data such as temperature, humidity, light intensity and air quality index from distributed sensor nodes. These data are cleaned and preprocessed to remove outliers and noise, ensuring the quality and consistency of the environmental data. The classified and sorted environmental data stream is a data set arranged by time and geographical location, which is used for subsequent time series clustering analysis. The time series clustering algorithm is a method of grouping similar time series, aiming to reveal the typical change patterns in different regions. Spatial association rule mining technology is used to analyze the interaction between environmental parameters between adjacent or close regions, and to understand the complex relationship between environmental parameters by discovering hidden spatial association rules. Comprehensively evaluating the environmental characteristics of each region and its changing trends over time refers to comprehensively evaluating the environmental conditions of each region based on the above analysis results.

[0057] In the embodiment of the present application, the environmental data stream is first cleaned and preprocessed to remove outliers and noise to ensure the quality and consistency of the data. Then, the preprocessed environmental data is classified and sorted according to time and geographical location to form an orderly data stream. Then, the environmental parameters in different regions are clustered using a time series clustering algorithm to identify the change pattern of the environmental parameters and generate typical change rules in each region. Subsequently, the spatial association rule mining technology is combined to analyze the interaction of environmental parameters between adjacent or close regions to obtain the spatial association rules between the regions. Finally, according to these spatial association rules, the environmental characteristics of the environmental parameters in each region and the change trend over time are comprehensively evaluated, thereby providing a scientific basis for garden management and decision-making.

[0058] Here is a specific example: In an intelligent management project for an urban park, the team wanted to more accurately understand and predict environmental change patterns in different areas in order to better plan plant layout and maintenance strategies. To this end, they adopted the above improvement plan. First, the team cleaned and preprocessed the environmental data streams collected from various sensor nodes. This step includes identifying and removing outliers (such as temperature readings that are significantly beyond the normal range) and using smoothing filters to reduce the impact of random noise, ensuring the quality and consistency of environmental data. For example, extreme high temperature values ​​detected on a certain day were identified as outliers and excluded, while the humidity data with large fluctuations were smoothed using the moving average method. Next, the team sorted and organized the preprocessed environmental data by time and geographic location to form an orderly data stream. This allows environmental parameters in each time period and geographic coordinates to be accurately tracked and compared. For example, the temperature and humidity records of a specific location from 8 to 9 a.m. every day can be extracted separately for analysis. Then, the team used a time series clustering algorithm to cluster the environmental parameters in different areas. Through this process, the team identified several typical patterns of environmental change, such as areas with significant day-night temperature differences, areas significantly affected by seasons, and areas with non-periodic changes caused by human activities. For example, some areas near the recreation area showed non-periodic temperature and humidity changes related to holidays, while areas near the lake showed obvious day-night temperature differences. On this basis, the team combined spatial association rule mining technology to analyze the interaction of environmental parameters between adjacent or close areas. By mining spatial association rules, the team discovered implicit connections between different areas, such as increased humidity in a certain area may affect the soil moisture content in neighboring areas, thereby affecting plant growth. In addition, the team also noticed the impact of wind direction on air quality and temperature distribution. These findings help to formulate more reasonable irrigation and ventilation plans. Finally, based on these spatial association rules, the team comprehensively evaluated the environmental characteristics of each area and its changing trends over time. For example, the team found that some areas had lower temperatures during the day but higher temperatures at night due to the shadows of buildings, which put forward special requirements for the selection of suitable plant species. Through a comprehensive evaluation of all areas, the team can more accurately understand the microclimate characteristics of each area, and adjust the plant layout and maintenance strategies accordingly to ensure the health and beauty of the garden environment. Through this method, the team not only improved their understanding of the change patterns of different areas in the garden environment, but also demonstrated how to achieve more scientific and efficient garden management and ecological protection through the combination of intelligent technology and data analysis methods. This method can not only be applied to the management of urban parks, but can also be extended to other fields involving ecological protection and resource management, such as agriculture, forestry, etc.

