Composite multi-technology collaborative harmless beast expelling system and method

Abstract

The invention relates to the technical field of wild animal protection and personal safety protection, and discloses a combined type multi-technology collaborative harmless beast expelling system and method, the system comprises a sensing layer, a control layer, an execution layer and an energy and communication layer; the sensing layer compositely detects a target through a thermal imaging camera, a millimeter wave radar and a vibration sensor; the control layer performs identification and threat evaluation on the target by using an AI algorithm; the execution layer comprises acoustic, optical, olfactory and physical barrier modules; the method comprises the following steps: cooperatively monitoring a protection area through multiple sensors; identifying beast species and evaluating threat levels; an optimal expelling technology combination scheme is intelligently generated based on the decision-making library; according to the threat level from low to high, multi-mode expelling means such as sound, light and smell are started or superposed in sequence, and hierarchical deterrence is formed; and the strategy is dynamically and adaptively adjusted according to the expelling effect. By means of the beast expelling device, efficient and anti-adaptive intelligent expelling without physical damage to beast is achieved.

Description

Technical Field

[0001] This invention relates to the field of wildlife protection and personal safety protection technology, and more specifically to a composite, multi-technology collaborative, harmless wild animal repelling system and method. Background Technology

[0002] With the improvement of the ecological environment and increased efforts in wildlife protection, the populations of predators (such as tigers, bears, wild boars, and wolves) have recovered somewhat. Their activity ranges are increasingly intersecting with human production and living areas, leading to frequent human-wildlife conflicts. Traditional methods of driving away predators have many drawbacks: Physical barriers (such as walls and electric fences): are costly, disrupt ecological connectivity, and have limited effectiveness against predators with climbing and digging abilities; Chemical repellency methods (such as spraying irritating agents): have a short-lived effect, are greatly affected by wind direction and rainfall, may pollute the environment, and pose potential hazards to humans and other non-target organisms; Sound-based deterrence methods (such as setting off firecrackers or using loudspeakers): The sound is monotonous, and wild animals can easily adapt (become accustomed) to it, thus rendering it ineffective; moreover, it causes serious noise pollution. Lethal methods (such as shootings and traps): contradict conservation principles, may raise legal and ethical issues, and disrupt the ecological balance.

[0003] Based on this, the present invention proposes a composite multi-technology collaborative non-harmful wild animal repelling system and method to solve the above problems. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a composite multi-technology collaborative non-damaging wild animal repelling system and method to solve the problems existing in the background art.

[0005] This invention provides the following technical solution: a composite, multi-technology collaborative, harmless wild animal deterrence system, comprising: The perception layer is used to detect intrusion targets and collect environmental data, including a multimodal sensor array and environmental sensors; The control layer, which is communicatively connected to the perception layer, is used to receive and process data, identify targets, assess threats, and generate deportation decisions. The execution layer, which is communicatively connected to the control layer, is used to execute expulsion commands and includes an acoustic expulsion module, an optical expulsion module, and an olfactory expulsion module. The energy and communication layer is used to power the system and provide remote communication capabilities.

[0006] As a further aspect of the present invention: the multimodal sensor array includes at least two of the following: a thermal imaging infrared camera, a millimeter-wave radar, and a vibration sensor; the environmental sensor includes a wind speed sensor and a wind direction sensor.

[0007] As a further aspect of the present invention: the control layer includes an edge computing gateway and a central controller; the central controller has an embedded AI decision algorithm module, which can optimize the selection of deportation strategies based on historical data and learning models.

[0008] As a further aspect of the present invention: the acoustic repellent module can generate a variety of sounds, including predator calls and strong sound waves; the optical repellent module includes an LED strobe light and / or a laser emitter that can project light spots onto the ground; the olfactory repellent module stores and can spray atomized natural animal repellents.

[0009] As a further aspect of the present invention: the execution layer further includes a physical barrier module, which includes an automatically rising and lowering barrier post or an inflatable pop-up model device.

[0010] As a further aspect of the present invention: a harmless method for repelling wild animals applied to any of the systems described in claims 1-5, characterized by comprising the following steps: S1: Continuously monitor the protected area through the sensing layer; S2: After a target is detected, the control layer performs target identification and threat level assessment; S3: Based on threat level and environmental data, intelligently generate one or more collaborative expulsion schemes; S4: According to the preset hierarchical strategy, control the execution layer to start one or more drive-off modules to carry out coordinated drive-off; S5: Dynamically adjust the expulsion plan based on the expulsion effect and record the learning process; S6: Send early warning information when necessary.

