A method and system for constructing mammal ecological corridors in urban green space systems

Abstract

This invention belongs to the field of ecological environment construction technology, specifically disclosing a method and system for constructing mammal ecological corridors in urban green space systems. The method includes: planning a corridor base containing peak paths and buffer chambers based on multi-source spatiotemporal data, and configuring functional vegetation; using phenologically matched trigger plants to attract animals along peak paths, and using responder vegetation in buffer chambers to respond to environmental disturbances and increase density to form dynamic barriers. Simultaneously, an adaptive operation system composed of a sensor network, edge computing nodes, and a cloud platform is deployed: edge nodes perceive animal wandering behavior in real time and drive audio-visual devices for flexible guidance; the cloud platform analyzes long-term performance data and automatically generates and executes adaptive management schemes for the corridor. This invention achieves real-time proactive guidance of animal behavior and adaptive management of environmental changes in the corridor, improving the actual effectiveness and intelligent management level of the ecological corridor.

Description

Technical Field

[0001] This invention belongs to the field of ecological environment construction technology, and relates to a method and system for constructing mammalian ecological corridors in urban green space systems. Background Technology

[0002] Urban green space systems are an important component of urban ecological infrastructure, and ecological corridors, as links connecting isolated green space patches, play a crucial role in maintaining urban biodiversity and ensuring gene exchange among wildlife populations. The construction of mammalian ecological corridors typically involves multiple technical fields such as horticulture, landscape ecology, and civil engineering. By scientifically configuring vegetation and constructing physical pathways, suitable migration environments are created for target species.

[0003] Existing technologies have made some progress in ecological corridor planning. Early methods were mostly based on geographic information systems for macro-level route selection. Recent improvements, such as the method for constructing a high-density urban ecological-recreation functional composite corridor network disclosed in Chinese Patent Publication No. CN119312965A, further consider the integration of human recreational activities and other composite functions into corridor planning, aiming to balance multiple needs through a more refined spatial layout. This approach represents a deeper understanding of the complex urban environment.

[0004] However, the aforementioned existing technical solutions have obvious functional limitations: 1. The corridor constructed by the existing methods is a passive physical space that cannot sense and respond to real-time dynamic events inside and around the corridor, such as brief bursts of bright light or sudden noise. When animals encounter such momentary disturbances during migration and show hesitation or retreat, the corridor itself does not have any ability to actively guide or provide immediate protection, which can easily lead to migration interruption.

[0005] 2. The urban environment is constantly changing, and the buildings, traffic, and human activity patterns around the corridor may change over time. Once the function of an existing corridor is set, it is difficult to adjust. It is impossible to actively adjust its internal functional configuration according to long-term changes in the external environment or the growth and succession of the vegetation itself, which may lead to a decline in its ecological effectiveness over time.

[0006] 3. The current management and effectiveness evaluation of corridors mainly rely on manual inspections and offline data analysis. This approach is slow to respond and makes it difficult to scientifically and quantitatively diagnose problems. The system lacks a built-in mechanism that can automatically monitor operational status, identify functional bottlenecks, and propose optimization suggestions. Summary of the Invention

[0007] In order to overcome the above-mentioned defects of the prior art and to achieve the above objectives, the present invention proposes the following technical solution: a method for constructing mammalian ecological corridors in urban green space systems, comprising: S1, acquiring activity trace data of target mammals, distribution data of urban disturbance sources, time data of key life activity periods of target mammals, time data of peak periods of urban human activities, and time data of natural rhythms, and performing spatiotemporal correlation on the acquired data to generate a spatiotemporal coupled database.

[0008] S2. Based on the spatiotemporal coupling database, the spatiotemporal overlap relationship between the peak periods of key activities and the peak periods of interference is analyzed, and a dynamic corridor base including peak path sections and buffer compartment sections is planned accordingly.

[0009] S3. At the planting sites of the dynamic corridor base, configure a functional modular vegetation system that includes trigger plant combinations and responder vegetation structures.

[0010] S4. Deploy a sensor network at the path decision nodes of the dynamic corridor base and configure the sensor network to collect animal movement signals in real time and generate animal behavior stream data.

[0011] S5. Deploy edge computing nodes and environmental regulation devices, and preset behavior guidance strategies in the edge computing nodes so that they are configured to receive animal behavior flow data, generate environmental regulation instructions according to the behavior guidance strategy, and control the environmental regulation devices to execute environmental regulation instructions, so as to achieve real-time guidance of the target mammal's movement path.

[0012] The second aspect of the present invention provides a system for constructing mammalian ecological corridors in urban green space systems, comprising: a data acquisition and processing module, which acquires activity trace data of target mammals, distribution data of urban disturbance sources, time data of key life activity periods of target mammals, time data of peak periods of urban human activities, and time data of natural rhythms, and performs spatiotemporal correlation on the acquired data to generate a spatiotemporally coupled database.

[0013] The corridor base planning module, based on a spatiotemporal coupled database, analyzes the spatiotemporal overlap between peak periods of key activities and peak periods of interference, and plans a dynamic corridor base that includes peak path sections and buffer compartment sections accordingly.

[0014] The vegetation system configuration module configures a functional modular vegetation system, including trigger plant combinations and responder vegetation structures, at the planting sites on the dynamic corridor base.

[0015] The behavior perception module deploys a sensor network at the path decision nodes of the dynamic corridor base and configures the sensor network to collect animal movement signals in real time and generate animal behavior flow data.