[0059] Optionally, in order to further improve the accuracy and comprehensiveness of the recognition of the environmental parameter change pattern, in some embodiments, step 102 uses a time series clustering algorithm to identify the change pattern of the environmental parameters in different areas divided by the garden environment according to the environmental data stream, and the change pattern includes periodic change or non-periodic change, and combines the spatial association rule mining technology to analyze the interaction of environmental parameters between adjacent or similar areas, and determines the environmental characteristics of the environmental parameters of each area and the change trend over time according to the change pattern and the interaction, including: The environmental data stream is cleaned and preprocessed to remove outliers and noise to ensure the quality and consistency of the environmental data. At the same time, the environmental data is classified and sorted according to time and geographical location to obtain a preprocessed environmental data stream; based on the preprocessed environmental data stream, a time series clustering algorithm is used to perform cluster analysis on the environmental parameters in different areas divided by the garden environment, identify the change pattern of the environmental parameters, and generate typical change laws in each area; based on the typical change laws in each area, combined with spatial association rule mining technology, the interaction between environmental parameters of adjacent or similar areas is analyzed to obtain spatial association rules between the areas; based on the spatial association rules, the environmental characteristics of the environmental parameters of each area and the change trends over time are comprehensively evaluated.

[0060] In this embodiment, the environmental data stream includes real-time monitoring data such as temperature, humidity, light intensity and air quality index from distributed sensor nodes. These data are cleaned and preprocessed to remove outliers and noise, ensuring the quality and consistency of the environmental data. The classified and sorted environmental data stream is a data set arranged by time and geographical location, which is used for subsequent time series clustering analysis. The time series clustering algorithm is a method of grouping similar time series, aiming to reveal the typical change patterns in different regions. Spatial association rule mining technology is used to analyze the interaction between environmental parameters between adjacent or close regions, and to understand the complex relationship between environmental parameters by discovering hidden spatial association rules. Comprehensively evaluating the environmental characteristics of each region and its changing trends over time refers to comprehensively evaluating the environmental conditions of each region based on the above analysis results.

[0061] In the embodiment of the present application, the environmental data stream is first cleaned and preprocessed to remove outliers and noise to ensure the quality and consistency of the data. Then, the preprocessed environmental data is classified and sorted according to time and geographical location to form an orderly data stream. Then, the environmental parameters in different regions are clustered using a time series clustering algorithm to identify the change pattern of the environmental parameters and generate typical change rules in each region. Subsequently, the spatial association rule mining technology is combined to analyze the interaction of environmental parameters between adjacent or close regions to obtain the spatial association rules between the regions. Finally, according to these spatial association rules, the environmental characteristics of the environmental parameters in each region and the change trend over time are comprehensively evaluated, thereby providing a scientific basis for garden management and decision-making.

[0062] Here is a specific example: In an intelligent management project for an urban park, the team wanted to more accurately understand and predict environmental change patterns in different areas in order to better plan plant layout and maintenance strategies. To this end, they adopted the above improvement plan. First, the team cleaned and preprocessed the environmental data streams collected from various sensor nodes. This step includes identifying and removing outliers (such as temperature readings that are significantly beyond the normal range) and using smoothing filters to reduce the impact of random noise, ensuring the quality and consistency of environmental data. For example, extreme high temperature values ​​detected on a certain day were identified as outliers and excluded, while the humidity data with large fluctuations were smoothed using the moving average method. Next, the team sorted and organized the preprocessed environmental data by time and geographic location to form an orderly data stream. This allows environmental parameters in each time period and geographic coordinates to be accurately tracked and compared. For example, the temperature and humidity records of a specific location from 8 to 9 a.m. every day can be extracted separately for analysis. Then, the team used a time series clustering algorithm to cluster the environmental parameters in different areas. Through this process, the team identified several typical patterns of environmental change, such as areas with significant day-night temperature differences, areas significantly affected by seasons, and areas with non-periodic changes caused by human activities. For example, some areas near the recreation area showed non-periodic temperature and humidity changes related to holidays, while areas near the lake showed obvious day-night temperature differences. On this basis, the team combined spatial association rule mining technology to analyze the interaction of environmental parameters between adjacent or close areas. By mining spatial association rules, the team discovered implicit connections between different areas, such as increased humidity in a certain area may affect the soil moisture content in neighboring areas, thereby affecting plant growth. In addition, the team also noticed the impact of wind direction on air quality and temperature distribution. These findings help to formulate more reasonable irrigation and ventilation plans. Finally, based on these spatial association rules, the team comprehensively evaluated the environmental characteristics of each area and its changing trends over time. For example, the team found that some areas had lower temperatures during the day but higher temperatures at night due to the shadows of buildings, which put forward special requirements for the selection of suitable plant species. Through a comprehensive evaluation of all areas, the team can more accurately understand the microclimate characteristics of each area, and adjust the plant layout and maintenance strategies accordingly to ensure the health and beauty of the garden environment.

