Intelligent breeding system and method for full-barn feeding down producing goats

Abstract

The invention relates to the technical field of livestock breeding intellectualization, and discloses an intelligent breeding system and method for full-barn feeding cashmere goats. The method comprises the following steps: collecting position coordinates and attitude information of down producing goats in a breeding house, and collecting environmental parameters of a plurality of position points; identifying a sheep flock subset with a local aggregation situation according to the position coordinates, and synchronously obtaining environmental parameters associated with the geographic position of the subset; deducing an expected behavior evolution trend of the sheep flock subset under the action of the environmental parameters in combination with the attitude information and the associated environmental parameters; judging whether the distribution of the whole cashmere goats needs to be subjected to global intervention or not according to the trend; when it is judged that intervention is needed, the movement guiding priority of each cashmere goat is calculated; and generating a differentiated path guide strategy for each goat based on the priority and executing the strategy to adjust the distribution state of the down producing goats in the shed. According to the invention, active prediction and personalized accurate guidance of sheep flock behaviors are realized, and the intelligent level of breeding management and animal welfare are improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent livestock farming technology, specifically to an intelligent farming system and method for fully indoor cashmere goats. Background Technology

[0002] In a fully intensive cashmere goat farming model, abnormal flocking is often an early sign of environmental discomfort, potential disease, or stress. Existing farming management techniques typically rely on regular manual inspections or simple video monitoring to detect flock gatherings. The drawbacks of this approach are its reliance on human experience, delayed detection, and difficulty in directly correlating observed flock behavior with specific, real-time local environmental data. Environmental monitoring systems and behavioral observation systems often operate independently, making it impossible for managers to accurately identify which specific environmental parameters in which location might be triggering the behavior at the initial stage of flock gathering. This results in a lack of precise data support for intervention decisions, leading to missed opportunities for optimal prevention.

[0003] When undesirable aggregation of sheep is observed, conventional intervention methods often involve uniform herding or using fixed audio-visual equipment for a full-scale warning. These methods lack specificity and are prone to causing unnecessary disturbance to all sheep, potentially triggering greater stress responses. Furthermore, because they fail to differentiate between individual sheep's states during aggregation, they cannot efficiently and orderly guide the flock to disperse. The intervention process is crude and its effects are inconsistent, potentially failing to fundamentally eliminate the environmental factors causing aggregation and even exacerbating overcrowding in certain areas. Livestock management requires an intelligent technology that can predict flock behavior trends in advance and implement precise, flexible guidance based on individual differences. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent breeding system and method for fully indoor cashmere goats, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides an intelligent method for raising cashmere goats entirely in intensive housing, the method comprising: The location coordinates and posture information of cashmere goats were collected inside the breeding shed, and environmental parameters were collected at multiple locations within the breeding shed. Based on the collected location coordinates, a subset of sheep flocks exhibiting local clustering patterns was identified; For the identified subset of sheep flocks exhibiting local clustering patterns, environmental parameters associated with the geographical location of the sheep flock subset are acquired simultaneously. By combining the posture information of a subset of sheep with synchronously acquired associated environmental parameters, the expected behavioral evolution trend of the subset of sheep under the influence of the associated environmental parameters is deduced. Based on the expected behavioral evolution trend, determine whether a global intervention is needed to control the distribution of all cashmere goats in the breeding shed; When it is determined that a global intervention is required, the movement guidance priority for each cashmere goat in the breeding shed is calculated. Based on movement guidance priority, generate differentiated path guidance strategies for each cashmere goat; Implement the differentiated path guidance strategy to adjust the distribution of cashmere goats in the barn.

[0006] Preferably, the location coordinates and posture information of cashmere goats are collected inside the goat pen, and environmental parameters at multiple locations within the pen are also collected, specifically including: Images are periodically captured by visual sensors deployed on the top of the breeding shed, and the real-time position coordinates of each cashmere goat, as well as the posture information such as head orientation and limb extension, are extracted from the images. Through a sensor network distributed on the ground of the breeding shed and at different heights, environmental parameters such as temperature, humidity, light intensity and concentration of harmful gases are collected synchronously at corresponding locations. Establish a timestamp alignment mechanism to ensure that the location coordinates and posture information of each cashmere goat collected have the same time reference as the environmental parameters of each location point.

[0007] Preferably, based on the collected location coordinates, a subset of sheep flocks exhibiting local clustering patterns is identified, specifically including: Set a distance determination threshold; For each cashmere goat in the barn, calculate the Euclidean distance between its location coordinates and the location coordinates of all other cashmere goats in the barn; Filter out all other cashmere goats whose Euclidean distance from the current cashmere goat is less than the distance determination threshold; If the number of selected cashmere goats exceeds the set threshold, the current cashmere goat and all other selected cashmere goats will be defined together as a subset of the flock with a local clustering pattern. Traverse all cashmere goats in the breeding shed to identify all subsets of the flock that exhibit localized clustering patterns.

