Big data-based dust retention plant combination design method and system

Abstract

The invention relates to a big data-based dust retention plant combination design method and system, and relates to the technical field of big data processing and plant ecology, and the method comprises the following steps: S1, collecting real-time environment data, including air quality data, meteorological data, soil data and traffic flow data, of a target area through a distributed sensor network, the air quality data at least comprise PM2.5 concentration, PM10 concentration and dust fall amount. Multi-source environment data are collected in real time through a distributed sensor network, a matching relation between environment characteristics and plant dust retention capacity is analyzed by combining a dust retention plant database and utilizing a random forest algorithm, the efficient dust retention effect of a plant combination is ensured, the plant combination is calculated and optimized through the ecological niche overlapping degree, resource competition is reduced, and the dust retention effect of the plant combination is improved. And a scheme for balancing the dust retention efficiency and the cost is generated through a multi-objective optimization algorithm in combination with landscape planning requirements and cost constraints.

Description

Technical Field

[0001] The present application relates to the field of material processing technology, and in particular to a dust-retaining plant combination design method and system based on big data. Background Art

[0002] With the acceleration of urbanization, improving air quality has become a key link in urban ecological construction. As natural air purifiers, plants' dust-retention capacity has important application value in urban greening, ecological restoration, and landscape design. Based on the development of big data technology, combining massive environmental data with plant ecological characteristics to achieve precise screening and combination design of dust-retaining plants has become an emerging technical direction for improving urban air quality. By integrating multi-source environmental data, analyzing plant dust-retention capacity parameters, and optimizing combination schemes, the dust-retention efficiency of plants can be significantly improved, providing a scientific basis for urban ecological planning. Currently, research in this field focuses on how to use big data to explore the matching relationship between environmental characteristics and plant dust-retention capacity, and how to achieve dynamic optimization of plant combinations through intelligent algorithms.

[0003] However, existing methods for designing dust-retaining plant combinations still have significant limitations. Traditional designs rely primarily on manual experience and on-site inspections, lacking systematic analysis of large-scale environmental data such as air quality, meteorological, and soil data. This makes it difficult to accurately identify the matching requirements between the environmental characteristics of different regions and the dust-retaining capacity of plants. At the same time, existing plant combinations often adopt fixed patterns, based solely on basic plant properties such as cold tolerance and photophilia. They ignore the impact of niche differences such as root distribution, canopy structure, and growth cycle on the synergistic dust-retaining effect of plants. This can lead to resource competition conflicts among plants within the combination, and the inability to fully utilize the dust-retaining efficiency of the group. In addition, traditional methods lack a dynamic adjustment mechanism, making it difficult to adapt to fluctuations in environmental conditions caused by factors such as changes in traffic flow and seasonal changes. Furthermore, the design process does not adequately consider the balance of multiple objectives, such as landscape aesthetics and cost constraints, resulting in limited overall benefits of the design scheme. To address this, we propose a dust-retaining plant combination design method and system based on big data. Summary of the Invention

[0004] The purpose of this application is to provide a dust-retaining plant combination design method and system based on big data.

[0005] In the first aspect, the present application provides a method for designing a dust-retaining plant combination based on big data, which adopts the following technical solutions: A dust-retaining plant combination design method based on big data includes the following steps: S1. Collecting real-time environmental data of the target area through a distributed sensor network, including air quality data, meteorological data, soil data, and traffic flow data. The air quality data includes at least PM2.5 concentration, PM10 concentration, and dust fall. The meteorological data includes wind speed, wind direction, precipitation, temperature and humidity. The soil data includes pH value, organic matter content, and heavy metal content. S2. Construct a dust-retention plant database to store plant dust-retention capacity parameters, including the dust retention capacity of a single plant, leaf morphological characteristics, dust-retention adsorption efficiency, and growth adaptability indicators. The growth adaptability indicators are obtained through a combination of wind tunnel experiments and field monitoring. S3. Based on environmental data and a dust-retaining plant database, a random forest algorithm is used to analyze the matching relationship between the environmental characteristics of the target area and the dust-retaining capacity of plants, and to generate an initial plant combination scheme. The input features of the random forest algorithm include PM2.5 concentration gradient, soil heavy metal migration rate, and the spatiotemporal distribution of traffic dust. S4. Optimize the initial combination scheme by calculating plant niche overlap, including root complementarity, canopy shading complementarity, and growth cycle overlap, and select plant combinations with a comprehensive competition index below a preset threshold; S5. Combining landscape planning requirements and cost constraints, a multi-objective optimization algorithm is used to generate a final solution that balances dust retention efficiency and cost, and the dynamic adjustment results are output through a three-dimensional visualization interface.