[0063] Through this method, the team not only improved their understanding of the change patterns of different areas in the garden environment, but also demonstrated how to achieve more scientific and efficient garden management and ecological protection through the combination of intelligent technology and data analysis methods. This method can not only be applied to the management of urban parks, but can also be extended to other fields involving ecological protection and resource management, such as agriculture, forestry, etc.

[0064] This application takes into account that in garden management and ecological planning, it is crucial to accurately assess the fitness of plants to current and future environmental conditions. Traditional methods often rely on expert experience and limited historical data, and it is difficult to fully consider complex and changeable environmental factors and their impact on plant growth. With the development of intelligent sensor networks and big data analysis technology, we are now able to obtain a large amount of high-precision environmental monitoring data in real time, and combine it with advanced machine learning algorithms for in-depth analysis. However, how to effectively integrate these data and generate reliable plant fitness assessments remains a challenge. To meet this challenge, we propose a set of new formulas based on random forest classifiers and Bayesian optimization algorithms, aiming to more accurately assess the fitness of plants to current and future environmental conditions. This formula not only takes into account the changes in environmental parameters, but also introduces a preset plant physiological ecology knowledge base to ensure that the evaluation results are more scientific and reasonable. In addition, by introducing weight coefficients and exponential factors, the model can flexibly adjust the importance of different environmental variables, thereby providing a more detailed and accurate fitness assessment. Therefore, a new optional scheme is proposed, which includes: Optionally, step 104 uses a random forest classifier and a Bayesian optimization algorithm based on the environmental state map, and combines a preset plant physiological ecology knowledge base to evaluate the adaptability of various plants in the garden environment to the current environmental conditions and the predicted future environmental conditions, and generates a fitness score matrix, including: A random forest classifier is used based on the environment state vector in the environment state map. , evaluate the adaptability of various plants in the garden environment to the current environmental conditions; wherein the adaptability of various plants in the garden environment to the current environmental conditions is calculated by the following formula: ; in, Indicates at a point in time Seasonal Plants The current fitness score of the plant The likelihood of growth and survival under current environmental conditions. Represents the number of decision trees in the random forest. Random forest is an ensemble learning method that improves the accuracy and robustness of the model by making joint decisions with multiple decision trees. Represents the index of the decision tree, from 1 to . Represents the parameter that controls the exponential decay speed, which determines the distance function The degree of influence on the final fitness score. Larger A higher value will cause the exponential decay to be faster, thus reducing the impact of unsuitable situations. Indicates plants The characteristic vector contains various physiological and ecological parameters of the plant, such as the optimal temperature range, water requirement, photosynthesis efficiency, etc. These parameters describe the growth requirements and characteristics of the plant. It represents the ideal fitness mean, which is used to set a benchmark value to measure the gap between the actual fitness and the ideal fitness, and it represents the ideal growth conditions. Indicates at a point in time The environmental state vector, including the temperature ,humidity , light intensity ,Air Quality Index Environmental parameters, which describe the current environmental conditions. Indicates The decision tree for the environment state vector and plant feature vector The classification result is 0 (unsuitable) or 1 (suitable), indicating that the plant Is it suitable for the current environmental conditions? Indicates The distance function between the classification results of a decision tree and the ideal fitness measures the gap between the actual fitness and the ideal fitness. The smaller the distance, the closer the plant is to the ideal growth condition.

[0065] The following is a brief introduction to the design reasons of each sub-item of the formula: Fractional part : This is to average the results of all decision trees. Because random forest uses multiple decision trees for prediction, the prediction results of each tree may be different. Taking the average value can reduce the deviation caused by a single decision tree and improve the stability and accuracy of the overall prediction.

[0066] Sum part :It is a summary of the results of all decision trees. By summing them, we can comprehensively consider the evaluation of plant fitness by each decision tree to ensure that the evaluation results are more comprehensive and reliable. This part is to synthesize the results of all decision trees. Each decision tree has independently evaluated the fitness of the plant. By summing them, we can summarize these evaluation results to ensure the comprehensiveness and reliability of the evaluation process.

[0067] Index part : The exponential function is used to convert the distance function Converts to a value between 0 and 1. Exponential decay factor Controls the effect of distance on fitness score. When it is smaller, the index value is larger, indicating that the plant is close to the ideal growth condition; on the contrary, when When it is larger, the index value is smaller, indicating that the plant is far away from the ideal growth conditions.

[0068] Distance function : The distance function is used to quantify the gap between actual fitness and ideal fitness. It helps us understand how well the plant matches the ideal growth conditions under the current environmental conditions, thereby more accurately evaluating the plant's fitness.