[0008] Preferably, for the identified subset of sheep flocks exhibiting local clustering patterns, environmental parameters associated with the geographical location of the sheep flock subset are simultaneously acquired, specifically: Determine the geographical coverage of the identified sheep flock subset; Within the geographical coverage area, locate the deployment locations of all sensor nodes; The system acquires environmental parameters such as temperature, humidity, illuminance, and concentration of harmful gases collected by the sensor nodes after timestamp alignment. The acquired set of environmental parameters is labeled as the geographical location-associated environmental parameters of the sheep flock subset.

[0009] Preferably, by combining the posture information of a subset of sheep with synchronously acquired associated environmental parameters, the expected behavioral evolution trend of the subset of sheep under the influence of the associated environmental parameters is deduced, specifically including: Analyze the rate of change in head orientation consistency and limb extension of cashmere goats within a subset of the flock; The rate of change of head orientation consistency and limb extension is correlated and mapped with the temperature change gradient, humidity value, light intensity and harmful gas concentration in the geographical location-related environmental parameters. By querying a pre-established behavior-environment mapping rule base, the expected behavior category of a subset of sheep under the current posture and environmental parameter combination is output; the expected behavior category includes maintaining aggregation, slow diffusion, rapid collision, or migration to a specific area. The expected behavior categories and their possible evolution paths are defined as the expected behavior evolution trends of the sheep subset.

[0010] Preferably, based on the expected behavioral evolution trend, it is determined whether a global intervention is needed to control the distribution of all cashmere goats in the breeding shed, specifically as follows: When the expected behavioral trend of any subset of sheep flocks is rapid collision, or when the expected behavioral trend of more than half of the sheep flocks is migration to the same specific area, it is determined that a global intervention is needed on the distribution of all cashmere goats in the breeding shed. Otherwise, it is determined that no global intervention is needed, and only local observation of individual sheep subsets is required.

[0011] Preferably, when it is determined that a global intervention is needed, the movement guidance priority for each cashmere goat in the barn is calculated, and the specific process is as follows: For each cashmere goat, identify all the subsets of the flock to which it belongs; Obtain the expected behavior category for each sheep subset to which it belongs; Different urgency coefficients are assigned to the expected behavior categories of rapid collision, migration to a specific area, slow spread, and maintaining aggregation. Calculate the movement guidance priority value of the cashmere goat, which is equal to the sum of the urgency coefficients of all the flock subsets to which it belongs, and then multiply it by the average movement speed of the cashmere goat in the most recent time period. All cashmere goats are sorted in descending order based on their movement guidance priority value; the higher the value, the higher the movement guidance priority.

[0012] Preferably, based on movement guidance priority, a differentiated path guidance strategy is generated for each cashmere goat, specifically including: Plan a straight or near-straight path to the target area for the cashmere goat subset with the highest priority for movement guidance, where the target area is an open area with low current aggregation. Plan segmented, progressive paths for a subset of cashmere goats with medium priority for movement guidance, and set intermediate transition points in the paths; Plan a fine-tuned path for the cashmere goat subset with the lowest priority for movement, so that it makes a small detour around its current position; Each planned path must ensure that it does not conflict with the fixed facilities in the breeding shed or the pre-set paths of other cashmere goats at the same time and space point.

[0013] Preferably, the differentiated path guidance strategy is implemented to adjust the distribution of cashmere goats within the breeding shed, specifically including: The planned straight or near-straight paths, segmented progressive paths, and fine-tuning paths are converted into specific control command sequences for adjustable feeders, controllable lights, directional sound devices, and automatic access control in the breeding sheds. Control of the adjustable feeder, controllable lights, directional sound devices, and automatic access control is initiated in descending order of movement guidance priority with time delays. During execution, the deviation between the actual location of the cashmere goats and the planned path is continuously monitored. If the deviation exceeds the tolerance, a real-time correction process for the cashmere goat path guidance strategy is triggered, and the subsequent control command sequence is updated.

[0014] Preferably, when the processor executes the computer program, it implements the steps of the intelligent breeding method for fully house-fed cashmere goats as described in any of the above-mentioned methods.

[0015] Compared with the prior art, the beneficial effects of the present invention are: By integrating and deploying a network of position sensors, attitude sensors, and environmental sensors, multi-dimensional data is collected and fused in real time. The algorithm dynamically identifies locally clustered subsets of the emerging sheep flock based on their location coordinates and simultaneously locks down the real-time environmental parameters of the micro-region containing these subsets. Combining the sheep's posture information with associated environmental data, the algorithm uses a built-in behavioral model for real-time analysis and calculation to deduce the short-term behavioral evolution trend of these sheep flock subsets under specific environmental conditions. This enables the management system to identify potential risk situations before the flock exhibits obvious overcrowding, stress, or health problems, shifting management actions from reactive response to proactive early warning.