[0011] As a further aspect of the present invention: the hierarchical strategy in step S4 is as follows: Primary removal: The acoustic removal module is activated first; Intermediate removal: If the primary removal is ineffective, activate the optical removal module. Advanced Repellent: If the intermediate repellent is ineffective, then activate the olfactory repellent module and / or the physical barrier module.

[0012] As a further aspect of the present invention: when generating the repellency scheme in step S3, wind direction data needs to be considered to ensure that the olfactory repellent can drift toward the target direction.

[0013] As a further aspect of the present invention: in step S5, the dynamic adjustment includes: if the expulsion is effective, reducing the intensity; if it is ineffective, upgrading the scheme or switching the technology combination.

[0014] As a further aspect of the present invention: the system can distinguish between different types of ferocious beasts and call different preset expulsion strategy libraries for different types of ferocious beasts.

[0015] The technical effects and advantages of this invention are as follows: This invention employs non-harmful methods such as sound, light, and smell to drive away predators, avoiding physical harm to them and meeting ecological protection requirements. Simultaneously, the synergistic effect of multiple technologies significantly improves the effectiveness and reliability of the drive-away process, overcoming the problem of single-technology failure. Furthermore, the system possesses target recognition and intelligent decision-making capabilities, automatically selecting and adjusting drive-away strategies based on threat level and environmental conditions for precise response. Its rich combination of technologies and variable operating modes effectively prevent predators from adapting to fixed drive-away methods, ensuring long-term effectiveness. In addition, the invention combines radar and thermal imaging technologies to overcome the monitoring limitations of traditional optical equipment at night and in inclement weather, achieving uninterrupted protection. Attached Figure Description

[0016] The invention will now be further described with reference to the accompanying drawings.

[0017] Figure 1 This is a block diagram of the overall architecture of a composite multi-technology collaborative non-damaging wild animal repelling system of the present invention; Figure 2 This is a flowchart illustrating the steps of a composite, multi-technology collaborative, harmless method for driving away wild animals according to the present invention. Figure 3 This is a schematic diagram of the expulsion strategy logic of the present invention. Detailed Implementation

[0018] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. These embodiments are only used to explain the present invention and are not intended to limit the scope of protection of the present invention.

[0019] Example 1 This embodiment is used to provide a detailed description of the technical solution of the present invention: Please see Figure 1 As shown, a composite, multi-technology collaborative, harmless predator repelling system includes: 1. Perception layer: Multimodal sensor array: Thermal imaging infrared camera: Employs an uncooled vanadium oxide (VOx) or silicon nitride (a-Si) microbolometer focal plane array with a pixel count of no less than 384×288, thermal sensitivity (NETD) better than 50mK, operating wavelength of 8-14μm, and integrates an automatic defrosting coating and temperature control system to ensure normal operation in extremely cold environments. It also has a built-in preliminary AI recognition function, which can initially distinguish between humans, livestock, and wild animals (such as tigers, wild boars, bears, and wolves) based on thermal profile features. Millimeter-wave radar: It adopts the frequency modulated continuous wave (FMCW) system, operates at the 24GHz or 77GHz frequency band, and has multi-target tracking (MTT) capability. It can simultaneously detect and track the radial range, radial velocity, azimuth angle and Doppler micro-motion characteristics of multiple targets. The detection range is not less than 200 meters, the range accuracy is better than ±0.5 meters, and the velocity accuracy is better than ±0.1 m / s. The radar outputs raw point cloud data and tracked target data. Vibration sensor: It adopts a high-precision triaxial accelerometer based on MEMS technology, buried at a depth of 0.3-0.7 meters, with a sensitivity of not less than 2g. It can effectively collect and filter out low-frequency (1-50Hz) ground vibration signals generated by the footsteps of large animals (>50kg), and eliminate interference caused by wind, rain, small animals and vehicles through algorithms. Environmental sensors: Wind speed sensor: ultrasonic or mechanical type, measuring range 0-60m / s, accuracy ±0.3m / s; Wind direction sensor: Employs an absolute encoder with an accuracy of ±3°; Humidity sensor: Employs a capacitive polymer thin-film sensor with a measurement range of 0-100%RH and an accuracy of ±2%RH; Optional upgrades: precipitation sensor and atmospheric pressure sensor to provide the system with more comprehensive environmental data; 2. Control Layer: Edge computing gateway: Employ embedded processors with AI acceleration cores (such as ARM Cortex-A72+NPU). Run the multi-sensor data fusion (MSDF) algorithm: receive and time-synchronize radar point cloud data, infrared image frames and vibration signals, and use Kalman filter and nearest neighbor data association (NNDA) algorithms to associate and fuse target information from different sources to generate a unified target track with confidence, including fields such as position, speed, type and threat index; Perform a preliminary threat assessment: Based on the fused target trajectory, generate a primary threat level (e.g., low, medium, high) according to preset rules (e.g., target type is a predator, speed is greater than 1 m / s, direction is towards the core of the protected area).