[0069] Index part The multiplication operation in is to adjust the contribution weight of each decision tree. The exponential function converts the distance into a fitness score, so that the greater the distance, the lower the score; the smaller the distance, the higher the score. This conversion method can better reflect the actual changes in plant fitness and control the exponential decay speed. , the sensitivity of the evaluation can be flexibly adjusted.

[0070] Through the Bayesian optimization algorithm, based on the preset plant physiological ecology knowledge base , evaluate the adaptability of various plants in the garden environment to the predicted future environmental conditions; wherein the adaptability of various plants in the garden environment to the predicted future environmental conditions is calculated by the following formula: ; in, Indicates at a point in time Seasonal Plants The future fitness of the plant The likelihood of growth and survival under predicted future environmental conditions. It means to find the conditional probability The largest Value, here Represent the fitness class or classification of plants, i.e. choose the most likely fitness level. Represents the state vector of the given future environment and Knowledge Base Under these conditions, plants Fitness The probability distribution of , which is calculated by a Bayesian optimization algorithm, takes into account future environmental changes and plant physiological ecology information. Indicates at a point in time The predicted future environmental state vector contains the temperature ,humidity , light intensity and air quality index and other environmental parameters, which describe the predicted future environmental conditions. Represents the preset plant physiological ecology knowledge base, which contains information on the adaptability of plants to various environmental parameters. This knowledge base helps the model understand the performance of plants under different environmental conditions. Indicates the weight of the impact of different environmental parameters on plant fitness, and introduces the index ,It is used to adjust the degree of influence of each environmental factor on plant fitness.,Indices can be obtained through historical data training or set by domain experts, and they can change over time and plant species; The following is a brief introduction to the design reasons of each sub-item of the formula: Maximum Likelihood Estimation : This part uses the Bayesian optimization algorithm to estimate the probability distribution of plant fitness under future environmental conditions and select the most likely fitness level. , which ensures that we have selected the fitness assessment results that are most suitable for future environmental conditions.

[0071] Probability distribution : This probability distribution combines the predicted future environment state vector and the preset plant physiological ecology knowledge base , to provide a fitness estimate based on statistics and domain knowledge. It helps us quantify the likelihood of a plant surviving under specific future environmental conditions.

[0072] Environmental parameters influence weight :Different environmental parameters have different effects on plant fitness. By introducing weights , we can flexibly adjust the importance of each environmental parameter to make the evaluation more realistic. For example, some plants may be very sensitive to temperature, while others are more dependent on light intensity.

[0073] Index Adjustment Factor :The multiplication operation combines the effects of all environmental parameters to form an overall fitness evaluation. The influence of each environmental parameter is expressed by its weight and adjustment index Weighting is performed to ensure that the final fitness score fully reflects the effects of all environmental factors.

[0074] Multiplication part The purpose of this multiplication is to consider the impact of multiple environmental parameters. By multiplying, we can combine the weights and adjustment indexes of each environmental parameter to form a comprehensive fitness score. This multiplication method not only considers the independent impact of each environmental parameter, but also ensures that the interactions between them are correctly reflected.

[0075] The above formula is designed to comprehensively consider the impact of multiple environmental parameters on future plant fitness, and combine these impacts through the Bayesian optimization algorithm and the multiplication operation. The maximum likelihood estimation part ensures that we select the most likely fitness evaluation result, while the probability distribution part provides an evaluation basis based on statistics and domain knowledge. The multiplication part comprehensively considers the impact of various environmental parameters through multiplication operations to ensure that the evaluation results are more comprehensive and accurate. This method not only improves the accuracy of future plant fitness assessments, but also provides a scientific basis for garden management and ecological protection.

[0076] According to the adaptability of various plants in the garden environment to the current environmental conditions and the adaptability of various plants in the garden environment to the predicted future environmental conditions, a fitness scoring matrix is ​​constructed; wherein the fitness scoring matrix is ​​defined as: ; in, Indicates at a point in time Seasonal Plants The comprehensive fitness score of Represents the weight coefficient under the current environmental conditions, which is used to adjust the influence of the current fitness score. Indicates at a point in time Seasonal Plants The current fitness score of is calculated by using the random forest classifier and the Bayesian optimization algorithm. The exponential factor representing the fitness score under the current environmental conditions is used to adjust the degree of influence of the fitness score. Indicates the effective time point of the current environmental conditions. When , the current fitness score is used. represents the weight coefficient under future environmental conditions, which is used to adjust the impact of future fitness scores. Indicates at a point in time Seasonal Plants The future fitness score of is calculated by using the random forest classifier and the Bayesian optimization algorithm. An exponential factor representing the fitness score under future environmental conditions, used to adjust the impact of the fitness score. Indicates the effective time point of future environmental conditions. , the future fitness score is used.