[0006] Preferably, the step S1 includes: Traffic flow data is used to extract vehicle density and road dust contribution rate through image recognition technology, and then integrated with meteorological data to generate pollution source heat maps; Data preprocessing also includes noise suppression and missing value filling, and Kriging interpolation method is used to generate a spatially continuous environmental dataset.

[0007] Preferably, the calculation of the niche overlap in step S4 includes: The soil resource competition index was calculated based on the difference in root distribution depth, the light competition index was calculated based on the canopy projected area, and the intensity of temporal resource conflict was quantified by combining the number of days of overlapping growth cycles. If any index exceeds the threshold, the conflicting plants are removed and the combination is re-matched.

[0008] Preferably, the multi-objective optimization algorithm in step S5 is NSGA-II, and the optimization objectives include the dust retention efficiency improvement rate, seedling procurement cost, maintenance cost and landscape aesthetic score. The Pareto optimal solution set is output for user interactive selection, and the generation of similar area solutions is accelerated through the transfer learning model.

[0009] In the second aspect, the present application provides a dust-retaining plant combination design system based on big data, which adopts the following technical solutions: A dust-retaining plant combination design system based on big data, comprising: The data acquisition module integrates temperature sensors, humidity sensors, gas concentration sensors, and traffic flow monitoring units to collect PM2.5 concentration, humidity, dust fall, and vehicle dust data in the target area; The data interaction module is connected to the data acquisition module to filter outliers, align timestamps, and perform spatial interpolation on the data, generate gridded datasets, and synchronize them with the cloud server through the API interface; The environmental conditioning module includes an air supply system, a spray system, an air production system, and a dehumidification and ventilation system. The air supply system uses a variable frequency fan to adjust the air supply strategy to suppress dust diffusion. The spray system dynamically adjusts the atomization particle size based on humidity feedback. The air production system injects negative oxygen ions to enhance the dust retention capacity of plants. The dehumidification and ventilation system uses a multi-stage filtration unit to purify polluted air. The driver module includes parallel linear MOSFET circuits to provide stable power to each module and switch to the backup circuit when the voltage fluctuates; The system dynamically optimizes plant combination plans and environmental adjustment parameters through machine learning models, and records operation logs based on blockchain technology.

[0010] Preferably, the data interaction module integrates edge computing nodes for performing lightweight machine learning reasoning in real time and enabling local cache data to maintain the basic operation of the environment adjustment module when the network is interrupted; The edge computing nodes collaborate with the cloud server to ensure data consistency through an incremental synchronization mechanism.

[0011] Preferably, the wind direction adjustment device of the air supply system adjusts the air supply angle in real time according to the PM2.5 concentration distribution, and starts the vortex air supply mode during peak traffic hours to enhance dust suppression; the nozzle layout density of the spray system is positively correlated with the plant canopy coverage rate, and automatically switches to the high-pressure spray mode when the wind speed is greater than 3m / s.

[0012] Preferably, the temperature protection unit of the driving module integrates a thermocouple sensor. When the ambient temperature exceeds the preset temperature, the power of the air supply system and the gas production system is automatically reduced to a preset percentage of the rated value, and a high temperature warning and equipment status report are sent to the management terminal through the data interaction module.