[0016] The real-time guidance module deploys edge computing nodes and environmental regulation devices, and presets behavior guidance strategies in the edge computing nodes. It is configured to receive animal behavior flow data, generate environmental regulation instructions according to the behavior guidance strategy, and control the environmental regulation devices to execute the environmental regulation instructions, so as to realize real-time guidance of the target mammal's movement path.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention improves the adaptability and long-term effectiveness of ecological corridors through the deep integration of spatiotemporal collaborative design and dynamic response mechanism. The method establishes a spatiotemporal coupling database to pre-identify and plan peak path sections and buffer chamber sections that respond to animal rhythms and urban disturbance dynamics. This non-uniform design based on prediction enables the corridor to have structural resilience to cope with complex urban environmental changes from the beginning of its construction, ensuring that it can continue to play an effective ecological connectivity function in long-term operation.

[0018] (2) This invention achieves synergistic enhancement and active regulation of corridor ecological functions by combining a functional modular vegetation system with automated control. The trigger plant combination attracts and guides animals using natural plant phenology, while the responder vegetation structure actively forms a physical barrier through triggered directional growth under environmental stress. This approach, which combines biological rhythms with engineering needs, transforms vegetation from merely a static food source or shelter into an active functional component in the corridor system, achieving an organic unity of multiple functions such as food supply, dynamic shielding, and behavioral guidance, thereby improving the overall efficiency of the system.

[0019] (3) This invention achieves immediate and precise fine-tuning of corridor usage by introducing a real-time guidance closed loop based on animal behavior flow data. By deploying sensor networks and environmental adjustment devices at key decision nodes, the system can perceive the hesitant behavior of animals in real time and immediately take non-invasive, flexible guidance measures. This neural reflex-like mechanism helps solve the problem of passage obstacles caused by unpredictable local disturbances in traditional corridors, reduces the passage pressure on animals, and thus improves the actual utilization rate of corridors and the success rate of animal passage. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of the implementation steps of the method of the present invention.

[0022] Figure 2 This is a schematic diagram of the system module connections of the present invention. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Example 1 Please see Figure 1 As shown, the present invention proposes a method for constructing mammalian ecological corridors in urban green space systems, including: S1, acquiring activity trace data of target mammals, distribution data of urban disturbance sources, time data of key life activity periods of target mammals, time data of peak periods of urban human activities, and time data of natural rhythms, and performing spatiotemporal correlation on the acquired data to generate a spatiotemporally coupled database.

[0025] In a preferred embodiment, the step of performing spatiotemporal correlation on the acquired data to generate a spatiotemporally coupled database includes: acquiring activity trace data containing geographic coordinates and timestamps through infrared camera monitoring and trace investigation.

[0026] By combining remote sensing mapping and ground monitoring, data on the distribution of urban interference sources, including light intensity and noise decibel values, can be obtained.

[0027] Data on key life activity periods, peak periods of urban human activities, and natural rhythms were obtained from ecological databases, urban management databases, and meteorological and astronomical databases, respectively.

[0028] The study area is divided into pre-defined geographical grid units.

[0029] For each geographic grid cell, based on a pre-set evaluation model, and combined with key life activity period time data, peak period time data of urban human activities, and natural rhythm time data, the animal activity intensity index obtained from activity trace data analysis, the normalized light intensity value, and the noise decibel value are weighted and calculated to obtain the dynamic spatiotemporal suitability index characterizing the grid cell at a specific time point.

[0030] The data, which includes the spatial coordinates, timestamps, and corresponding dynamic spatiotemporal suitability indices of each grid cell, are combined to generate a spatiotemporal coupled database.

[0031] Specifically, the purpose of generating the spatiotemporal coupled database in this embodiment is to standardize and fuse multi-source heterogeneous static spatial data with dynamic temporal data, providing a quantitative decision-making basis for subsequent non-uniform planning of the corridor base. This process includes three sequential engineering steps.

[0032] The first step involves collecting and vectorizing basic data. To obtain activity trace data, infrared camera arrays will be deployed in a grid pattern every 50 to 100 meters in potential animal activity areas, such as urban parks, woodlands, and riverbanks. This will be combined with trace surveys conducted along pre-defined transects to collect raw data including animal images, footprints, and feces. All data will include GPS geographic coordinates and collection timestamps. To obtain urban disturbance source distribution data, traffic road vector data from the urban geographic information system (GIS) will be used, along with hourly traffic flow data provided by traffic management departments. Simultaneously, drones equipped with calibrated photometers will be used to fly at night to create urban light intensity distribution maps with an accuracy within 5 meters. Decibel meters will be deployed at key locations such as along main roads and in commercial areas to collect environmental noise data for more than 72 consecutive hours, generating noise contour maps. This step ultimately outputs a series of vectorized layer files with geographic coordinates and time attributes.

[0033] The second step involves synchronizing and formatting multi-dimensional time data. To obtain time data for key life activity periods, the annual breeding season, juvenile rearing period, and subadult dispersal period of target mammals such as raccoon dogs and weasels will be retrieved from the ecological databases of local forestry departments or research institutions, typically accurate to the ten-day period. To obtain time data for peak periods of urban human activities, public holiday schedules, calendars of major cultural and sports events, and weekday morning and evening rush hours derived from public transportation card swipe data will be extracted from urban management databases. To obtain natural rhythm time data, standard meteorological and astronomical database interfaces will be called to obtain sunrise and sunset times, lunar phase cycles, especially the dates of new moon and full moon for the coming year. This step unifies all discrete time data into time series in hourly or daily units, outputting a standardized timetable.

