Urban roof photovoltaic greening suitability intelligent evaluation system

Abstract

This invention relates to the fields of photovoltaic energy utilization, intelligent assessment algorithms, and urban planning technology, specifically an intelligent assessment system for the suitability of urban rooftop photovoltaic greening. It includes a multi-source data acquisition module, a data preprocessing module, an intelligent assessment model construction module, a suitability level determination module, an optimization scheme generation module, and a visualization output module. The multi-source data acquisition module is used to acquire roof structure parameters, photovoltaic potential parameters, greening suitability parameters, and land use data of urban rooftops. This invention achieves the dual goals of "power generation gain + ecological improvement" through the synergistic optimization of photovoltaics and greening. On the one hand, plant transpiration reduces the temperature of photovoltaic panels, improving power generation efficiency; on the other hand, photovoltaic panel shading improves the plant growth environment, increases plant survival rate and carbon sequestration capacity, while simultaneously increasing urban green area, mitigating the heat island effect, and contributing to urban carbon emission reduction and ecological livability construction.

Description

Technical Field

[0001] This invention relates to the fields of photovoltaic energy utilization, intelligent assessment algorithms and urban planning technology, specifically an intelligent assessment system for the suitability of urban rooftop photovoltaic greening. Background Technology

[0002] With the deepening of the "dual carbon" goals and the continuous upgrading of urban sustainable development needs, urban rooftops, as scarce idle space resources, are increasingly demonstrating their comprehensive utilization value. The "photovoltaic-greening integration" model, which combines photovoltaic power generation technology with ecological greening projects, can not only efficiently utilize solar energy resources for power generation, but also increase urban green area, alleviate the heat island effect, and improve the living environment, becoming an important direction for urban green development.

[0003] However, existing technologies have significant limitations in practical applications and are difficult to meet the precise needs of photovoltaic greening collaborative construction: The assessment is often limited to a single dimension and lacks synergistic considerations: Existing technologies tend to focus on single-dimensional assessments, either calculating only the potential of rooftop photovoltaics (such as solar radiation and power generation capacity) or analyzing only the feasibility of rooftop greening (such as soil conditions and plant adaptability), without constructing a comprehensive assessment system for the coordinated development of photovoltaics and greening. This single-dimensional assessment ignores the mutual influence and synergistic effects between the two. For example, shading by photovoltaic panels can reduce water consumption from plant transpiration and improve plant growth stability, while plant transpiration can lower the surface temperature of photovoltaic panels and improve power generation efficiency. Consequently, the assessment results fail to reflect the comprehensive value and actual feasibility of the project.

[0004] The lack of core influencing factors and insufficient assessment accuracy are issues: Existing assessment methods fail to systematically integrate key indicators such as roof structural safety (e.g., load-bearing limit, slope, and degree of structural aging), physiological adaptability of greening plants (e.g., drought tolerance, shade tolerance, and root depth), and environmental synergistic effects (e.g., carbon sequestration, rainwater retention rate, and cooling effect). For example, some technologies determine photovoltaic feasibility solely based on roof area and solar radiation, without considering potential safety hazards caused by insufficient roof load-bearing capacity; some greening feasibility analyses fail to consider local meteorological conditions and plant physiological characteristics, resulting in low plant survival rates and poor greening effects.

[0005] Insufficient feasibility and lack of targeted optimization solutions: Existing assessment results are mostly qualitative conclusions (such as "buildable" or "not buildable"), without providing specific and actionable optimization solutions for different limiting factors. For example, for roofs with load-bearing capacity nearing their limits, there are no clear recommendations for lightweight photovoltaic panel types or modular greening schemes; for roofs with severe shading, no suggestions are given for adjusting the layout of photovoltaic panels or adapting plant varieties. This leads to a disconnect between assessment results and actual project implementation, resulting in a significant waste of rooftop resources or project failure due to unreasonable solutions.