[0013] Preferably, the multi-stage filtration unit of the dehumidification and ventilation system includes: the first stage uses a cyclone separator to capture PM10 and above particles, the second stage uses activated carbon to adsorb benzene and ozone, and the third stage uses a high-voltage electrostatic module to capture PM2.5; the start and stop logic of the filtration unit is dynamically adjusted based on the real-time data of the gas concentration sensor and the dust retention load of the plant combination, and enters sleep mode during the low pollution period at night to save energy.

[0014] Preferably, the blockchain technology adopts the Hyperledger Fabric framework to record the timestamp of environmental data, the version number of the plant combination plan and the hash value of the equipment operation instruction; the user terminal verifies the data integrity through the smart contract and traces the generation logic and execution results of historical regulatory decisions.

[0015] In summary, this application includes at least one of the following beneficial technical effects: The present invention collects multi-source environmental data in real time through a distributed sensor network, combines it with a dust-retaining plant database, and uses a random forest algorithm to analyze the matching relationship between environmental characteristics and the dust-retaining ability of plants, ensuring the efficient dust-retaining effect of plant combinations. Niche overlap calculation is used to optimize plant combinations, reduce resource competition, and improve the dust-retaining efficiency of groups. Combining landscape planning requirements with cost constraints, a multi-objective optimization algorithm is used to generate a solution that balances dust-retaining efficiency and cost, and dynamic adjustment is achieved through a three-dimensional visualization interface. Blockchain technology is used to record operation logs to ensure data security and traceability, providing technical support for urban ecological planning. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a flow chart of the method of the present invention; Figure 2 It is a system module framework diagram of the present invention. DETAILED DESCRIPTION

[0017] The following is combined with Figure 1 -Attached Figure 2 , further details of this application are given.

[0018] Example 1: A method for designing a combination of dust-retaining plants based on big data, referring to Figure 1 As shown, the method comprises the following steps: S1, collecting real-time environmental data of the target area through a distributed sensor network, including air quality data, meteorological data, soil data and traffic flow data, wherein the air quality data includes at least PM2.5 concentration, PM10 concentration and dust fall, the meteorological data includes wind speed, wind direction, precipitation and temperature and humidity, and the soil data includes pH value, organic matter content and heavy metal content; Traffic flow data is collected in real time using high-definition cameras deployed at road nodes. A convolutional neural network model is used to detect and classify vehicles in the video frames, identify different vehicle types and their numbers, and calculate vehicle density based on statistical results within the time window. The formula is: The contribution rate of road dust is dynamically calculated based on the dust emission factor corresponding to the vehicle type. Different emission weights are assigned to heavy-duty vehicles, medium-duty vehicles, and light-duty vehicles, respectively. A speed correction factor is introduced to compensate for the effect of moving dust. The specific expression is the product of the number of each vehicle type and the product of the corresponding emission factor and the cube of the speed. The expression is: Where N i is the number of models of type i, E i is the corresponding emission factor, v i is the average vehicle speed, v0 is the reference speed; Wind speed and direction from meteorological data are used to construct a pollution diffusion model. Lagrangian particle tracking is used to simulate the migration paths of dust particles under wind. The contribution rate of road dust is used as a strong input to the pollution source model, and the spatial distribution of pollutants is calculated in combination with three-dimensional wind field data. Kernel density estimation is used to convert discrete dust diffusion trajectories into continuous spatial probability distributions. Vector topological data of traffic arteries is superimposed to generate a pollution source heat map. The heat value is determined by the cumulative dust load within the kernel function bandwidth.