[0077] Indicates at a point in time Seasonal Plants The comprehensive fitness score of the plant. Overall growth and survival probability at a specific point in time. and Represent the weight coefficients of current fitness and future fitness respectively. These weights can reflect the plant The overall growth and survival probability at a specific point in time. Set realistically to reflect the importance of current and future environmental conditions. Indicates at a point in time Seasonal Plants The current fitness score of a plant. It assesses the fitness of a plant based on the current environmental conditions. Indicates at a point in time Seasonal Plants Future fitness score. It assesses the fitness of plants based on predicted future environmental conditions. and : They are adjustment indices for current fitness and future fitness, respectively. They can be used to amplify or reduce the impact of the fitness score, allowing the model to respond more flexibly to changes under different conditions. and : The current and future threshold time points, respectively. They define when to use the current fitness score and when to switch to the future fitness score.

[0078] The following is a brief introduction to the design reasons of each sub-item of the formula: Current fitness section : This part applies to time points Does not exceed current threshold It uses the fitness score under the current environmental conditions And pass it through the weight and adjustment index Adjustments are made to produce a composite fitness score. This design allows us to assess plants in real time based on current environmental conditions.

[0079] Future fitness part : This part applies to time points Exceeding the current threshold but not exceeding future thresholds It uses the fitness score under predicted future environmental conditions , and also through the weight and adjustment index This design allows us to consider possible future environmental changes in advance and make corresponding management decisions accordingly.

[0080] Weight coefficient and :The weight coefficient is introduced to balance the importance of current fitness and future fitness. For example, in short-term planning, current fitness may be more important, while in long-term planning, future fitness is more critical. The weight coefficient can be flexibly adjusted according to actual needs.

[0081] Adjustment Index and : The tuning exponent is used to control the degree of influence of the fitness score. A larger exponent makes the score more sensitive to changes, while a smaller exponent makes it relatively stable. This helps to adjust the sensitivity of the model for different plant species and environmental conditions.

[0082] The multiplication operation in the formula is used to combine the weight coefficient with the fitness score, and further modify the final score by adjusting the exponent. This approach not only reflects the independent impact of each factor, but also ensures that the interaction between them is correctly reflected. For example, by multiplying, we can amplify or reduce the importance of a certain fitness score to better adapt to different application scenarios. Through the above decomposition and explanation, we can see that the formula The design aims to provide a dynamic, time-varying method for evaluating plant fitness. It not only considers the immediate fitness under current environmental conditions, but also foresightedly considers the impact of future environmental changes on plant fitness. By introducing weight coefficients and adjustment indexes, the model can flexibly adapt to different plant species and environmental conditions, providing strong support for garden management and ecological protection. This method not only improves the accuracy of fitness assessment, but also enhances the ability to respond to future environmental changes.

[0083] Through these adjustments, not only the complexity of the model is increased, but also the impact of environmental conditions on plant fitness can be more accurately reflected. The introduction of weights and indexes allows the model to consider the importance of different environmental variables and their different degrees of impact on plant growth, thereby providing a more detailed and accurate fitness assessment method.

[0084] Here is a specific example: suppose in a city park, you need to evaluate the adaptability of a certain type of plant (such as ginkgo trees) under different environmental conditions and predict the impact of future environmental changes on its growth. The specific values ​​are as follows: Current environment state vector :temperature ,humidity , light intensity ,Air Quality Index . Pre-set plant physiological ecology knowledge base The optimum temperature range for Ginkgo trees is , water requirement is , photosynthesis efficiency is high. Weight and index: temperature ,humidity , light intensity ,Air Quality Index ;index 1.5, . Random Forest Classifier Parameters: . Bayesian optimization algorithm parameters: .

[0085] Specific calculation process: Current fitness evaluation: , assuming that the average fitness obtained by the random forest classifier is . Future fitness prediction: , assuming that the future fitness predicted by the Bayesian optimization algorithm is . Generate a total fitness score drop of: , calculated: The total reduction of the current fitness score is , the total fitness score decreases in the future .