[0006] Therefore, in response to the problems of existing urban rooftop photovoltaic greening technologies, such as single evaluation dimensions, failure to consider the synergistic effect of photovoltaics and greening, lack of core influencing factors, insufficient evaluation accuracy, poor implementation of evaluation results, and lack of targeted optimization solutions, the evaluation results cannot reflect the comprehensive value and actual feasibility of the project, which can easily lead to waste of rooftop resources or project construction failure, and it is difficult to meet the precise needs of photovoltaic greening synergistic construction, an intelligent evaluation system for the suitability of urban rooftop photovoltaic greening is provided. Summary of the Invention

[0007] To address the aforementioned technical problems, this invention provides an intelligent assessment system for the suitability of urban rooftop photovoltaic greening. Specifically, the technical solution of this invention includes: It includes a multi-source data acquisition module, a data preprocessing module, an intelligent evaluation model construction module, a suitability level determination module, an optimization scheme generation module, and a visualization output module; The multi-source data acquisition module is used to acquire roof structure parameters, photovoltaic potential parameters, greening adaptation parameters, and land use data of urban rooftops. The data preprocessing module is used to clean, normalize, and extract core features from the collected multi-source data; the intelligent evaluation model construction module is based on the fusion algorithm of "random forest algorithm + improved BP neural network", which integrates four dimensions of indicators: photovoltaic power generation potential, roof structure safety, green plant growth adaptability, and environmental impact synergy, to construct a multi-dimensional coupled intelligent evaluation model. The suitability level determination module classifies buildings into four levels—high suitability, medium suitability, low suitability, and unsuitability—based on the comprehensive evaluation score output by the intelligent evaluation model. The optimization scheme generation module automatically generates personalized optimization schemes for different suitability levels and core limiting factors; The visualization output module generates evaluation results and optimization schemes through GIS map visualization and standardized report generation.

[0008] Preferably, the acquisition parameters of the multi-source data acquisition module include: Roof structural parameters: roof load-bearing limit, roof slope, roof area, roof material, structural service life, and structural damage status; Photovoltaic potential parameters: total annual solar radiation, direct radiation, diffuse radiation, surrounding shading rate, sunshine duration, and usable area for photovoltaic installation; Greening adaptation parameters: average annual temperature, annual precipitation, air humidity, soil type, soil fertility, soil moisture content, plant drought tolerance / shade tolerance / root depth / transpiration; Land use data: land use type and urban planning restrictions for the plot; The data collection method employs a multi-channel combination approach: real-time sensor acquisition, satellite / GIS data access, database retrieval, and supplementary field surveys.

[0009] Preferably, in the intelligent evaluation model construction module: The weights of the four dimensions were determined by combining the Analytic Hierarchy Process (AHP) and the CRITIC method, and the sum of the weights of the four dimensions was 1. The model training dataset has a sample size of ≥10,000, and the ratio of training set to test set is 7:3. The model evaluation uses accuracy, precision, recall, and F1 score as validation metrics, with accuracy ≥ 90%.

[0010] Preferably, the comprehensive evaluation score of the suitability level determination module is 0-100 points, and the level classification standard is as follows: A score of 80-100 indicates a high suitability for construction. A score of 60-79 indicates a moderately suitable building level; A score of 40-59 indicates a low suitability for construction. A score of 0-39 indicates an unsuitable building level.

[0011] Preferably, the types of limiting factors targeted by the optimization scheme generation module include roof structure type, photovoltaic potential type, greening adaptation type, and environmental synergy type, and the corresponding optimization schemes include: Roof structure restrictions: We recommend lightweight photovoltaic panels, modular greening planting boxes, or adjusting the installation angle of photovoltaic panels and using non-slip greening substrate; Limitations on photovoltaic potential: Optimize photovoltaic panel layout, adopt adjustable-angle photovoltaic brackets, or select high-efficiency photovoltaic cells, or reduce the photovoltaic installation area; Greenery compatibility restrictions: Recommended plant varieties are suitable, or soil improvement and modular soilless cultivation technology can be used; Environmental synergy constraints: Optimize the spacing between photovoltaic panels and green plants, select plants with high transpiration / high carbon sequestration, or adjust the proportion of plant varieties.

[0012] Preferably, the data preprocessing module processes data in the following ways: Outliers were removed using box plots, and missing data were filled in using linear interpolation or random forest interpolation. A standardized formula was used to convert parameters of different dimensions into data in the [0,1] interval. Principal component analysis (PCA) is used to extract core features and reduce data redundancy.