[0034] The third step involves performing correlation mapping and spatiotemporal coupling calculations. The core of this step is establishing an evaluation model that quantitatively correlates spatial animal activity with disturbance intensity and temporal rhythm dynamics. The system divides the study area into 10m x 10m geographic grid units and calculates the suitability of each grid unit at a specific time point. Specifically, a dynamic spatiotemporal suitability index is used. The evaluation is conducted using the following formula: in, It represents the overall suitability index of a specific grid at a specific time. , , These are weighting coefficients, representing the importance of animal activity, light disturbance, and noise disturbance in the model, respectively. Their values ​​are preset by experts based on the ecological habits of the target species, and the sum of the three is 1. For example, for a species that is extremely photophobic, the weights might be set to 1. , , . The normalized animal activity intensity index reflects animals' preferences for specific grid cells as habitats, foraging grounds, or passageways. It is obtained by performing kernel density analysis on activity trace data within the grid, and its value is scaled to the [0,1] interval. This represents the area with the highest animal activity intensity within the study area. This represents areas where no activity was detected. Kernel density analysis is a spatial statistical method used to estimate the spatial distribution density of sample points. This represents the suitability of animal activity at a specific point in time. It is a moderating coefficient used to reflect the interaction between animal biorhythms and urban anthropogenic activity rhythms. Its specific value is determined based on a pre-defined species sensitivity model. This model integrates three types of time data: key life activity periods such as the breeding season, the rearing period, and the dispersal period. The value should be increased; such as during peak periods of urban human activity, such as weekday morning and evening rush hours and large-scale events on holidays. The value should be lowered; for natural rhythms such as day and night, especially for nocturnal animals, the nighttime... The value should be higher than during the day. For example, it can be set to be higher during off-peak traffic hours at night during the target species' dispersal period. During daytime traffic peak hours, It may be set to 0.1 or lower. and These are the normalized values ​​of the light intensity and noise decibels for the grid at that time point, respectively, derived from the aforementioned urban interference source distribution data.

[0035] By using the comprehensive suitability index calculation formula, the system assigns a series of dynamically changing suitability indices to each grid cell in the geographic information database. The final output of this process is the spatiotemporal coupled database, which is a multidimensional data set containing spatial coordinates, timestamps, and corresponding suitability indices, directly supporting the subsequent accurate identification of peak path segments and buffer zone segments.

[0036] S2. Based on the spatiotemporal coupling database, the spatiotemporal overlap relationship between the peak periods of key activities and the peak periods of interference is analyzed, and a dynamic corridor base including peak path sections and buffer compartment sections is planned accordingly.

[0037] In a preferred embodiment, the planning forms a dynamic corridor base that includes peak path sections and buffer compartment sections, including: calculating and generating potential migration paths connecting different green space nodes based on a spatiotemporal coupled database using a minimum cumulative resistance model, and marking the potential migration paths as peak path sections.

[0038] Spatial overlay analysis is performed on the peak path segment and the area where the interference value exceeds the preset interference threshold during the peak interference period, which is extracted from the spatiotemporal coupling database. The spatially overlapping conflict area is identified and marked as the buffer compartment segment.

[0039] By integrating peak path sections and buffer compartment sections, a dynamic corridor base spatial layout map is generated to guide subsequent engineering construction.

[0040] Specifically, this embodiment plans to form a dynamic corridor base with a non-uniform connectivity structure. The engineering purpose is to transform the spatiotemporal coupling database generated in the previous step into a constructable blueprint with clear spatial orientation and functional zoning. This process is achieved through three sequential engineering steps, transforming the abstract suitability index into specific corridor paths and functional zoning.

[0041] The first step is to identify and generate peak path segments. The core of this step is to use a minimum cumulative resistance model to simulate and calculate the most likely path chosen by the target mammal during migration in a specific season, within the geographic space represented by the spatiotemporally coupled database. In engineering terms, the system first uses the suitability index of the breeding or dispersal season from the spatiotemporally coupled database. It is converted into a raster cost surface through an inverse function. The formula is ,in This represents the passage resistance or cost of a geographic grid cell for a target mammal. The higher the value, the more difficult it is for the animal to cross that grid cell; The lower the value, the easier and more likely it is to cross; it is the direct input data for the minimum cumulative resistance model. This is a scaling constant or scaling factor, preset by the model builder or user based on experience or computational needs. For example, if The value is between [0,1], set It can The values ​​are distributed over a wider range. To ensure the stability of numerical calculations, a tiny positive number, such as 0.000001, is set to avoid a zero denominator. This transformation ensures that areas with higher suitability have lower access costs. Subsequently, the grid cost surface and preset source green space nodes are used as inputs to generate a cumulative cost surface using the minimum cumulative resistance algorithm. Then, using preset target green space nodes as targets, the optimal paths connecting different green space nodes are generated based on the cumulative cost surface through cost path analysis. Finally, the optimal paths with cumulative resistance values ​​below a preset migration feasibility threshold are selected and marked as peak path segments.

[0042] The second step involves identifying conflict zones and delineating buffer zones. The engineering objective of this step is to accurately locate high-risk points along the corridor path caused by human activity. In the geographic information platform, the system spatially overlays a layer composed of peak path segments generated in the previous step with a layer extracted from a spatiotemporally coupled database representing areas where the interference values ​​(light intensity or noise decibels) exceed a preset interference threshold during peak interference periods. Areas where the two layers overlap spatially are identified as conflict zones. Considering the diffusion of interference effects, the system extends a safe buffer distance outward from the boundary of the conflict zone. This distance is set according to the type of interference; for example, for light and noise interference on main traffic arteries, the buffer distance can be set to 20 to 50 meters. The peak path segment covered by this buffer range is ultimately marked as a buffer zone.

[0043] The third step is to integrate and output the spatial layout map of the dynamic corridor base. This step involves integrating the separate paths and functional sections into a unified guiding engineering drawing. The system loads the preset green space node spatial distribution data provided by the urban planning department as the base map, and then uses the marked peak path sections as connecting lines to physically link these isolated green space nodes. Finally, the designated buffer chamber sections are precisely overlaid on the corresponding positions of the peak path sections and highlighted with different legends. The final output, a vectorized composite map containing green space nodes, peak path sections, and buffer chamber sections, is the spatial layout map of the dynamic corridor base. It clearly defines the framework of the corridor and the key areas that need to be reinforced and shielded, directly guiding subsequent vegetation configuration and engineering construction.