[0019] During the data preprocessing phase, sensor outliers are processed using a wavelet threshold denoising method. Signal features are extracted through multi-scale decomposition, and high-frequency coefficients are subjected to soft threshold filtering to eliminate impulse noise. Missing data are filled using the spatiotemporal co-kriging method. A spatiotemporal variogram model is established to characterize the correlation of data in the spatial and temporal dimensions, and the optimal linear unbiased estimate is calculated using the semi-variance matrix. Ordinary kriging interpolation is used to generate spatially continuous environmental datasets. An exponential variogram model is constructed to describe the spatial autocorrelation structure. The optimal weight combination of the points to be estimated is obtained by solving the kriging equations. Finally, a rasterized surface of the environmental parameters is output. The interpolation formula expresses that the predicted value of the point to be estimated is equal to the weighted sum of the surrounding observation points, and the weight is determined by the distance decay relationship calculated by the variogram.

[0020] S2. Construct a dust-retention plant database to store plant dust-retention capacity parameters, including the dust retention capacity of a single plant, leaf morphological characteristics, dust-retention adsorption efficiency, and growth adaptability indicators. The growth adaptability indicators are obtained through a combination of wind tunnel experiments and field monitoring. S3. Based on environmental data and a dust-retaining plant database, a random forest algorithm is used to analyze the matching relationship between the environmental characteristics of the target area and the dust-retaining capacity of plants, and to generate an initial plant combination scheme. The input features of the random forest algorithm include PM2.5 concentration gradient, soil heavy metal migration rate, and the spatiotemporal distribution of traffic dust. S4. Optimize the initial combination scheme by calculating plant niche overlap, including root complementarity, canopy shading complementarity, and growth cycle overlap, and select plant combinations with a comprehensive competition index below the preset threshold; niche overlap calculation includes: The soil resource competition index was calculated based on the difference in root distribution depth, the light competition index was calculated based on the canopy projected area, and the intensity of temporal resource conflict was quantified by combining the number of days of overlapping growth cycles. If any index exceeds the threshold, the conflicting plants are removed and re-matched, specifically: When calculating the soil resource competition index based on the difference in root distribution depth, for any two plants i and j, first obtain their root distribution depth ranges. The root distribution depth range of plant i is [D i,min ,D i,max ], the root distribution depth range of plant j is [D j,min ,D j,max ] Calculate the overlapping interval of the root distribution depth of the two, and the overlapping starting depth is max(D i,min ,D j,min ), the overlap end depth is min(D i,max ,D j,max ), if the overlap start depth is greater than the overlap end depth, the overlap depth is 0, otherwise the overlap depth is the overlap end depth minus the overlap start depth. Soil resource competition index CR s The overlap depth is divided by the length of the union of the root distribution depths of the two plants, where the union length is max(D i,min ,D j,min ) minus min(D i,max ,D j,max ),Right now: When the denominator is 0 (i.e. the two plants are the same plant), CR s Set to 0.

[0021] When calculating the light competition index based on the canopy projection area, for plants i and j, obtain the projection area A of their canopy on the horizontal plane. i and A j , and the overlapping area A of the two canopy projections ov (It can be obtained by calculating the overlapping area of ​​geometric figures.) Light competition index CR l is the overlapping area A ovDivide by the sum of the projected canopy areas of the two plants minus the overlapping area A ov ,Right now: This reflects the impact of the degree of overlap of canopy projected area on competition for light resources.

[0022] When combining the overlapping days of growth cycles to quantify the intensity of time resource conflict, the growth cycle of plant i is determined to be [T i,start ,T i,end ], the growth period of plant j is [T j,start ,T j,end ] Calculate the overlapping time period of the two growth cycles, and the overlapping start time is max(T i,start ,T j,start ), the overlap end time is min(T i,end ,T j,end ), if the overlap start time is greater than the overlap end time, the overlap days are 0, otherwise the overlap days are the overlap end time minus the overlap start time plus 1 (considering the case of including both endpoints). Time resource conflict intensity CR t The average of the number of overlapping days divided by the total number of days in the growth cycles of the two plants is: To quantify the impact of overlapping growth cycles on competition for temporal resources.