[0013] Preferably, the improved BP neural network in the intelligent evaluation model introduces a momentum factor to optimize the convergence speed; The number of decision trees in the random forest algorithm is 100-200; The number of hidden layer nodes in a BP neural network is 32-64.

[0014] Preferably, the content output by the visualization output module includes: GIS visualization map: Marking the suitability level of each rooftop (different colors to distinguish them), core limiting factors, and a schematic diagram of photovoltaic greening layout; Standardized assessment report: includes basic roof information, data collection results, comprehensive assessment score, suitability level, analysis of limiting factors, details of optimization schemes, investment estimate and benefit forecast.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention integrates four dimensions and 20+ core indicators, including photovoltaic power generation potential, roof structure safety, adaptability of green plants, and synergy of environmental impact, breaking through the limitations of single-dimensional evaluation. It fully considers the mutual influence and synergistic effect between photovoltaics and greening, ensuring that the evaluation results comprehensively reflect the actual feasibility and comprehensive value of rooftop photovoltaic greening construction. 2. This invention is based on the fusion of multi-source data and the fusion algorithm of "random forest + improved BP neural network" to achieve deep coupling calculation of multi-dimensional factors, which greatly improves the accuracy of the model and significantly enhances the evaluation precision; the automated data processing and evaluation process avoids the subjectivity and lag of manual evaluation, improves evaluation efficiency, and provides fast and accurate support for project decision-making. 3. This invention provides personalized and operable optimization solutions for different suitability levels and core limiting factors, covering all aspects such as equipment selection, plant recommendations, layout design, and cost estimation, solving the problem of "disconnect between assessment and implementation", reducing project construction risks, and improving the utilization rate of rooftop resources; 4. This invention achieves the dual goals of "power generation gain + ecological improvement" through the synergistic optimization of photovoltaics and greening. On the one hand, plant transpiration reduces the temperature of photovoltaic panels, improving power generation efficiency; on the other hand, the shading provided by photovoltaic panels improves the plant growth environment, increases plant survival rate and carbon sequestration capacity, while also increasing urban green area, mitigating the heat island effect, and contributing to urban carbon emission reduction and the construction of an ecologically livable city. Attached Figure Description

[0016] The present invention will be further explained below with reference to the accompanying drawings and embodiments: Figure 1 This is a module architecture diagram of the intelligent assessment system for the suitability of urban rooftop photovoltaic greening according to the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0018] Please see Figure 1 This invention provides an intelligent assessment system for the suitability of urban rooftop photovoltaic greening, comprising: a multi-source data acquisition module; This module forms the foundation of the evaluation system, used to comprehensively and accurately acquire various core data required for urban rooftop photovoltaic greening construction. It employs a multi-channel data collection approach combining real-time sensor acquisition, satellite / GIS data access, database retrieval, and on-site surveys to ensure the comprehensiveness, accuracy, and timeliness of the data. Specific collection parameters and methods are as follows: Roof structural parameters include: roof load-bearing capacity (unit: kN / m²), roof slope (unit: °), roof area (unit: m²), roof material (concrete, steel structure, roofing membrane, tiles, etc.), structural service life (unit: years), and structural damage status (no damage, minor damage, severe damage). The roof load-bearing capacity and structural service life are obtained by accessing building archive databases; the roof material and structural damage status are confirmed through on-site surveys; the roof slope and roof area are obtained by combining GIS system mapping data with on-site measurements; and some key structural parameters (such as real-time load-bearing stress) are collected in real-time by installing structural monitoring sensors.

[0019] Photovoltaic potential parameters include total annual solar radiation (kWh / m²), direct radiation (kWh / m²), diffuse radiation (kWh / m²), surrounding building / vegetation shading rate (%), sunshine duration (h / year), and usable photovoltaic panel installation area (m²). Total solar radiation, direct radiation, and diffuse radiation are obtained by accessing satellite remote sensing data (such as AW3D30DSM, NASADEM) and local meteorological station databases; surrounding shading rate is calculated using UAV aerial surveying combined with light environment simulation software (such as Radiance); sunshine duration and usable photovoltaic installation area are determined by GIS system data and on-site surveys after removing rooftop obstructions (such as ventilation openings and cooling towers).