[0016] After the system determines that global intervention is necessary, the algorithm calculates a differentiated movement guidance priority based on the real-time status of each sheep and the overall layout. Based on this priority, the system generates and executes differentiated path guidance strategies, applying guidance stimuli of varying intensity, direction, and timing to sheep of different priorities through controllable guidance devices. High-priority sheep receive more direct and rapid evacuation guidance, while low-priority sheep receive gentler guidance signals, thus creating an orderly and progressive spatial distribution adjustment within the flock. This personalized guidance method significantly reduces flock-wide stress caused by uniform herding, making the adjustment of the flock's distribution more efficient and stable, improving animal welfare and management efficiency. Attached Figure Description

[0017] Figure 1 This is a schematic diagram illustrating the working principle of the intelligent intensive cashmere goat farming method described in this invention. Figure 2 A flowchart for collecting position coordinates, attitude information, and environmental parameters; Figure 3 A flowchart for identifying sheep subsets exhibiting localized clustering patterns; Figure 4 A grouping comparison chart of different priority path planning parameters for cashmere goats; Figure 5 A bar chart showing the correlation between the priority of movement guidance for cashmere goats and the expected behavior category. Detailed Implementation

[0018] 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.

[0019] Please see Figure 1This invention provides an intelligent method for raising cashmere goats in a fully enclosed environment. The method includes: First, simultaneously collecting the location coordinates and posture information of all individual cashmere goats inside the goat pen, while also collecting environmental parameters from multiple discrete locations within the pen. Based on the collected location coordinate data, the system uses an algorithm to identify spatially clustered subsets of goats. For each identified clustered subset, the system simultaneously retrieves environmental parameter data closely related to the geographical area of ​​that subset. Subsequently, the system comprehensively analyzes the posture information of individuals within the subset and the simultaneously acquired associated environmental parameters, and uses a built-in model or rule base to deduce the expected behavioral evolution trend of the subset under the current environmental parameters. Based on a comprehensive evaluation of the expected behavioral evolution trends of all identified subsets, the system determines whether the overall distribution of cashmere goats in the pen requires global intervention. If global intervention is required, the system calculates a movement guidance priority value for each cashmere goat in the pen. Based on the calculated priority, the system generates differentiated movement path guidance strategies for cashmere goats with different priorities. Ultimately, the system drives the relevant controllable equipment in the breeding shed to execute these differentiated path guidance strategies, thereby proactively adjusting and optimizing the distribution of cashmere goats throughout the breeding shed.

[0020] In one embodiment of the present invention, see [reference] Figure 2 The study collected the location coordinates and posture information of cashmere goats inside the goat shed, and also gathered environmental parameters from multiple locations within the shed. The specific implementation process was as follows: Visual sensors deployed on the roof of the shed periodically captured panoramic images of the shed. Image recognition and analysis algorithms were used to extract the real-time location coordinates of each cashmere goat and its posture information, including head orientation and limb extension, from consecutive image frames. A sensor network distributed across the floor and at different heights of the shed simultaneously collected environmental parameters such as temperature, humidity, illuminance, and harmful gas concentrations at each sensor location. To ensure spatiotemporal consistency of the data, a timestamp alignment mechanism was established, assigning a unified timestamp to each frame of image data and each batch of sensor data. This ensured that the collected location coordinates and posture information of each cashmere goat, along with the environmental parameters at each location, shared the same time reference.

[0021] In practice, three high-definition panoramic vision sensors are evenly deployed along the centerline of the roof of the cashmere shed. The sensors are installed at a height of 3.5 meters, each with a horizontal field of view of 120 degrees, covering the entire ground area of ​​the shed. The sensors periodically capture digital images of the shed's interior at a fixed frequency of 10 frames per second and upload them to the central processing unit. The central processing unit runs an image recognition algorithm based on a pre-trained convolutional neural network model. This model detects and segments the outline of each cashmere goat from each frame, then analyzes its real-time position coordinates, with the origin of the three-dimensional coordinate system at one corner of the shed, in meters. Simultaneously, the image recognition algorithm calculates the head orientation angle by analyzing the angle between the cashmere goat's head outline and the midline of its torso, and quantifies limb extension by calculating the ratio of the cashmere goat's outline area to the area of ​​its smallest bounding rectangle. The head orientation angle and limb extension together constitute the posture information.