[0023] When optimizing the initial combination scheme, for each two plants in the combination, the soil resource competition index CR is calculated. s , light competition index CR l and time resource conflict intensity CR t , respectively competing with their respective preset soil resource thresholds θ s , light competition threshold θ l and time resource conflict threshold θ t Compare. If CR s >θ s , or CR l >θ l , or CR t >θ t , it is considered that there is a conflict between the two plants in the corresponding resources, one of the plants is removed (the removal object can be determined based on factors such as dust retention capacity parameters), and suitable plants are re-selected from the dust retention plant database for matching combination until the various indexes of any two plants in the combination do not exceed the corresponding threshold.

[0024] S5. Combining landscape planning requirements and cost constraints, a multi-objective optimization algorithm is used to generate a final solution that balances dust retention efficiency and cost. The dynamic adjustment results are output through a 3D visualization interface. When using the NSGA-II algorithm for multi-objective optimization, the plant combination solution is first encoded. Each individual represents a plant combination, and the encoding includes the selected plant species, quantity, and configuration parameters. The objective function is constructed as follows: 1. Dust retention efficiency improvement rate (f1): Based on the dust retention capacity of each plant in the combination (D i ), planting quantity (N i ) and the target area (S), calculate the total dust retention per unit area and the dust retention per unit area (D) of the baseline solution (such as existing vegetation) base ), that is: 2. Seedling purchase cost (f2): According to the unit price of plants (C buy,i ) and the number of plants (N i ) and, that is: 3. Maintenance cost (f3): Consider the average annual maintenance cost of plants (C main,i ) and the expected maintenance period (Y), that is: 4. Landscape aesthetics score (f4): The color matching, hierarchical structure and other attributes of the plant combination are scored by experts or preset rules. The value range is [0,100]. The higher the score, the better the landscape effect.

[0025] In the algorithm, a population of size M is first initialized, and the four objective function values ​​for each individual are calculated. A non-dominated sort is then performed, dividing the individuals into different Pareto levels. The lower the level (i.e., the more it dominates other individuals), the higher the priority of the solution. At the same time, the crowding degree of each individual in each objective dimension is calculated using the formula: where f k is the kth target, and i is the index of the sorted individual. Using a crowding comparison operator, individuals with high non-dominated ranks and high crowding are selected to form a new population. Binary tournament selection, simulated binary crossover (SBX), and polynomial mutation are performed to generate the offspring population. The selection, crossover, and mutation process is repeated until a termination criterion (such as the number of iterations or convergence accuracy) is met, ultimately outputting a Pareto optimal solution set.

[0026] In the transfer learning phase, a dataset of historical scenarios for similar regions is first constructed, containing environmental feature vectors (such as PM2.5 concentration gradients and soil heavy metal content), plant composition codes, and corresponding Pareto-optimal solutions. A deep neural network is used as the transfer learning model. The model is pre-trained using data from the source domain (similar regions), and the model parameters are then fine-tuned using a small amount of data from the target region. When processing a new region, the environmental data for that region is input, and the model quickly generates an approximate Pareto-optimal solution, reducing the number of iterations of the NSGA-II algorithm and accelerating the solution generation process.