[0020] Green space suitability parameters include: average annual temperature (°C), annual precipitation (mm), air humidity (%), soil type (loam, sandy soil, clay), soil fertility (organic matter content, %), soil moisture content (%), and physiological characteristics of selectable plant varieties (drought tolerance, shade tolerance, growth cycle, root depth, transpiration). Average annual temperature, annual precipitation, and air humidity are obtained by accessing the city's meteorological database; soil type, soil fertility, and soil moisture content are obtained through rooftop soil sample testing combined with real-time monitoring by soil sensors; plant physiological characteristics are obtained by retrieving information from a professional plant physiological characteristics database, and a list of plant varieties adapted to the local climate and rooftop environment is selected.

[0021] Land use data includes the land use type of the plot to which the rooftop is located (residential land, public management and public service land, commercial service facilities land, industrial land, logistics and warehousing land, etc.) and urban planning restrictions (such as green coverage requirements, photovoltaic construction indicators, building height restrictions, etc.). This data is obtained by accessing urban AOI (Area of ​​Interest) data and planning department databases to ensure that the assessment results comply with the overall urban planning requirements.

[0022] In one embodiment of the present invention, a data preprocessing module is included. Due to issues such as inconsistent dimensions, missing data, and outliers in multi-source data, this module performs standardization processing to ensure data quality and consistency, providing reliable data input for subsequent intelligent evaluation models. The specific processing method is as follows: Data cleaning: Box plot method is used to identify and remove abnormal data caused by sensor failure, survey error, and abnormal data transmission (e.g., the load-bearing data of a roof exceeds the structural design standard by more than 30%, which is judged as abnormal data); for missing data (e.g., some roof soil moisture content data are missing), linear interpolation method (suitable for time-series continuous data, such as meteorological data) or random forest interpolation method (suitable for non-linear correlation data, such as soil parameters) is selected according to the data type to complete the data and ensure data integrity.

[0023] Data Normalization: Due to the significant differences in the dimensions of different parameters (e.g., solar radiation is measured in kWh / m², while soil moisture content is measured in %), direct coupling calculations are not feasible. Therefore, all parameters must be converted to standardized data within the [0,1] interval. For positive indicators (larger values ​​are more conducive to building suitability, such as total solar radiation and load-bearing capacity), the formula is: xnorm = xmax − xmin; for negative indicators (larger values ​​are less conducive to building suitability, such as shading rate and structural service life), the formula is: xnorm = xmax − xmin, where x is the original data, xmin is the minimum value of the parameter, and xmax is the maximum value of the parameter.

[0024] Feature extraction: Principal component analysis (PCA) is used to reduce the dimensionality of the normalized high-dimensional data and extract core features. Specifically, the variance contribution of each parameter is calculated, and key features with a cumulative variance contribution of over 85% are selected (such as total solar radiation, roof load-bearing limit, plant drought tolerance, shading rate, etc.). Redundant data is eliminated to reduce the computational load of the model and improve evaluation efficiency.

[0025] In one embodiment of the present invention, an intelligent assessment model construction module is included. This module is the core of the invention and is based on a fusion algorithm of "random forest algorithm + improved BP neural network". It integrates four dimensions of indicators: photovoltaic power generation potential, roof structure safety, green plant growth adaptability, and environmental impact synergy, to construct a multi-dimensional coupled intelligent assessment model, thereby achieving accurate assessment of the suitability of rooftop photovoltaic greening. The specific construction process is as follows: Dimensional Indicator System Construction: Four primary dimensions and corresponding secondary sub-indicators are clearly defined to ensure that the evaluation dimensions comprehensively cover the core influencing factors of photovoltaic-greening collaborative construction. The primary dimensions include photovoltaic power generation potential (W1), roof structure safety (W2), adaptability to green plant growth (W3), and environmental impact synergy (W4), and satisfy the condition W1+W2+W3+W4=1; the secondary sub-indicators are as follows: Photovoltaic power generation potential (W1): includes total solar radiation, surrounding shading rate, and available area for photovoltaic installation; Roof structural safety (W2): includes roof load-bearing limit, roof slope, structural service life, and roof material stability; Green plant growth suitability (W3): This includes the degree of matching with meteorological conditions (the degree of matching between annual average temperature, annual precipitation, air humidity and plant growth requirements), soil suitability (the degree of matching between soil type, fertility, water content and plant requirements), and the degree of matching with plant physiological characteristics (the degree of matching between drought tolerance, shade tolerance and roof environment). Environmental impact synergy (W4): This includes the cooling gain of photovoltaic panels (the reduction in photovoltaic panel temperature due to plant transpiration), plant carbon sequestration, rainwater interception rate, and the degree of mitigation of the urban heat island effect.