[0086] Figure 2 A structural diagram of a garden environment monitoring data analysis system is provided for an embodiment of the present application, such as Figure 2 As shown, the system includes: 21. A monitoring module, which is used to use distributed sensor nodes deployed at various key locations in the garden environment to monitor the environmental parameters corresponding to the garden environment in real time to obtain environmental data streams, wherein the environmental parameters include temperature, humidity, light intensity and air quality index; 22. An analysis module, for identifying, based on the environmental data stream, a change pattern of environmental parameters in different areas divided by the garden environment using a time series clustering algorithm, wherein the change pattern includes periodic change or non-periodic change, and combining spatial association rule mining technology to analyze the interaction of environmental parameters between adjacent or close areas, and determining the environmental characteristics of the environmental parameters of each area and their change trend over time based on the change pattern and the interaction; 23. A dimension reduction module, for applying a multidimensional scaling algorithm to reduce the dimension of the multivariate vector corresponding to the environmental parameters of the environmental data stream at each time point based on the environmental characteristics of each area and the change trend over time, and visually displaying it in three-dimensional space to generate an environmental state map; 24. An evaluation module, for evaluating the adaptability of various plants in the garden environment to the current environmental conditions and the predicted future environmental conditions based on the environmental state map, using a random forest classifier and a Bayesian optimization algorithm, and combining a preset plant physiological ecology knowledge base, and generating a fitness scoring matrix; 25. A search module, which is used to search for the best plant configuration combination according to the preset ecological goal and the fitness score matrix through genetic algorithm and particle swarm optimization algorithm to obtain a preliminary plant layout plan, and to comprehensively evaluate the plant layout plan according to the influence relationship between various types of plants and environmental impact by implementing network analysis method, and to dynamically adjust the preliminary plant layout plan in combination with the adaptive strategy of reinforcement learning to determine the optimal plant layout plan; 26. An implementation module, used to implement the layout of various types of plants in the garden environment based on the optimal plant layout plan.

[0087] Figure 2 The garden environment monitoring data analysis system can be executed Figure 1 The implementation principle and technical effect of the garden environment monitoring data analysis method described in the embodiment are not repeated here. The specific way in which each module and unit performs operations in the garden environment monitoring data analysis system in the above embodiment has been described in detail in the embodiment of the method, and will not be described in detail here.

[0088] In one design, Figure 2 The garden environment monitoring data analysis system of the embodiment shown can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32; The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0089] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 The garden environment monitoring data analysis method of the illustrated embodiment.

[0090] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0091] The system embodiment described above is merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.

[0092] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0093] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A garden environment monitoring data analysis method, characterized in that: include: Using distributed sensor nodes deployed at various key locations in the garden environment, real-time monitoring of environmental parameters corresponding to the garden environment is performed to obtain environmental data streams, wherein the environmental parameters include temperature, humidity, light intensity, and air quality index; According to the environmental data stream, a time series clustering algorithm is used to identify the change pattern of environmental parameters in different areas divided by the garden environment, wherein the change pattern includes periodic change or non-periodic change, and the interaction of environmental parameters between adjacent or close areas is analyzed in combination with spatial association rule mining technology, and the environmental characteristics of the environmental parameters of each area and the change trend over time are determined according to the change pattern and the interaction; Based on the environmental characteristics of each area and its changing trend over time, a multidimensional scaling algorithm is applied to reduce the dimension of the multivariate vector corresponding to the environmental parameters of the environmental data stream at each time point, and visualize it in three-dimensional space to generate an environmental state map; Based on the environmental state map, using a random forest classifier and a Bayesian optimization algorithm, combined with a preset plant physiological ecology knowledge base, the adaptability of various plants in the garden environment to the current environmental conditions and the predicted future environmental conditions is evaluated, and a fitness scoring matrix is ​​generated; According to the preset ecological goals and the fitness score matrix, the optimal plant configuration combination is searched through the genetic algorithm and the particle swarm optimization algorithm to obtain a preliminary plant layout plan, and the plant layout plan is comprehensively evaluated according to the influence relationship between various types of plants and the environmental impact by implementing the network analysis method, and the preliminary plant layout plan is dynamically adjusted in combination with the adaptive strategy of reinforcement learning to determine the optimal plant layout plan; The layout of various types of plants in the garden environment is achieved based on the optimal plant layout plan.