[0044] S3. At the planting site of the dynamic corridor base, a functional modular vegetation system including trigger plant combinations and responder vegetation structures is configured, wherein the phenological period of the trigger plant combinations matches the key life activity period, and the responder vegetation structures are used to perform morphological densification growth in response to preset environmental disturbance threshold signals.

[0045] In a preferred embodiment, the configuration comprises a functional modular vegetation system including trigger plant combinations and responder vegetation structures, comprising: selecting a variety of native plants that match the peak foraging periods during the fruit ripening period and key life activity periods to form trigger plant combinations, and planting them in a point-like manner along the peak path sections.

[0046] Plants whose growth rate is regulated by nutrient stimuli were selected to form the responder vegetation structure, and were linearly planted along the boundary of the buffer chamber section near the disturbance source.

[0047] An automatic drip irrigation system is configured for the responder vegetation structure, and the automatic drip irrigation system is connected to a sensor for monitoring environmental disturbance parameters, so that when the monitored value exceeds the preset environmental disturbance threshold signal, the automatic drip irrigation system is triggered to apply nutrient stimulation.

[0048] Specifically, the modular vegetation system configured in this embodiment aims to physically integrate biological components with preset ecological functions with automated control hardware according to the spatial layout map of the dynamic corridor base, thereby constructing a living ecological interface with active induction and passive response capabilities within the corridor.

[0049] The first step of this project is the precise planting of trigger plant combinations. The project team will use a local native plant phenological database to select species whose fruit ripening periods highly overlap with the peak foraging periods of target mammals, such as paper mulberry, firethorn, and wild grape. These species will be combined and planted in clusters along the peak foraging routes, with each cluster spaced approximately 10 to 20 meters apart. This planting method aims to create biobeacons within the corridor, attracting animals to venture deeper, thus forming the trigger plant combinations, through seasonal food availability.

[0050] The second step of this project is to construct the responder vegetation structure. In this project, fast-growing climbing plants sensitive to water and fertilizer conditions, such as Virginia creeper, or clump-forming shrubs, such as specific varieties of rose, are selected as the main species for the responder vegetation structure. These plants will be planted in a high-density linear fashion, at 3 to 5 plants per linear meter, along one side of the buffer chamber section adjacent to the disturbance source, such as road or commercial area lights. The aim is to pre-deploy a biological substrate capable of forming a physical barrier through rapid proliferation.

[0051] The third step of this project is to deploy an automated environmental response system. Here, a zone-controlled automated drip irrigation system will be installed and cover all responder vegetation structures. This system includes at least two independent infusion lines, connected to a clean water source and a specific nutrient stimulation storage tank containing high-nitrogen liquid fertilizer, respectively. Simultaneously, in the buffer zone section, environmental sensors linked to the automated drip irrigation system controller will be installed, including sensors monitoring noise levels in decibels and light intensity. The system's control logic is programmed so that when the sensor readings exceed a preset environmental interference threshold for a period of time, such as more than 10 minutes (e.g., nighttime light intensity exceeding 15 lux or background noise exceeding 70 decibels), the controller will automatically open the solenoid valve connected to the specific nutrient stimulation storage tank, precisely delivering nutrients to the roots of the responder vegetation structures in the corresponding area. This stimulates accelerated growth and thickens the vegetation canopy without human intervention, achieving dynamic and adaptive shielding against interference sources.

[0052] In a further preferred embodiment, the triggering of the automatic drip irrigation system to apply nutrient stimulation includes: preset control logic, when the sensor detects that the ambient noise decibel value continuously exceeds the preset noise threshold, the automatic drip irrigation system applies a first nutrient formula, the first nutrient formula being a high-nitrogen, high-phosphorus water-soluble fertilizer for promoting the increase of branch and leaf density, so as to trigger the responder vegetation structure to promote the morphological densification growth of branch and leaf density.

[0053] When the sensor detects that the ambient light intensity value continuously exceeds the preset light pollution threshold, the automatic drip irrigation system applies a second nutrient formula, which is a plant growth regulator containing components that promote vertical growth, to trigger the responder vegetation structure to carry out morphological densification growth that promotes vertical growth rate.

[0054] Specifically, this embodiment employs differentiated regulation of the morphological density growth of the responder vegetation structure. The engineering objective is to precisely and efficiently shape the physical morphology of the vegetation barrier according to the specific type of interference source, thereby achieving optimal shielding against different types of interference. This process is implemented through two parallel automated control loops.

[0055] The first control loop targets noise interference shielding. In this loop, acoustic sensors deployed in the buffer compartment continuously monitor ambient noise levels in decibels. The noise threshold for these sensors is preset to an empirical value between 65 and 75 decibels, corresponding to the level at which persistent traffic noise might have a repulsive effect on most small to medium-sized mammals. When the noise level detected by the sensors consistently exceeds this threshold for a preset time, such as 15 minutes, the control system determines it as a valid noise interference event. At this point, the automatic drip irrigation system is triggered to deliver targeted nutrients. The system opens the valve connecting to the first nutrient formulation, a high-nitrogen, high-phosphorus water-soluble fertilizer whose nitrogen, phosphorus, and potassium ratios are pre-adjusted to preferentially promote plant vegetative growth, i.e., increased leaf density. By applying this formulation in pulses, it can significantly promote the budding and leaf area expansion of responder vegetation structures, such as clump-forming shrubs, thereby effectively increasing the acoustic absorption and scattering capacity of the vegetation barrier within several weeks.