[0026] Indicator weight determination: A combination of the Analytic Hierarchy Process (AHP) and the CRITIC method was used to determine the weights of each dimension and sub-indicators, taking into account both subjective experience and objective data characteristics. First, experts in photovoltaic engineering, building structures, urban greening, and environmental engineering were invited to score the relative importance of each indicator using AHP, constructing a judgment matrix and calculating preliminary weights. Then, the CRITIC method was used to calculate objective weights based on data variability (standard deviation) and indicator conflict (correlation coefficient). Finally, the subjective and objective weights were weighted and merged in a 4:6 ratio to obtain the final indicator weights, ensuring the scientific and reasonable allocation of weights.

[0027] Algorithm selection and optimization: A fusion model of "random forest algorithm + improved BP neural network" is adopted to give full play to the advantages of both algorithms. Random Forest Algorithm: Used for feature importance ranking and preliminary evaluation. By constructing 100-200 decision trees (optimized according to dataset size), it votes on the core input features to select the indicators that have the most significant impact on suitability, reducing the interference of single indicators and outputting preliminary evaluation results; Improved BP Neural Network: Used for precise calculations with multi-dimensional coupling. A momentum factor (value 0.1-0.3) is introduced into the traditional BP neural network to optimize the algorithm's convergence speed and avoid getting trapped in local optima. The number of hidden layer nodes is set to 32-64 (adjusted according to the feature dimension), the sigmoid function is used as the activation function, and the output layer displays a comprehensive evaluation score (0-100 points).

[0028] Model training and validation: Dataset Construction: Collect historical project data from different climate regions (tropical, subtropical, temperate, and cold temperate), different city types (first-tier, second-tier, and third-tier cities), and different roof types (residential, commercial, and industrial) across the country. This includes successful cases (such as photovoltaic greening projects that have been operating stably) and failed cases (such as projects that were halted due to insufficient load-bearing capacity or large-scale plant death). Construct a dataset containing 10,000+ samples, which is divided into training and test sets in a 7:3 ratio. Model training: Input the training set data into the fusion model, iteratively update the model parameters through the backpropagation algorithm, and adjust hyperparameters such as the number of decision trees in the random forest, the number of hidden layer nodes in the BP neural network, and the momentum factor until the model converges; Model validation: Accuracy, precision, recall, and F1 score are used as validation metrics to evaluate the test set data. The model accuracy is required to be ≥90%, ensuring the reliability and stability of the evaluation results.

[0029] In one embodiment of the present invention, a suitability level determination module is included: Based on the comprehensive evaluation score (0-100 points) output by the intelligent assessment model, combined with industry technical standards, engineering practice experience, and urban planning requirements, a suitability level is determined to provide clear qualitative conclusions for project decision-making. The specific level classification standards are as follows: High suitability rating (80-100 points): The roof structure is safe and stable (sufficient load-bearing capacity, no damage, reasonable slope), with ample photovoltaic potential (high solar radiation, low shading rate, and large usable area), good adaptability to greening (highly compatible with meteorological conditions, soil conditions, and plant physiological characteristics), and significant synergistic effect between photovoltaics and greening (meeting standards for cooling gain, carbon sequestration, and rainwater interception). This type of roof can directly promote integrated photovoltaic and greening construction without additional optimization or adjustment.

[0030] Medium-level suitability (60-79 points): The core conditions of the roof are met, but there are 1-2 minor limiting factors (such as slight shading, insufficient soil fertility, and limited usable area for photovoltaics). After targeted optimization and adjustment, such roofs can be used for photovoltaic greening and construction, and their suitability can be upgraded to a high-level suitability.