2. The method according to claim 1, characterized in that Also includes: Acquire historical environmental data, and use a long short-term memory network to construct a time series of environmental data based on the historical environmental data to predict the evolution path of environmental parameters in the future and obtain a prediction result; Formulate a personalized maintenance plan using a decision tree regression model based on the prediction results and the optimal plant layout plan; Based on the personalized maintenance plan, an automatic control system framework is constructed, and an adaptive control algorithm is designed to automatically adjust the control parameters of the automatic control system framework according to the current environmental parameters obtained in real time, so as to realize intelligent regulation and control, so as to realize intelligent monitoring and management of the garden environment.

3. The method according to claim 1, characterized in that According to the preset ecological goals and the fitness score matrix, the optimal plant configuration combination is searched through genetic algorithm and particle swarm optimization algorithm to obtain a preliminary plant layout plan, and the plant layout plan is comprehensively evaluated according to the influence relationship between various types of plants and environmental impact by implementing network analysis method, and the preliminary plant layout plan is dynamically adjusted in combination with the adaptive strategy of reinforcement learning to determine the optimal plant layout plan, including: According to the specific requirements of the garden design, a series of ecological goals are set, including biodiversity, carbon absorption, and landscape aesthetics. A genetic algorithm and a particle swarm optimization algorithm are used to perform a global search process on potential plant configuration combinations with the fitness score matrix as input to obtain a preliminary plant configuration solution set that meets the ecological goals and has a high comprehensive fitness score, and the optimal solution is selected from the global search process as the preliminary plant layout solution; Using the implementation network analysis method, based on the influence relationship between various types of plants and the impact of various types of plants on the environment, the preliminary plant layout plan is comprehensively evaluated to ensure that the plant layout meets the ecological goals and obtain an intermediate plant layout plan; Combined with the adaptive strategy of reinforcement learning, the intermediate plant layout plan is dynamically adjusted. The intermediate plant layout plan is continuously and dynamically adjusted through a reward mechanism. After multiple rounds of iterations and optimization, the optimal plant layout plan is determined.

4. The method according to claim 3, characterized in that Combined with the adaptive strategy of reinforcement learning, the intermediate plant layout scheme is dynamically adjusted. The intermediate plant layout scheme is continuously and dynamically adjusted through the reward mechanism. After multiple rounds of iterations and optimization, the optimal plant layout scheme is determined, including: Constructing a reinforcement learning environment, using the state of the current plant layout and environmental conditions in the intermediate plant layout scheme as the state space, defining adjustment operations including adding, removing or replacing plant species at specific locations as the action space, and establishing an environmental feedback system based on the environmental monitoring data and plant growth data received in real time; In the reinforcement learning environment, the state space and action space are quantified according to the ecological goal to generate a reward function, and appropriate penalty items are set for operations that violate the ecological goal or cause adverse effects to generate a reward mechanism; based on the new state information provided by the environmental feedback system, the reward value is calculated for each executed action to ensure that the reward mechanism can truly reflect the actual effect of the plant layout; Using a reinforcement learning algorithm, trial-and-error learning is performed on the states and actions in the reinforcement learning environment, and the reinforcement learning model is continuously updated through the reward mechanism to gradually improve the quality of the plant layout and obtain a preliminary adjustment strategy; After multiple rounds of iterations, the comprehensive evaluation process of the preliminary adjustment strategy is repeatedly performed. When several consecutive rounds of iterations fail to significantly improve the layout quality, the adjustment is stopped and the optimal plant layout solution is output.

5. The method according to claim 1, characterized in that Based on the environmental state map, the random forest classifier and Bayesian optimization algorithm are used, combined with the preset plant physiological ecology knowledge base, to evaluate the adaptability of various plants in the garden environment to the current environmental conditions and the predicted future environmental conditions, and generate a fitness score matrix, including: Using principal component analysis technology, a multivariate vector obtained from the environmental state map is subjected to preliminary dimensionality reduction processing to obtain a compressed environmental parameter vector, and based on the environmental parameter vector, a deep learning-driven nonlinear dimensionality reduction processing is further performed using an autoencoder to obtain an optimized environmental parameter representation; Based on the gradient boosting decision tree algorithm, the optimized environmental parameter representation and the corresponding plant growth conditions are learned and processed to identify the key environmental features that affect the growth of specific plants. After the training is completed, the feature contribution analysis value of each feature in the model is calculated using an open source feature contribution analysis library. By analyzing the feature contribution analysis values, a plant growth prediction model with enhanced generalization ability is obtained; The Bayesian optimization algorithm is combined with Gaussian process regression to perform uncertainty modeling, and the hyperparameter settings of the plant growth prediction model are optimized. After multiple rounds of iterations, a set of hyperparameter configurations that makes the plant growth prediction model perform best on the validation set is determined, thereby obtaining an optimized plant growth prediction model with higher prediction accuracy and reduced resource consumption; According to the future environmental state map generated by the climate change model, the fitness of plants under future environmental conditions is predicted using a long short-term memory network, and the generated future environmental parameter time series prediction values ​​are input into the optimized plant growth prediction model to obtain future plant fitness prediction results; Based on the shortest path algorithm in graph theory, the fitness prediction results of all plants under current and predicted future environmental conditions are integrated, the interactions and competition relationships between plants are analyzed, and a fitness scoring matrix is ​​generated to assist in formulating a reasonable layout strategy.