[0056] The second control loop targets light pollution shielding. In this loop, a deployed photosensor operates at night, continuously monitoring ambient light intensity. The light pollution threshold is set between 10 and 20 lux, a range corresponding to the intensity at which artificial light sources significantly interfere with the navigation and circadian rhythms of nocturnal animals at night. When the sensor detects that the light intensity consistently exceeds this threshold, the system identifies it as a valid light pollution event. At this point, the automatic drip irrigation system is triggered, but only the valve connecting to the second nutrient formulation is opened. This formulation is a special preparation that regulates apical dominance in plants, potentially containing a low nitrogen-phosphorus ratio and trace amounts of plant growth regulators, such as gibberellins. Applying this formulation prioritizes promoting the elongation and vertical growth rate of responder vegetation structures, such as climbing plants, enabling them to rapidly climb pre-set trellises or adjacent trees, thereby creating an efficient vertical visual barrier between the interfering light source and the corridor. In this way, the system can adaptively construct a physically optimal noise or light-shading barrier through biological means.

[0057] S4. Deploy a sensor network at the path decision nodes of the dynamic corridor base and configure the sensor network to collect animal movement signals in real time and generate animal behavior stream data.

[0058] In a preferred embodiment, the deployment of a sensor network at path decision nodes of the dynamic corridor base and the configuration of the sensor network to collect animal movement signals and generate animal behavior stream data in real time includes: capturing raw sensing signals containing target signals generated by environmental noise and target mammal activity through infrared microwave composite sensors or vibration sensors deployed at the path decision nodes.

[0059] The original sensing signal is filtered and feature extracted to obtain feature parameters including movement direction, movement speed and dwell time.

[0060] The feature parameters are combined according to the time series to form animal behavior stream data that describes the movement patterns of animals at path decision nodes.

[0061] Specifically, in this embodiment, animal movement signals are collected in real time through a sensor network to generate animal behavior stream data. The engineering objective is to transform low-fidelity physical disturbance signals into a structured data stream that can characterize the animal's kinematic features, providing real-time input for subsequent behavior guidance decisions. This process is completed on a local edge computing node through three closely linked steps.

[0062] The first step is to capture the raw sensing signal. Infrared-microwave composite sensors or high-sensitivity vibration sensors are deployed in 2x2 or 3x3 arrays at path decision nodes, such as corridor forks or entrances to underground passages. These sensors passively monitor the background physical field. When a target mammal, such as an individual weighing 2 to 10 kg, enters the effective sensing range of the sensor array, its movement causes a slight change in the local geomagnetic field or transmits vibration waves with a specific frequency range to the ground. The sensor array converts these continuous physical field changes into analog voltages or digital signals; this signal is the raw sensing signal, which includes both environmental noise and the target signal.

[0063] The second step involves signal processing and feature parameter extraction. The system first applies digital signal processing algorithms to the received raw sensor signal. A bandpass filter with a preset cutoff frequency filters out high-frequency noise unrelated to animal movement, such as wind noise, as well as low-frequency background drift. Next, a signal envelope detection algorithm identifies the valid signal segment. Based on this, the animal's direction of movement is calculated by comparing the time difference between the signal peak arrival times at different sensors in the array; its speed is estimated by calculating the time difference between the animal passing two sensors with known spacing in the array; and the duration of stay is determined by the duration the signal amplitude exceeds a preset activation threshold. These three calculated quantitative indicators constitute the feature parameters.

[0064] The third step involves the structured encapsulation of animal behavior stream data. The system binds each set of discrete feature parameter data extracted in the previous step with the event timestamp and sensor node ID, forming a complete data record. As the animal continuously moves within the corridor, generating a series of such data records, the system concatenates these records in chronological order. This structured time-series data set, containing information on time, spatial location, direction of movement, speed, and dwell time, constitutes the final generated animal behavior stream data, dynamically describing the animal's movement patterns and decision-making behavior within the corridor, and is transmitted in real-time to the subsequent analysis and decision-making modules.

[0065] S5. Input animal behavior flow data to the edge computing node. The edge computing node generates environmental regulation instructions based on the behavior guidance strategy and controls the environmental regulation device set at the path decision node to execute the environmental regulation instructions in order to guide the movement path of the target mammal in real time.

[0066] In a preferred embodiment, the process of generating environmental regulation instructions and controlling the environmental regulation device to execute environmental regulation instructions includes: receiving animal behavior flow data and determining whether the animal behavior flow data exhibits wandering characteristics through pattern recognition.

[0067] If the determination is yes, then the environmental adjustment instruction corresponding to the loitering characteristic is retrieved from the instruction mapping table.

[0068] The environmental adjustment command is a control signal used to trigger the environmental adjustment device to release pheromones to simulate odors or to adjust the light emission mode of the low-light guide lamp.

[0069] Specifically, in this embodiment, edge computing nodes generate environmental adjustment instructions based on preset behavior guidance strategies. The engineering purpose is to automatically translate the real-time monitored animal hesitation behavior into specific, non-invasive guidance interventions to overcome the obstacles to the animal's passage at key decision points in the corridor.

[0070] This project begins with the real-time analysis of received animal behavior stream data by edge computing nodes. Within a preset time window, such as 1 to 3 minutes, the system performs pattern recognition on the data to determine the presence of loitering characteristics. These loitering characteristics are quantified in engineering as a combination of a series of behavioral parameters; for example, within the sensing area of ​​a specific path decision node, the animal's average moving speed is less than 0.2 meters per second, and its direction of movement changes more than three times within 30 seconds. Once these conditions are simultaneously met, the system confirms that a loitering event has been triggered.

[0071] After confirming the loitering characteristics, the system immediately generates intervention commands based on a pre-set command mapping table. This command mapping table is a query database stored locally on the edge computing nodes, which pre-associates loitering characteristics occurring at different sensor node locations with a set of optimal guidance actions. For example, if it is determined that an animal is loitering at the north exit of the corridor, the command mapping table will directly return an activation command for the environmental regulation device pointing to the north exit. The core of this step is to directly transform abstract behavioral judgments into environmental regulation commands with clear spatial orientation and action types.