[0020] Central controller: Employ industrial-grade programmable automation controllers (PACs) or high-performance embedded industrial control computers; The core of the AI ​​decision-making algorithm module is a hybrid decision engine that combines a rule-based expert system with a machine learning model; The strategy base is a multidimensional lookup table or knowledge graph whose decision dimensions include: target type (wild boar / bear / wolf / tiger), number of targets, target behavior pattern (loitering / charging / standing still), real-time distance, ambient lighting (day / night), wind speed and direction, and historical driving effects. For different combinations under each dimension, the policy library defines the corresponding expulsion response recipe, which precisely specifies: Technology combination: Which execution modules are enabled; Operating parameters: frequency of the acoustic module (e.g., low-frequency plosives of 500Hz-2kHz preferred for bears, and high-frequency whistles of 3k-8kHz preferred for wild boars), sound pressure level (105-120dB, adjustable with distance); strobe frequency of the optical module (5-15Hz), and scanning mode of the laser spot (random / zigzag); spray concentration and duration of the olfactory module; Startup sequence: Are the modules started simultaneously, or in a specific order (e.g., sound first, then light, then smell)? Area of ​​action: The servo gimbal of the control and execution module precisely guides the deflection method to the target location, avoiding unnecessary energy waste and area interference; 3. Execution layer: Acoustic dispersal module: It includes a high-power digital amplifier (≥200W), a waterproof horn speaker (sensitivity ≥110dB), and a digital audio processor (DSP). The DSP can generate or call up a variety of pre-stored acoustic sample libraries and synthesize strong sound waves of specific frequencies and envelopes in real time. Ultrasound (>20kHz) is used for interference, infrasound (<20Hz) is used to create discomfort, and audible sound waves are used to simulate predators and danger signals. All sound wave outputs are rigorously calculated and calibrated to ensure that their peak sound pressure (SPL) is below 140dB at the target location. This threshold is recognized as a safe upper limit that will not cause permanent hearing damage to the vast majority of animals. Optical drive-away module: High-power LED strobe light: It adopts a COBLED light source with an integrated heat dissipation substrate, with a luminous flux of >20,000 lumens and a color temperature of 5,000-6,000K. It is driven by a MOSFET circuit to achieve high-speed flashing with millisecond-level response. Laser emitter: Employs a Class-3R or lower safety laser (output power <5mW), with a wavelength of 532nm (green) or 650nm (red) for excellent visibility; equipped with a Galvo-Scanner system, which allows programmable control of the laser beam to scan dynamic and unpredictable geometric patterns (such as flashing halos or rapidly moving spots) at extremely high speeds within a preset area (e.g., 10 square meters of ground in front of a predator), greatly enhancing visual deterrence and preventing animals from habituating; Olfactory decoy module: It includes a liquid storage tank, a corrosion-resistant precision metering pump, an ultra-micro atomizing nozzle (producing droplets with a particle size of <30μm to ensure long-term suspension), and a compressed air system; The natural animal repellent is a food-grade capsaicin microcapsule water-soluble preparation or a mixture of organic acids extracted from humus. It has a strong stimulating effect on the olfactory mucosa of wild animals but is non-toxic and biodegradable. The spray is controlled by a controller that calculates the lead time and spray angle based on wind direction and speed data to achieve precise odor delivery. Physical barrier module (optional): Automatic lifting bollards: Driven by hydraulic or electric push rods, with a lifting speed ≥0.5m / s, and the surface is covered with flexible cushioning material; Inflatable deterrent model: Using a fast-release CO2 cylinder, it can inflate and pop up a model of a large predator (such as a human or wolf) silhouette within 0.5 seconds, triggering the predator's innate fear response by using the sudden change in shape. 4. Energy and Communication Layer: Solar power supply unit: Uses monocrystalline silicon solar panels with a conversion efficiency >22% and is equipped with an MPPT (maximum power point tracking) solar controller; Battery pack: Uses deep-cycle lithium iron phosphate (LiFePO4) batteries, with long cycle life, wide operating temperature range, and equipped with a battery management system (BMS). Communication module: Employs an industrial-grade 4G / 5G DTU, supporting VPN private network access. In areas without public network signal, it uses Low Power Wide Area Network (LPWAN) technology (such as LoRa) for local networking, and transmits critical alarm data and receives remote commands via satellite communication terminals (such as BeiDou RDSS).