6. The method according to claim 1, characterized in that According to the environmental data stream, a time series clustering algorithm is used to identify the change pattern of environmental parameters in different areas divided by the garden environment, wherein the change pattern includes periodic change or non-periodic change, and the interaction of environmental parameters between adjacent or close areas is analyzed in combination with spatial association rule mining technology. According to the change pattern and the interaction, the environmental characteristics of the environmental parameters of each area and the change trend over time are determined, including: Cleaning and preprocessing the environmental data stream to remove outliers and noise to ensure the quality and consistency of the environmental data, and at the same time classifying and arranging the environmental data by time and geographical location to obtain a preprocessed environmental data stream; According to the pre-processed environmental data stream, a time series clustering algorithm is used to perform cluster analysis on the environmental parameters in different areas divided by the garden environment, identify the change pattern of the environmental parameters, and generate a typical change law in each area; Based on the typical change rules in each area, combined with the spatial association rule mining technology, the interaction of environmental parameters between adjacent or close areas is analyzed to obtain the spatial association rules between the areas; According to the spatial association rules, the environmental characteristics of the environmental parameters of each area and their changing trends over time are comprehensively evaluated.

7. The method according to claim 6, characterized in that Based on the typical change rules in each area, combined with the spatial association rule mining technology, the interaction of environmental parameters between adjacent or close areas is analyzed to obtain the spatial association rules between the areas, including: According to the typical change rules in each area, a spatiotemporal correlation matrix is ​​constructed, wherein the spatiotemporal correlation matrix is ​​used to represent the environmental parameter values ​​and their changes in different areas at different time points; A suitable spatial association rule mining algorithm is selected to analyze the spatiotemporal association matrix, to mine the interaction of environmental parameters between adjacent or close areas, and to obtain spatial association rules.

8. A garden environment monitoring data analysis system, characterized in that: include: A monitoring module is used to use distributed sensor nodes deployed at various key locations in the garden environment to monitor the environmental parameters corresponding to the garden environment in real time to obtain an environmental data stream, wherein the environmental parameters include temperature, humidity, light intensity and air quality index; An analysis module, for identifying, based on the environmental data stream, a change pattern of environmental parameters in different areas divided by the garden environment using a time series clustering algorithm, wherein the change pattern includes periodic change or non-periodic change, and analyzing the interaction of environmental parameters between adjacent or close areas in combination with a spatial association rule mining technique, and determining the environmental characteristics of the environmental parameters of each area and their change trend over time based on the change pattern and the interaction; A dimension reduction module is used to apply a multidimensional scaling algorithm to reduce the dimension of the multivariate vector corresponding to the environmental parameters of the environmental data stream at each time point based on the environmental characteristics of each area and the change trend over time, and visualize it in three-dimensional space to generate an environmental state map; An evaluation module is used to evaluate the adaptability of various plants in the garden environment to the current environmental conditions and the predicted future environmental conditions based on the environmental state map, using a random forest classifier and a Bayesian optimization algorithm, and combining a preset plant physiological ecology knowledge base, and generate a fitness scoring matrix; A search module is used to search for the best plant configuration combination according to the preset ecological goal and the fitness score matrix through genetic algorithm and particle swarm optimization algorithm to obtain a preliminary plant layout plan, and to comprehensively evaluate the plant layout plan according to the influence relationship between various types of plants and environmental impact by implementing a network analysis method, and to dynamically adjust the preliminary plant layout plan in combination with a self-adaptive strategy of reinforcement learning to determine the optimal plant layout plan; An implementation module is used to implement the layout of various types of plants in the garden environment based on the optimal plant layout plan.

9. A computing device, characterized in that It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a garden environment monitoring data analysis method as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a computer, a garden environment monitoring data analysis method as described in any one of claims 1 to 7 is implemented.

Images