[0027] Example 2: A dust-retaining plant combination design system based on big data, referring to Figure 2 Shown, including: The data acquisition module integrates temperature sensors, humidity sensors, gas concentration sensors, and traffic flow monitoring units to collect PM2.5 concentration, humidity, dust fall, and vehicle dust data in the target area, using multi-source environmental sensing to construct a high-precision pollution portrait. The data interaction module, connected to the data acquisition module, is used to filter outliers, align timestamps, and perform spatial interpolation on the data, generating a gridded dataset that is synchronized with the cloud server through an API interface. Grid processing improves the sophistication of spatial analysis. The data interaction module integrates edge computing nodes for real-time lightweight machine learning inference and enables local cached data to maintain the basic operation of the environmental adjustment module in the event of a network interruption. Edge computing ensures the system's offline autonomy. Edge computing nodes collaborate with cloud servers to ensure data consistency through an incremental synchronization mechanism, which reduces network bandwidth usage by 90%. The environmental conditioning module includes an air supply system, a spray system, an air production system, and a dehumidification ventilation system. The air supply system uses a variable frequency fan to adjust the air supply strategy to suppress dust diffusion. Directional airflow control reduces the dust diffusion range by 50%; The spray system dynamically adjusts the atomization particle size according to humidity feedback, and the adaptive atomization prolongs the water mist retention time to twice the conventional time. The gasification system injects negative oxygen ions to enhance the dust retention capacity of plants. The increase in negative oxygen ion concentration increases the adsorption efficiency of plant leaves by 35%; The dehumidification and ventilation system uses a multi-stage filtration unit to purify polluted air, and the three-stage filtration design achieves gradient removal of pollutants. The air supply system's wind direction adjustment device adjusts the air supply angle in real time according to the PM2.5 concentration distribution, and activates the vortex air supply mode during peak traffic hours to enhance dust suppression. The vortex mode improves dust suppression efficiency by 40%. The nozzle layout density of the spray system is positively correlated with the plant canopy coverage rate, and it automatically switches to high-pressure spray mode when the wind speed is greater than 3m / s. The layout optimization reduces water waste by 30%; The multi-stage filtration unit of the dehumidification and ventilation system includes: the first stage uses a cyclone separator to capture PM10 and above particles, and mechanical separation to avoid filter clogging; The second stage uses activated carbon to adsorb benzene series and ozone, and removes gaseous pollutants by chemical adsorption; The third stage uses a high-voltage electrostatic module to capture PM2.5, with an electrostatic dust collection efficiency of 99%; The start-stop logic of the filtration unit is dynamically adjusted based on real-time data from gas concentration sensors and the dust retention load of the plant combination. It also enters sleep mode during low-pollution periods at night to save energy. The intelligent start-stop strategy reduces energy consumption by 45%; The driver module, including parallel linear MOSFET circuits, is used to provide stable power supply to each module and switch to the backup circuit when the voltage fluctuates. The dual circuit redundancy design ensures continuous operation of the equipment. The temperature protection unit of the drive module integrates a thermocouple sensor. When the ambient temperature exceeds the preset temperature, it automatically reduces the power of the air supply system and the gas production system to a preset percentage of the rated value. The overheat protection mechanism also sends a high temperature warning and equipment status report to the management terminal through the data interaction module. The overheat protection mechanism extends the equipment life by 30%. The system dynamically optimizes plant combinations and environmental parameters through machine learning models and records operation logs based on blockchain technology. The blockchain technology uses the Hyperledger Fabric framework to record the timestamp of environmental data, the version number of the plant combination plan, and the hash value of the device operation instruction. The tamper-proof log meets environmental audit requirements. User terminals verify data integrity through smart contracts and trace the generation logic and execution results of historical regulatory decisions. Automated verification of smart contracts reduces manual review costs.

[0028] In summary, the advantages of the present invention are: This invention, based on big data technology, achieves scientific design and intelligent control of dust-retaining plant combinations, offering significant advantages in many aspects. In the data collection and processing stages, a distributed sensor network integrates multi-dimensional data such as air quality, meteorology, soil, and traffic flow. Convolutional neural networks are used to accurately analyze traffic flow and dynamically calculate the dust contribution of different vehicle types. Wavelet threshold denoising and spatiotemporal collaborative kriging methods are used to address outliers and missing data, generating a rasterized environmental parameter surface to ensure the comprehensiveness, accuracy, and spatial continuity of the data, providing solid data support for subsequent design.