[0021] Please see Figure 2 As shown, a harmless method for repelling wild animals applied to the above system includes the following steps: S1: Continuous monitoring: After the system is powered on, the millimeter-wave radar first enters a low-power continuous scanning mode, the vibration sensor continuously monitors, and the thermal imaging camera and other high-power sensors are in sleep mode. When the radar or vibration sensor detects a valid disturbance, it sends an interrupt signal to the edge computing gateway to wake up the corresponding sensor (such as a thermal imaging camera) as needed for collaborative verification, which greatly reduces the overall power consumption of the system. S2: Target Identification and Threat Assessment Edge computing gateways perform timestamp alignment and coordinate system unification for multi-source asynchronous data; The thermal imaging images were used to perform target detection and classification using a pre-trained convolutional neural network (CNN) model (such as YOLOv5 or ResNet), with the recognition confidence threshold set to 0.8. Threat level assessment uses a weighted scoring system; for example: target is bear (weight +5), number >1 (+3), speed >3m / s and facing the core area (+4), distance <50 meters (+3); total score 0-5 is low threat, 6-10 is medium threat, and 11+ is high threat. S3: Intelligent Decision-Making The central controller's AI decision-making module receives the fused target trajectory and threat level; The decision-making process begins with context matching, querying the strategy library for cases most similar to the current context (goal, environment, history); If no perfect match is found, a reinforcement learning (RL)-based decision model is activated to simulate the expected returns (Q-values) of different expulsion strategies, and the strategy combination with the highest Q-value is selected. The strategy returns are calculated based on a multi-objective optimization function, including expulsion success rate, energy consumption, and stress response to the target. S4: Hierarchical Cooperative Expulsion: Primary deterrence: Activate the "warning tone" category in the acoustic module, such as playing a wolf howl that lasts for 5 seconds, with the sound pressure level gradually increasing from 90dB to 105dB; at the same time, the radar continuously monitors changes in the target's speed. Intermediate deterrence: If the target speed does not decrease or the distance shrinks, the optical module is activated synchronously while the sound continues; the LED strobe light operates at a frequency of 10Hz, and the laser galvanometer projects a rapidly jumping light spot on the target's path; the sound and light are randomly and slightly asynchronous in time to enhance unpredictability; Advanced deterrence: If the target is still approaching (e.g., distance <20 meters), it is determined to be a high threat; the controller immediately calculates the current wind direction vector, controls the olfactory module nozzles upwind to point towards the target, and sprays a mist three times, each lasting 2 seconds, with a 1-second interval; at the same time, the sound and light intensity is increased to the maximum (sound 120dB, light fully illuminated); and the physical barrier module can be commanded to suddenly rise in the target's path to form a final deterrent; S5: Performance Evaluation and Adaptive Learning Effectiveness evaluation indicators include: whether the target stopped moving forward, whether the velocity vector turned 180° (escape), and whether it left the protected area; Establish a deportation case library to record the context data (S), deportation actions (A), and final results (R) for each event; periodically use this data to train the decision model offline, update the weights and priorities of each strategy in the strategy library, and realize experience sharing and continuous iterative optimization of system performance; S6: Early Warning and Reporting: The alert information adopts a multi-level push mechanism; low-level threats can only report logs locally; medium-level threats trigger APP push; high-level threats or failed deportation (target breaks through the final boundary) immediately trigger SMS, telephone and email alerts with on-site captured images, and provide the target's last known location and trajectory prediction.