[0031] Low suitability rating (40-59 points): Key limiting factors exist (such as roof load-bearing capacity approaching its limit, insufficient solar radiation, and poor plant compatibility), making it impossible to carry out complete integrated photovoltaic and greening construction. Significant optimization of the plan is required (such as replacing equipment with lightweight equipment or selecting specially adapted plants) or partial construction (such as only building a photovoltaic system or only carrying out greening projects) to avoid blind construction leading to resource waste or safety hazards.

[0032] Unsuitable for construction (0-39 points): Significant safety hazards exist (e.g., insufficient roof load-bearing capacity, severe structural damage, excessive slope that cannot be adjusted) or core conditions are not met (e.g., insufficient effective sunshine duration of less than 800 hours per year, lack of green soil foundation and inability to lay greenery). Photovoltaic greening construction is prohibited on such roofs. It is recommended to maintain the original usage status or carry out structural modifications before reassessment.

[0033] In one embodiment of the present invention, an optimization scheme generation module is included: This module automatically generates personalized and actionable optimization solutions for different suitability levels and core limiting factors (roof structure, photovoltaic potential, greening compatibility, and environmental synergy), achieving closed-loop support from "suitability assessment" to "solution implementation." Specific optimization solutions are as follows: Roof structural limitations (insufficient load-bearing capacity, unreasonable slope, structural damage): Insufficient load-bearing capacity: It is recommended to use lightweight photovoltaic panels (such as thin-film photovoltaic panels, weighing ≤10kg / m²) to replace traditional crystalline silicon photovoltaic panels; select modular greening planting boxes (each box weighing ≤50kg, including soil and plants) to control the greening load per unit area; prioritize the layout of photovoltaic panels and planting boxes in areas of the roof with strong local load-bearing capacity (such as above beams and columns).

[0034] Unreasonable slope (slope >30° or <5°): When the slope is >30°, adjust the installation angle of the photovoltaic panels to complement the roof slope (e.g., if the roof slope is 35°, set the photovoltaic panel installation angle to 25°), and use anti-slip fixing brackets; add anti-slip clips to the green planting boxes, and select creeping or low-growing plants to prevent them from slipping. When the slope is <5°, optimize the roof drainage design (add drainage channels), leave a 10-15cm ventilation gap at the bottom of the photovoltaic panels, add water-retaining agents to the greening substrate and set up an impermeable layer to prevent water accumulation and root rot.

[0035] Structural damage: Minorly damaged areas should be reinforced with carbon fiber before equipment can be installed; photovoltaic and greening installations are prohibited in severely damaged areas, and structural repair and renovation should be prioritized.

[0036] Limitations on photovoltaic potential (severe shading, insufficient solar radiation, limited usable photovoltaic area): Severe shading (shading rate >30%): Accurately locate the shading source through GIS system and drone mapping, optimize the layout of photovoltaic panels (avoid shading areas); adopt adjustable angle photovoltaic brackets (adjustment range 0-60°), dynamically adjust the angle according to the solar altitude angle to avoid shading; for areas where the shading source cannot be removed, prioritize the layout of shade-tolerant plants to reduce the installation density of photovoltaic panels.

[0037] Insufficient solar radiation (total annual radiation <1000kWh / m²): Select high-efficiency photovoltaic cells (conversion efficiency ≥22%) to improve power generation capacity; reduce the photovoltaic installation area, prioritize greening needs, and achieve a layout of "greening as the main focus and photovoltaic as a supplement"; combine energy storage equipment to store electrical energy and improve energy utilization efficiency.

[0038] Limited available photovoltaic area: Flexible photovoltaic panels are used to fit the curved roof surface, making full use of corner space; integrated photovoltaic greening components (integrated design of photovoltaic panels and planting boxes) are selected to improve space utilization.