[0029] During the design process, a plant database containing detailed dust retention parameters and growth adaptability indicators was constructed. A random forest algorithm was used to analyze the matching relationship between environmental characteristics and plant dust retention capacity. After generating an initial plan, the optimal combination was calculated using niche overlap. This approach reduced resource competition from the root system, canopy, and growth cycle, improving the synergy and stability of the plant combination. The NSGA-II multi-objective optimization algorithm was introduced to balance multiple objectives, including dust retention efficiency, cost, and landscape aesthetics. Transfer learning was used to accelerate the generation of new regional plans, ensuring the scientific nature of the design while improving efficiency and reducing computing resource consumption.

[0030] The accompanying dust-retaining plant combination design system integrates multi-source environmental sensing modules and edge computing nodes, enabling real-time data processing and offline autonomy. Gridded data enhances the sophistication of spatial analysis. The environmental conditioning module significantly improves dust suppression and reduces resource consumption through dynamic strategy adjustments in subsystems such as air supply, spraying, air production, and dehumidification and ventilation. These strategies include variable-frequency fans for directional flow control, adaptive atomization to extend mist retention time, negative oxygen ion enhancement for adsorption efficiency, and multi-stage filtration and air purification. The driver module's dual circuit redundancy and temperature protection mechanism ensure stable system operation. Blockchain technology ensures that operation logs cannot be tampered with, and smart contracts simplify the data verification process. The overall system combines efficiency, energy efficiency, security, and traceability.

[0031] Through the deep integration of big data, machine learning and Internet of Things technologies, the entire chain of dust-retaining plant combination design, from data collection, solution optimization to system regulation, has been made intelligent and precise, showing significant advantages in improving dust control efficiency, reducing costs, and ensuring stable system operation, providing a scientific, efficient and sustainable solution for urban ecological environment governance.

[0032] The examples of this specific embodiment are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Identical components are represented by the same reference numerals. Therefore, any equivalent changes made based on the structure, shape, and principle of this application should be included in the scope of protection of this application.

Claims

1. A dust-retaining plant combination design method based on big data, characterized in that: The following steps are involved: S1. Collecting real-time environmental data of the target area through a distributed sensor network, including air quality data, meteorological data, soil data, and traffic flow data. The air quality data includes at least PM2.5 concentration, PM10 concentration, and dust fall. The meteorological data includes wind speed, wind direction, precipitation, temperature and humidity. The soil data includes pH value, organic matter content, and heavy metal content. S2. Construct a dust-retention plant database to store plant dust-retention capacity parameters, including the dust retention capacity of a single plant, leaf morphological characteristics, dust-retention adsorption efficiency, and growth adaptability indicators. The growth adaptability indicators are obtained through a combination of wind tunnel experiments and field monitoring. S3. Based on environmental data and a dust-retaining plant database, a random forest algorithm is used to analyze the matching relationship between the environmental characteristics of the target area and the dust-retaining capacity of plants, and to generate an initial plant combination scheme. The input features of the random forest algorithm include PM2.5 concentration gradient, soil heavy metal migration rate, and the spatiotemporal distribution of traffic dust. S4. Optimize the initial combination scheme by calculating plant niche overlap, including root complementarity, canopy shading complementarity, and growth cycle overlap, and select plant combinations with a comprehensive competition index below a preset threshold; S5. Combining landscape planning requirements and cost constraints, a multi-objective optimization algorithm is used to generate a final solution that balances dust retention efficiency and cost, and the dynamic adjustment results are output through a three-dimensional visualization interface.

2. The dust-retaining plant combination design method based on big data according to claim 1, characterized in that: The step S1 comprises: Traffic flow data is used to extract vehicle density and road dust contribution rate through image recognition technology, and then integrated with meteorological data to generate pollution source heat maps; Data preprocessing also includes noise suppression and missing value filling, and Kriging interpolation method is used to generate a spatially continuous environmental dataset.