[0022] Example 2 This embodiment uses the example of driving away wild boars, black bears, and other ferocious beasts to protect farmland and livestock pens on the edge of a village. 1. System Hardware Composition and Deployment Perception layer deployment: Thermal imaging infrared camera (optional models, such as the FLIRA series): Deployed on poles in protected areas (such as farmland boundaries), at a height of about 3-4 meters, with a field of view covering the main intrusion directions; its working wavelength is 8-14μm, with an effective detection distance of more than 100 meters, and it can provide clear imaging in complete darkness at night; its working mode is set to start recording and identification after being triggered by radar to avoid continuous power consumption. Millimeter-wave radar (using K-band frequency modulated continuous wave FMCW radar): mounted on the same pole as the infrared camera; the radar detection angle is 120°, the maximum detection range is 200 meters, and it has good resistance to rain, snow, and fog interference; its primary task is to detect moving objects and make a preliminary judgment on their speed and direction of movement, thereby triggering the infrared camera for precise identification. Vibration sensors (using high-sensitivity seismographs): buried approximately 0.5 meters underground every 20 meters along the boundary line that may be breached, forming an invisible vibration sensing barrier; used to detect ground vibrations caused by wild boars pawing the ground, bears walking, etc., as a supplement to radar and infrared systems, especially suitable for areas with dense vegetation that obstructs the view. Environmental sensors: A small weather station is installed on the main pole of the system, integrating wind speed, wind direction, and temperature and humidity sensors; the data is uploaded to the control layer in real time; Control layer deployment: Edge computing gateway: An industrial-grade gateway with a built-in high-performance AI processing unit (such as NVIDIA Jetson Xavier NX); deployed in a waterproof and lightning-proof enclosure on site; running lightweight target detection algorithms (such as an improved model based on YOLOv5s) on it. The model has been pre-trained and optimized with thousands of infrared images of animals such as wild boars, black bears, and deer, and can achieve a recognition accuracy of ≥95%. Central controller: adopts industrial-grade PLC or embedded industrial computer, and is connected to the edge computing gateway via Ethernet; Execution layer deployment: Acoustic deterrence module: It uses a high-power (100W) outdoor waterproof horn speaker, which is placed at 50-meter intervals; its built-in memory card contains various audio sounds such as wolf howls, tiger roars, shotgun sounds, and electric shock explosions; it also integrates a set of strong sound wave generators, which can directionally emit ultrasonic waves (18-25kHz) or low-frequency sound waves (100-500Hz) with a sound pressure level of 105-110dB, and the action time is controlled within a short pulse (1-3 seconds) to ensure deterrence while avoiding harm; Optical deterrence module: High-brightness LED strobe light (brightness > 200 million candela) is selected and installed on the same pole as the speaker; the laser emitter (Class 1 or Class 2 safety laser, power < 1mW) is installed separately and its angle is adjusted so that it can project a fast-moving, flashing green or red light spot on the ground 5-10 meters in front of the target, creating visual interference and fright. Olfactory repellent module: Stainless steel storage tanks and corrosion-resistant spray nozzles are installed at key locations (such as farmland entrances); the animal repellent is a mixture of natural capsaicin extract and water; the spray system is controlled by a solenoid valve, with an intermittent short-duration spray mode (e.g., 3 seconds on, 10 seconds off), and the upwind nozzle is selected based on wind direction data to ensure that the droplets can drift towards the predators; Physical barrier module (optional in this embodiment): A set of automatically rising and lowering barrier posts are deployed on the path leading to the livestock shed; normally the posts are buried underground, but upon receiving an advanced threat command, the posts quickly rise to a height of 0.8 meters above the ground within 2 seconds, effectively blocking collisions without harming the animals; Energy and Communication Layer: The entire system is powered by a 400W solar panel paired with two 200Ah deep-cycle batteries, which can ensure normal operation for 7 days under continuous rainy weather. It uses a 4G DTU communication module (with a spare SIM card slot) for remote data transmission; in areas with extremely weak network signals, it can be replaced with a satellite communication module.