[0072] The environmental control command is a string of digital codes containing all the parameters for the specific action to be performed. If the command is to trigger a directional odor diffuser, the code will specify the activation of a specific diffuser located 5 to 10 meters ahead of the target path, releasing a low dose of pre-stored pheromones that simulate similar safety signals or the scent of local dominant foods such as mulberry fruit, for a duration typically set to 30 to 60 seconds. If the command is to adjust a specific wavelength of low-light guide light, such as red or amber light, the code will control the guide light to emit a weak light with an illuminance of less than 1 lux, simulating the moonlight spectrum, and may point it towards the entrance of the safety passage in a slow pulse pattern to avoid creating a disturbing strong light source.

[0073] In a further preferred embodiment, after generating the environmental regulation command and controlling the environmental regulation device to execute the environmental regulation command, the method further includes: during the monitoring period after the execution of the environmental regulation command, collecting subsequent animal movement signals of the path decision node through a sensor network.

[0074] Based on subsequent animal movement signals, the effectiveness of environmental regulation instructions was evaluated by calculating the guidance success score, and this score was used as feedback data on the guidance effect.

[0075] Based on feedback data of the guidance effect, the preset weight values ​​associated with environmental adjustment instructions are adjusted through edge computing nodes to adaptively optimize the behavior guidance strategy.

[0076] Specifically, the adaptive optimization adjustment performed after executing the environmental adjustment command in this embodiment aims to establish a closed-loop feedback mechanism based on the actual guidance effect, so that the edge computing node can learn autonomously and iteratively optimize the effectiveness of its guidance intervention, thereby transforming the system from a static rule enforcer into a dynamic learner.

[0077] This project first initiates a preset monitoring period within the edge computing node, limited to 3 to 5 minutes after the execution of the environmental regulation command. During this period, the sensor network continuously operates, collecting and generating subsequent animal movement signals. Next, the system enters the evaluation phase, quantifying the effectiveness of the environmental regulation command by analyzing these subsequent animal movement signals. In practice, this evaluation is achieved by calculating a guidance success score. This is achieved by ensuring that the success score is positively correlated with the animal's correct movement direction and movement efficiency.

[0078] in, This refers to the feedback data on the guiding effect, which is a dimensionless score. It is a binary success factor; if subsequent animal movement signals indicate that the animal has successfully passed the path decision node and is moving in the predetermined direction, then... It is 1 if it is true, otherwise it is 0. This is a preset reference passage time constant, which represents the expected or standard passage time for the target species within the monitoring section, for example, set to 15 seconds based on corridor design and animal habits. The time taken for the animal to completely leave the monitoring area after receiving intervention is calculated from the timestamps in subsequent animal movement signals. When the guidance success score is higher than a preset optimization threshold, the weight value is increased; conversely, the weight value is decreased. This formula ensures that only successful and efficient guidance will receive a high score.

[0079] Finally, the system uses the guidance success score to adaptively optimize and adjust the parameters in the preset behavior guidance strategy. Inside the edge computing node, the behavior guidance strategy is implemented as a weight table, which associates each lingering feature with all available environmental adjustment instructions, assigning a weight value to each pair of associations. The system compares the calculated guidance success score with a preset optimization threshold. If R is higher than the threshold, it indicates that the environmental adjustment instruction was executed effectively, and the system finds the weight entry associated with that instruction in a preset weight table or similar key-value storage structure, and increases its value, for example, by 0.1 or proportionally by 5%. If R is not higher than the threshold, it indicates that the execution effect was poor or mediocre, and the system reduces the corresponding weight value. Through this simple reinforcement learning-based mechanism, the system will be more inclined to choose instructions with historically good guidance performance in future decisions, thereby achieving adaptive optimization and continuous improvement of the strategy.

[0080] In a further preferred embodiment, the method further includes: periodically integrating long-term animal behavior stream data, growth status data of the functional modular vegetation system, and environmental disturbance monitoring data to generate a system operation dataset.

[0081] By performing trend analysis on the system operation dataset, failure sections in the dynamic corridor base that have continuously lower utilization efficiency than the preset functional threshold or pressure sections whose interference level has continuously exceeded the preset pressure threshold can be identified.

[0082] Based on the results of trend analysis, vegetation configuration adjustment schemes or local spatial structure optimization schemes are generated for failure or stress sections to initiate adaptive management of the dynamic corridor base.

[0083] Specifically, the adaptive management closed loop in this embodiment aims to upgrade the entire ecological corridor system from a passively maintained facility to a life cycle management system with long-term self-diagnosis and optimization capabilities, ensuring that it maintains a high level of ecological efficiency in the ever-changing urban environment.

[0084] This project begins with a periodic data integration step. The system will automatically perform data archiving and integration on a quarterly or semi-annual basis. The system will summarize all long-term animal behavior flow data collected by the sensor network within this period and process it into statistical indicators such as passage frequency, average passage speed, and loitering event incidence for each corridor segment. Simultaneously, the system will acquire growth status data of the functional modular vegetation system through multispectral sensors mounted on UAVs, primarily represented by a distribution map of the Normalized Difference Vegetation Index (NDVI) to quantify the health and density of the vegetation. Furthermore, the environmental disturbance monitoring data, namely historical records from light and noise sensors, will also be integrated into daily average disturbance levels and extreme disturbance event frequencies for each segment. After spatial alignment and temporal synchronization on the geographic information system platform, these three types of data will be integrated into a comprehensive system operation dataset.