[0039] Greenery compatibility limitations (climate incompatibility, soil incompatibility, low plant survival rate): Mismatched weather conditions (e.g., insufficient rainfall in arid regions, excessive humidity in humid regions): In arid regions, drought-tolerant plants (such as Crassulaceae and Caryophyllaceae plants) are recommended, along with water-saving irrigation systems (drip irrigation, micro-sprinkler irrigation); In humid regions, moisture-tolerant plants (such as Iridaceae and Araceae plants) are recommended, and roof drainage and ventilation designs should be optimized to reduce soil water accumulation.

[0040] Soil incompatibility (poor soil quality): Improve the soil by adding decomposed organic fertilizer, perlite, and water-retaining agents to improve soil fertility and aeration; for rooftops without soil foundation, use modular soilless cultivation technology (substrate is coconut coir, expanded clay, and rock wool) to reduce dependence on soil.

[0041] Low plant survival rate: Select local native plant varieties (strong adaptability and low maintenance cost); combine with mixed planting mode (trees + shrubs + herbs) to improve the stability of plant community; develop maintenance plan according to plant growth cycle (such as regular pruning, fertilization, and pest and disease control).

[0042] Environmental synergy limitations (insufficient cooling by photovoltaic panels, low carbon sequestration by plants, poor rainwater retention): Insufficient cooling by photovoltaic panels: Optimize the distance between photovoltaic panels and green plants (0.5-1m recommended) to ensure airflow; select plants with high transpiration rates (such as reeds and calamus) to reduce the ambient temperature through transpiration; use a reflective coating on the surface of the photovoltaic panels to reduce heat absorption.

[0043] Low carbon sequestration by plants: Increase the proportion of shrubs and trees (such as privet, osmanthus, and conifers) to enhance carbon sequestration capacity; extend the plant growth cycle (select perennial plants) to avoid frequent replacement of annual plants that could lead to interruption of carbon sequestration.

[0044] Poor rainwater retention rate: Select water-retaining greening substrate (add zeolite and vermiculite) to improve soil water retention capacity; set up a water storage layer and drainage filter at the bottom of the planting box to collect rainwater for irrigation and improve rainwater utilization rate.

[0045] In one embodiment of the present invention, a visualization output module is included: To facilitate users (urban planning departments, project construction units, and design units) in intuitively obtaining assessment results, this module combines GIS map visualization with standardized report generation to output assessment results, specifically including: GIS Visualization Map: Based on the city's GIS system, the suitability levels of each rooftop are marked with different colors (green for high suitability, blue for medium suitability, yellow for low suitability, and red for unsuitable). Clicking on a single rooftop allows you to view core limiting factors (such as "insufficient load-bearing capacity" and "high shading rate"), as well as a schematic diagram of the photovoltaic greening layout (including the installation location of photovoltaic panels, green planting areas, and equipment spacing). It supports zooming, panning, and filtering operations, facilitating both macro-planning and micro-level viewing.

[0046] Standardized Assessment Report: Automatically generates an assessment report in Word or PDF format, including the following core contents: basic roof information (address, land type, roof area, structural parameters), data collection results (specific values ​​of each dimension parameter, collection method), comprehensive assessment score and suitability level, detailed analysis of core limiting factors, details of personalized optimization solutions (including equipment selection, plant recommendations, layout design, cost estimation), and investment estimation and benefit prediction (power generation revenue, ecological benefit value, investment payback period). The report format is standardized and the content is detailed, which can be directly used as the basis for project initiation, approval, and design.

[0047] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. An intelligent assessment system for the suitability of urban rooftop photovoltaic greening, characterized in that, It includes a multi-source data acquisition module, a data preprocessing module, an intelligent evaluation model construction module, a suitability level determination module, an optimization scheme generation module, and a visualization output module; The multi-source data acquisition module is used to acquire roof structure parameters, photovoltaic potential parameters, greening adaptation parameters, and land use data of urban rooftops. The data preprocessing module is used to clean, normalize, and extract core features from the collected multi-source data; the intelligent evaluation model construction module is based on the fusion algorithm of "random forest algorithm + improved BP neural network", which integrates four dimensions of indicators: photovoltaic power generation potential, roof structure safety, green plant growth adaptability, and environmental impact synergy, to construct a multi-dimensional coupled intelligent evaluation model. The suitability level determination module classifies buildings into four levels—high suitability, medium suitability, low suitability, and unsuitability—based on the comprehensive evaluation score output by the intelligent evaluation model. The optimization scheme generation module automatically generates personalized optimization schemes for different suitability levels and core limiting factors; The visualization output module generates evaluation results and optimization schemes through GIS map visualization and standardized report generation.