[0023] Combination Figure 2 The complete workflow of this system is as follows: Monitoring and triggering: At night, the millimeter-wave radar detected a medium-sized target moving rapidly toward the farmland at a distance of 150 meters; the radar immediately sent a signal to the edge computing gateway; Identification and Assessment: The gateway wakes up the thermal imaging camera and focuses on the area to acquire infrared images; the AI ​​image recognition model determines within 0.5 seconds that the target is a "wild boar", the number is "1", the direction of movement is "directly towards the farmland", and the speed is "approximately 7m / s"; based on its behavior, the central controller assesses its threat level as "high"; Intelligent decision-making: The central controller reads meteorological data: the current wind is westerly, with a wind speed of 2 m / s; after querying the strategy library, it decides the optimal solution as follows: activate the acoustic module (play the hunting howl of wolves), activate the optical module (LED strobe + laser projection onto the ground in front of the wild boar), and activate the olfactory spray device located upwind to the west of the wild boar; Coordinated expulsion execution: First, upon hearing the sudden wolf howl (approximately 95 dB), the wild boar slowed down and hesitated. Almost simultaneously, intense, flashing white light and rapidly jumping green laser spots appeared in front of its eyes and along its path, severely disrupting and frightening its visual system. Because of the westerly wind, the spray nozzles located upwind of the wild boar sprayed capsaicin droplets with a pungent odor. The droplets were carried by the wind to the location of the wild boar, stimulating its sensitive sense of smell. The combined negative stimuli of sound, light, and smell acted on the wild boar within 2-3 seconds, which it could not adapt to. As a result, the wild boar gave up and turned to flee. Effectiveness evaluation and learning: Infrared cameras and radar continuously tracked and confirmed that the wild boar had escaped to a safe distance of more than 100 meters; the central controller recorded the event: "For a single wild boar, under (certain conditions), the 'sound + light + smell' combination strategy is effective"; this successful experience will be marked for priority recommendation in similar future scenarios; System Reset: All drive-off modules automatically stop working, the system returns to low-power monitoring state, and waits for the next trigger.

[0024] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A composite, multi-technology collaborative, harmless wild animal deterrence system, characterized in that, include: The perception layer is used to detect intrusion targets and collect environmental data, including a multimodal sensor array and environmental sensors; The control layer, which is communicatively connected to the perception layer, is used to receive and process data, identify targets, assess threats, and generate deportation decisions. The execution layer, which is communicatively connected to the control layer, is used to execute expulsion commands and includes an acoustic expulsion module, an optical expulsion module, and an olfactory expulsion module. The energy and communication layer is used to power the system and provide remote communication capabilities.

2. The composite multi-technology collaborative non-damaging wild animal repelling system according to claim 1, characterized in that: The multimodal sensor array includes at least two of the following: a thermal imaging infrared camera, a millimeter-wave radar, and a vibration sensor; the environmental sensors include a wind speed sensor and a wind direction sensor.

3. The composite multi-technology collaborative non-harmful wild animal repelling system according to claim 1, characterized in that: The control layer includes an edge computing gateway and a central controller; the central controller has an embedded AI decision-making algorithm module, which can optimize the selection of deportation strategies based on historical data and learning models.

4. The composite multi-technology collaborative non-harmful wild animal repelling system according to claim 1, characterized in that: The acoustic repellent module can generate a variety of sounds, including predator calls and strong sound waves; the optical repellent module includes an LED strobe light and / or a laser emitter that can project light spots onto the ground; the olfactory repellent module stores and can spray atomized natural animal repellents.

5. The composite multi-technology collaborative non-damaging wild animal repelling system according to claim 1, characterized in that: The execution layer also includes a physical barrier module, which includes automatically rising and falling barrier posts or inflatable pop-up model devices.

6. A harmless method for repelling wild animals applied to any of the systems described in claims 1-5, characterized in that, Includes the following steps: S1: Continuously monitor the protected area through the sensing layer; S2: After a target is detected, the control layer performs target identification and threat level assessment; S3: Based on threat level and environmental data, intelligently generate one or more collaborative expulsion schemes; S4: According to the preset hierarchical strategy, control the execution layer to start one or more drive-off modules to carry out coordinated drive-off; S5: Dynamically adjust the expulsion plan based on the expulsion effect and record the learning process; S6: Send early warning information when necessary.

7. The method for repelling wild animals without causing harm according to claim 6, characterized in that: The hierarchical strategy described in step S4 is as follows: Primary removal: The acoustic removal module is activated first; Intermediate removal: If the primary removal is ineffective, activate the optical removal module. Advanced Repellent: If the intermediate repellent is ineffective, then activate the olfactory repellent module and / or the physical barrier module.

8. The method for driving away wild animals without causing harm according to claim 6, characterized in that: When generating the repellent scheme in step S3, wind direction data must be taken into account to ensure that the olfactory repellent can drift towards the target direction.

9. The method for repelling wild animals without causing harm according to claim 6, characterized in that: In step S5, the dynamic adjustment includes: if the expulsion is effective, reducing the intensity; if it is ineffective, upgrading the scheme or switching the technology combination.

10. The method for repelling wild animals without causing harm according to claim 6, characterized in that: The system can distinguish between different types of wild animals and call different preset expulsion strategy libraries for different types of wild animals.

Images