[0085] Next, the system performs trend analysis on the system's operational dataset to identify corridor segments experiencing functional degradation. This analysis mainly includes two aspects. First, identifying failed segments: the system compares the current period's passage frequency with historical data from the same period. If the passage frequency of a segment decreases by more than a preset threshold, such as 50%, or falls below the minimum passage volume required to maintain population communication for several consecutive periods, the segment is automatically marked as a failed segment. Second, identifying stress segments: the system analyzes the time series of environmental disturbance data. If the increase in the frequency of nighttime light or noise disturbance events in a segment compared to the previous period reaches a preset growth threshold, and the decrease in the segment's passage frequency reaches a preset response threshold, the segment is marked as a stress segment.

[0086] Finally, based on the results of trend analysis, the system automatically generates executable optimization plans. If a segment is marked as a failure segment, but the increase in its disturbance level does not reach the preset growth threshold used to identify stress segments, the system, after analyzing its vegetation growth data, may generate a vegetation configuration adjustment plan, suggesting replanting or replacing the trigger plant combination to enhance food attraction. If a segment is marked as both a failure segment and a stress segment, the system will prioritize generating a vegetation configuration adjustment plan to strengthen the shielding, such as suggesting manually increasing the nutrient supply to the responder vegetation structure. In extreme cases, if the analysis shows that the intensity of the newly added disturbance source exceeds the adjustment capacity of the vegetation barrier, the system will generate a local spatial structure optimization plan for the dynamic corridor base, such as suggesting planning an alternative path around the strong disturbance source in the GIS model, or suggesting the installation of physical sound barriers. These plans are pushed to management personnel in the form of reports, thereby connecting data monitoring, problem diagnosis, and optimization decision-making, formally initiating and completing a closed loop of adaptive management of the dynamic corridor base.

[0087] Example 2 Please see Figure 2 As shown, based on Embodiment 1, the second aspect of the present invention provides a system for constructing mammal ecological corridors in urban green space systems, comprising: a data acquisition and processing module, a corridor base planning module, a vegetation system configuration module, a behavior perception module, and a real-time guidance module.

[0088] The data acquisition and processing module acquires activity trace data of the target mammal, distribution data of urban disturbance sources, time data of key life activity periods of the target mammal, time data of peak periods of urban human activities, and natural rhythm time data, and performs spatiotemporal correlation on the acquired data to generate a spatiotemporal coupled database.

[0089] The corridor base planning module, based on a spatiotemporal coupling database, analyzes the spatiotemporal overlap between peak periods of key activities and peak periods of interference, and plans a dynamic corridor base that includes peak path sections and buffer compartment sections accordingly.

[0090] The vegetation system configuration module configures a functional modular vegetation system, including trigger plant combinations and responder vegetation structures, at the planting sites of the dynamic corridor base.

[0091] The behavior perception module deploys a sensor network at the path decision nodes of the dynamic corridor base and configures the sensor network to collect animal movement signals in real time and generate animal behavior flow data.

[0092] The real-time guidance module deploys edge computing nodes and environmental regulation devices, and presets behavior guidance strategies in the edge computing nodes, configuring them to receive animal behavior flow data, generate environmental regulation instructions according to the behavior guidance strategies, and control the environmental regulation devices to execute environmental regulation instructions, so as to achieve real-time guidance of the target mammal's movement path.

[0093] The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, and all such modifications and additions should fall within the protection scope of the present invention.

Claims

1. A method for constructing mammal ecological corridors in urban green space systems, characterized in that, include: S1. Acquire activity trace data of the target mammal, urban disturbance source distribution data, key life activity period time data of the target mammal, peak period time data of urban human activities, and natural rhythm time data, and perform spatiotemporal correlation on the acquired data to generate a spatiotemporal coupled database. S2. Based on the spatiotemporal coupling database, the spatiotemporal overlap relationship between the peak periods of key activities and the peak periods of interference is analyzed, and a dynamic corridor base including peak path sections and buffer compartment sections is planned accordingly. S3. At the planting sites of the dynamic corridor base, configure a functional modular vegetation system that includes trigger plant combinations and responder vegetation structures. S4. Deploy a sensor network at the path decision nodes of the dynamic corridor base and configure the sensor network to collect animal movement signals in real time and generate animal behavior flow data. S5. Deploy edge computing nodes and environmental regulation devices, and preset behavior guidance strategies in the edge computing nodes so that they are configured to receive animal behavior flow data, generate environmental regulation instructions according to the behavior guidance strategy, and control the environmental regulation devices to execute environmental regulation instructions, so as to achieve real-time guidance of the target mammal's movement path.

2. The method for constructing mammal ecological corridors in urban green space systems according to claim 1, characterized in that, The step of performing spatiotemporal correlation on the acquired data to generate a spatiotemporally coupled database includes: Infrared camera monitoring and trace investigation are used to obtain activity trace data containing geographic coordinates and timestamps; By using remote sensing mapping and ground monitoring, data on the distribution of urban interference sources, including light intensity and noise decibel values, can be obtained. Data on key life activity periods, peak periods of urban human activities, and natural rhythms were obtained from ecological databases, urban management databases, and meteorological and astronomical databases, respectively. The study area is divided into pre-defined geographical grid units; For each geographic grid cell, based on a pre-set evaluation model, and combined with key life activity period time data, peak period time data of urban human activities, and natural rhythm time data, the animal activity intensity index obtained from activity trace data analysis, the normalized light intensity value, and the noise decibel value are weighted and calculated to obtain the dynamic spatiotemporal suitability index characterizing the grid cell at a specific time point. The data, which includes the spatial coordinates, timestamps, and corresponding dynamic spatiotemporal suitability indices of each grid cell, are combined to generate a spatiotemporal coupled database.