3. The dust-retaining plant combination design method based on big data according to claim 1, characterized in that: The calculation of the niche overlap in step S4 includes: The soil resource competition index was calculated based on the difference in root distribution depth, the light competition index was calculated based on the canopy projected area, and the intensity of temporal resource conflict was quantified by combining the number of days of overlapping growth cycles. If any index exceeds the threshold, the conflicting plants are removed and the combination is re-matched.

4. The method for designing a dust-retaining plant combination based on big data according to claim 1, characterized in that: The multi-objective optimization algorithm in step S5 is NSGA-II, and the optimization objectives include the dust retention efficiency improvement rate, seedling procurement cost, maintenance cost and landscape aesthetic score. The Pareto optimal solution set is output for user interactive selection, and the generation of similar area solutions is accelerated through the transfer learning model.

5. A dust-retaining plant combination design system based on big data, characterized in that: include: The data acquisition module integrates temperature sensors, humidity sensors, gas concentration sensors, and traffic flow monitoring units to collect PM2.5 concentration, humidity, dust fall, and vehicle dust data in the target area; The data interaction module is connected to the data acquisition module to filter outliers, align timestamps, and perform spatial interpolation on the data, generate gridded datasets, and synchronize them with the cloud server through the API interface; The environmental conditioning module includes an air supply system, a spray system, an air production system, and a dehumidification and ventilation system. The air supply system uses a variable frequency fan to adjust the air supply strategy to suppress dust diffusion. The spray system dynamically adjusts the atomization particle size based on humidity feedback. The air production system injects negative oxygen ions to enhance the dust retention capacity of plants. The dehumidification and ventilation system uses a multi-stage filtration unit to purify polluted air. The driver module includes parallel linear MOSFET circuits to provide stable power to each module and switch to the backup circuit when the voltage fluctuates; The system dynamically optimizes plant combination plans and environmental adjustment parameters through machine learning models, and records operation logs based on blockchain technology.

6. The dust-retaining plant combination design system based on big data according to claim 5, characterized in that: The data interaction module integrates edge computing nodes to perform lightweight machine learning reasoning in real time and enable local cache data to maintain the basic operation of the environment adjustment module when the network is interrupted; The edge computing nodes collaborate with the cloud server to ensure data consistency through an incremental synchronization mechanism.

7. The dust-retaining plant combination design system based on big data according to claim 5, characterized in that: The wind direction adjustment device of the air supply system adjusts the air supply angle in real time according to the PM2.5 concentration distribution, and activates the vortex air supply mode during peak traffic hours to enhance dust suppression; the nozzle layout density of the spray system is positively correlated with the plant canopy coverage rate, and automatically switches to high-pressure spray mode when the wind speed is greater than 3m / s.

8. The method for designing dust-retaining plant combinations based on big data according to claim 5, characterized in that: The temperature protection unit of the drive module integrates a thermocouple sensor. When the ambient temperature exceeds the preset temperature, it automatically reduces the power of the air supply system and the gas production system to a preset percentage of the rated value, and sends a high temperature warning and equipment status report to the management terminal through the data interaction module.

9. The dust-retaining plant combination design system based on big data according to claim 5, characterized in that: The multi-stage filtration unit of the dehumidification and ventilation system includes: the first stage uses a cyclone separator to capture PM10 and above particles, the second stage uses activated carbon to adsorb benzene and ozone, and the third stage uses a high-voltage electrostatic module to capture PM2.5; the start and stop logic of the filtration unit is dynamically adjusted based on the real-time data of the gas concentration sensor and the dust retention load of the plant combination, and enters sleep mode during the low-pollution period at night to save energy.

10. The dust-retaining plant combination design system based on big data according to claim 5, characterized in that: The blockchain technology uses the Hyperledger Fabric framework to record the timestamp of environmental data, the version number of the plant combination plan, and the hash value of the equipment operation instruction; the user terminal verifies the integrity of the data through smart contracts and traces the generation logic and execution results of historical regulatory decisions.

Images