2. The intelligent assessment system for the suitability of urban rooftop photovoltaic greening according to claim 1, characterized in that, The acquisition parameters of the multi-source data acquisition module include: Roof structural parameters: roof load-bearing limit, roof slope, roof area, roof material, structural service life, and structural damage status; Photovoltaic potential parameters: total annual solar radiation, direct radiation, diffuse radiation, surrounding shading rate, sunshine duration, and usable area for photovoltaic installation; Greening adaptation parameters: average annual temperature, annual precipitation, air humidity, soil type, soil fertility, soil moisture content, plant drought tolerance / shade tolerance / root depth / transpiration; Land use data: land use type and urban planning restrictions for the plot; The data collection method employs a multi-channel combination approach: real-time sensor acquisition, satellite / GIS data access, database retrieval, and supplementary field surveys.

3. The intelligent assessment system for the suitability of urban rooftop photovoltaic greening according to claim 1, characterized in that, In the intelligent evaluation model construction module: The weights of the four dimensions were determined by combining the Analytic Hierarchy Process (AHP) and the CRITIC method, and the sum of the weights of the four dimensions was 1. The model training dataset has a sample size of ≥10,000, and the ratio of training set to test set is 7:

3. The model evaluation uses accuracy, precision, recall, and F1 score as validation metrics, with accuracy ≥ 90%.

4. The intelligent assessment system for the suitability of urban rooftop photovoltaic greening according to claim 1, characterized in that, The comprehensive evaluation score of the suitability assessment module is 0-100 points, and the classification criteria are as follows: A score of 80-100 indicates a high suitability for construction. A score of 60-79 indicates a moderately suitable building level; A score of 40-59 indicates a low suitability for construction. A score of 0-39 indicates an unsuitable building level.

5. The intelligent assessment system for the suitability of urban rooftop photovoltaic greening according to claim 1, characterized in that, The optimization scheme generation module targets limiting factors including roof structure, photovoltaic potential, greening compatibility, and environmental synergy, and the corresponding optimization schemes include: Roof structure restrictions: We recommend lightweight photovoltaic panels, modular greening planting boxes, or adjusting the installation angle of photovoltaic panels and using non-slip greening substrate; Limitations on photovoltaic potential: Optimize photovoltaic panel layout, adopt adjustable-angle photovoltaic brackets, or select high-efficiency photovoltaic cells, or reduce the photovoltaic installation area; Greenery compatibility restrictions: Recommended plant varieties are suitable, or soil improvement and modular soilless cultivation technology can be used; Environmental synergy constraints: Optimize the spacing between photovoltaic panels and green plants, select plants with high transpiration / high carbon sequestration, or adjust the proportion of plant varieties.

6. The intelligent assessment system for the suitability of urban rooftop photovoltaic greening according to claim 1, characterized in that, The data preprocessing module's processing methods include: Outliers were removed using box plots, and missing data were filled in using linear interpolation or random forest interpolation. A standardized formula was used to convert parameters of different dimensions into data in the [0,1] interval. Principal component analysis (PCA) is used to extract core features and reduce data redundancy.

7. The intelligent assessment system for the suitability of urban rooftop photovoltaic greening according to claim 1, characterized in that, The improved BP neural network in the intelligent evaluation model introduces a momentum factor to optimize the convergence speed; The number of decision trees in the random forest algorithm is 100-200; The number of hidden layer nodes in a BP neural network is 32-64.

8. The intelligent assessment system for the suitability of urban rooftop photovoltaic greening according to claim 1, characterized in that, The content output by the visualization output module includes: GIS visualization map: Marking the suitability level of each rooftop (different colors to distinguish them), core limiting factors, and a schematic diagram of photovoltaic greening layout; Standardized assessment report: includes basic roof information, data collection results, comprehensive assessment score, suitability level, analysis of limiting factors, details of optimization schemes, investment estimate and benefit forecast.

Images