3. The method for constructing mammal ecological corridors in urban green space systems according to claim 1, characterized in that, The plan forms a dynamic corridor base comprising peak path sections and buffer compartment sections, including: Based on a spatiotemporal coupled database, potential migration paths connecting different green space nodes are calculated and generated using a minimum cumulative resistance model, and these potential migration paths are marked as peak path segments. Spatial overlay analysis is performed on the peak path segment and the area where the interference value exceeds the preset interference threshold during the peak interference period, which is extracted from the spatiotemporal coupling database. The spatially overlapping conflict area is identified and marked as the buffer compartment segment. By integrating peak path sections and buffer compartment sections, a dynamic corridor base spatial layout map is generated to guide subsequent engineering construction.

4. The method for constructing mammal ecological corridors in urban green space systems according to claim 1, characterized in that, The configuration comprises a functional modular vegetation system consisting of trigger plant combinations and responder vegetation structures, including: Select a variety of native plants that match the peak foraging periods during the fruit ripening period and the critical life activity period to form a trigger plant combination, and plant them in spots along the peak path section. Plants whose growth rate is regulated by nutrient stimulation are selected to form a responder vegetation structure, and linear planting is carried out along the boundary of the buffer chamber section near the interference source. An automatic drip irrigation system is configured for the responder vegetation structure, and the automatic drip irrigation system is connected to a sensor for monitoring environmental disturbance parameters so that when the monitored value exceeds a preset environmental disturbance threshold signal, the automatic drip irrigation system is triggered to apply nutrient stimulation.

5. The method for constructing mammal ecological corridors in urban green space systems according to claim 4, characterized in that, The triggering of the automatic drip irrigation system to apply nutrient stimulation includes: The preset control logic is that when the sensor detects that the ambient noise decibel value continuously exceeds the preset noise threshold, the automatic drip irrigation system applies the first nutrient formula, which is a high-nitrogen and high-phosphorus water-soluble fertilizer used to promote the increase of branch and leaf density, so as to trigger the responder vegetation structure to promote the morphological densification growth of branch and leaf density. When the sensor detects that the ambient light intensity value continuously exceeds the preset light pollution threshold, the automatic drip irrigation system applies a second nutrient formula, which is a plant growth regulator containing components that promote vertical growth, to trigger the responder vegetation structure to carry out morphological densification growth that promotes vertical growth rate.

6. The method for constructing mammal ecological corridors in urban green space systems according to claim 1, characterized in that, The deployment of a sensor network at path decision nodes on the dynamic corridor base, and the configuration of the sensor network to collect animal movement signals and generate animal behavior stream data in real time, includes: The raw sensory signals, which include environmental noise and target mammal activity, are captured by infrared microwave composite sensors or vibration sensors deployed at path decision nodes. The original sensing signal is filtered and feature extracted to obtain feature parameters including movement direction, movement speed and dwell time; The feature parameters are combined according to the time series to form animal behavior stream data that describes the movement patterns of animals at path decision nodes.

7. The method for constructing mammal ecological corridors in urban green space systems according to claim 1, characterized in that, The generation of environmental adjustment commands and the control of the environmental adjustment device to execute the environmental adjustment commands include: Receive animal behavior stream data and determine whether the animal behavior stream data exhibits wandering characteristics through pattern recognition; If the determination is yes, then the environmental adjustment instruction corresponding to the loitering characteristic is retrieved from the instruction mapping table; The environmental adjustment command is a control signal used to trigger the environmental adjustment device to release pheromones to simulate odors or to adjust the light emission mode of the low-light guide lamp.

8. A method for constructing mammal ecological corridors in urban green space systems according to claim 7, characterized in that, After generating the environmental adjustment command and controlling the environmental adjustment device to execute the environmental adjustment command, the method further includes: During the monitoring period following the execution of environmental regulation instructions, subsequent animal movement signals are collected through a sensor network at the path decision nodes; Based on subsequent animal movement signals, the execution effect of environmental regulation instructions is evaluated by calculating the guidance success score, and this score is used as feedback data on the guidance effect. Based on feedback data of the guidance effect, the preset weight values ​​associated with environmental adjustment instructions are adjusted through edge computing nodes to adaptively optimize the behavior guidance strategy.

9. A method for constructing mammal ecological corridors in urban green space systems according to claim 1, characterized in that, Also includes: Regularly integrate long-term animal behavior data, growth status data of functional modular vegetation systems, and environmental disturbance monitoring data to generate a system operation dataset. Trend analysis of system operation datasets identifies failure sections or pressure sections in the dynamic corridor base where the utilization efficiency is consistently below the preset functional threshold or the interference level consistently exceeds the preset pressure threshold. Based on the results of trend analysis, vegetation configuration adjustment schemes or local spatial structure optimization schemes are generated for failure or stress sections to initiate adaptive management of the dynamic corridor base.

10. A system for constructing mammal ecological corridors in urban green spaces, characterized in that, include: The data acquisition and processing module acquires activity trace data of the target mammal, urban disturbance source distribution data, key life activity period time data of the target mammal, peak period time data of urban human activities, and natural rhythm time data, and performs spatiotemporal correlation on the acquired data to generate a spatiotemporal coupled database. The corridor base planning module, based on a spatiotemporal coupling database, analyzes the spatiotemporal overlap between peak periods of key activities and peak periods of interference, and plans a dynamic corridor base that includes peak path sections and buffer compartment sections accordingly. The vegetation system configuration module configures a functional modular vegetation system, including trigger plant combinations and responder vegetation structures, at the planting sites on the dynamic corridor base. The behavior perception module deploys a sensor network at the path decision nodes of the dynamic corridor base and configures the sensor network to collect animal movement signals in real time and generate animal behavior flow data. The real-time guidance module deploys edge computing nodes and environmental regulation devices, and presets behavior guidance strategies in the edge computing nodes. It is configured to receive animal behavior flow data, generate environmental regulation instructions according to the behavior guidance strategy, and control the environmental regulation devices to execute the environmental regulation instructions, so as to realize real-time guidance of the target mammal's movement path.

Images