[0022] In some embodiments, a distributed sensor network is also constructed within the breeding shed to collect environmental parameters. The sensor network includes 20 temperature and humidity sensing nodes deployed on the floor of the breeding shed, arranged in a grid with a 10-meter spacing between adjacent nodes. The sensor network also includes 15 environmental monitoring nodes deployed at different heights, with 8 nodes deployed at a height of 1 meter above the ground and 7 nodes deployed at a height of 2.5 meters above the ground. Each environmental monitoring node integrates a light intensity sensor and a harmful gas concentration sensor, specifically ammonia and hydrogen sulfide concentration sensors. All temperature and humidity sensing nodes and environmental monitoring nodes are connected to a central processing unit via a wired network and synchronously collect temperature, humidity, light intensity, and ammonia and hydrogen sulfide concentration values ​​at their respective locations at a sampling frequency of 10 times per second, uploading the collected environmental parameter data packets in real time.

[0023] Optionally, to ensure the consistency of cashmere goat posture information and environmental parameters in the time dimension, the system establishes a strict timestamp alignment mechanism. The central processing unit has a built-in high-precision clock and maintains clock synchronization with all visual sensors and sensor network nodes through a network time protocol. In specific implementation, each frame of image data captured by the visual sensors is marked with a timestamp accurate to the millisecond level, synchronized by the central processing unit, when it is transmitted; similarly, each environmental parameter data packet generated by the temperature and humidity sensing node or environmental monitoring node is also marked with a timestamp of the same precision and time source when it is transmitted. The data receiving module of the central processing unit sorts and aligns all incoming position coordinate data, posture information data, and environmental parameter data according to the timestamps, ensuring that in subsequent processing, the position coordinates, posture information, and environmental parameters of each cashmere goat used in any calculation cycle point to the same millisecond-level time reference point, and the time deviation is controlled within 10 milliseconds.

[0024] It is understandable that the above data acquisition process generates massive amounts of time-series data. To quantify the accuracy of data alignment, a time synchronization error coefficient is defined. The formula for calculating the time synchronization error coefficient is: Where: symbol Represents the time synchronization error coefficient, with the sign... Represents the total number of data packets received within a processing cycle, symbol Representing the The timestamp values ​​recorded by the vision system in each data packet, symbol Represents the absolute value operator, symbol Representing the The data packet contains the timestamp value recorded by the sensor network. The central processing unit calculates the time synchronization error coefficient in real time. and time synchronization error coefficient The data was maintained at a level below 10 milliseconds. Through periodic image capture and analysis by visual sensors, multi-point synchronous acquisition by sensor networks, and data alignment mechanism based on precise timestamps, the system obtained the location coordinates, head orientation, and limb extension information of each cashmere goat in the breeding shed with a unified time reference, as well as the temperature, humidity, illuminance, and harmful gas concentration information of multiple discrete locations in the breeding shed with a unified time reference.

[0025] In one embodiment of the present invention, see [reference] Figure 3Based on the collected location coordinates, subsets of sheep flocks exhibiting localized clustering patterns are identified. The specific implementation process involves: setting a distance threshold; calculating the Euclidean distance between each cashmere goat in the pen and the location coordinates of all other cashmere goats in the pen; filtering out all other cashmere goats whose Euclidean distance to the current cashmere goat is less than the distance threshold; if the number of filtered cashmere goats exceeds a set threshold, then the current cashmere goat and all other filtered cashmere goats are collectively defined as a subset of sheep flocks exhibiting localized clustering patterns. This process is repeated for all cashmere goats in the pen to identify all subsets of sheep flocks exhibiting localized clustering patterns.

[0026] For the identified subset of sheep exhibiting localized clustering, environmental parameters associated with the geographical location of the sheep subset are simultaneously acquired. Specifically, this involves: determining the geographical coverage of the identified sheep subset, typically the outer envelope formed by the coordinates of all individuals within the subset; locating the deployment locations of all sensor nodes within this geographical coverage; acquiring the environmental parameters (temperature, humidity, illuminance, and harmful gas concentration) collected by these sensor nodes after timestamp alignment; and labeling the acquired set of environmental parameters as the geographical location-associated environmental parameters for the sheep subset.

[0027] In practice, the system sets a fixed distance threshold of 3 meters. For each cashmere goat in the pen, the system calculates the Euclidean distance between its location coordinates and the location coordinates of all other cashmere goats in the pen. This calculation is performed by traversing the location coordinate matrix. The system filters out all other cashmere goats whose Euclidean distance to the current target cashmere goat is less than 3 meters and records their identifiers. If the number of filtered cashmere goats exceeds the set threshold of 5, the common membership relationship between the current target cashmere goat and all other filtered cashmere goats is defined as a candidate cluster, and this candidate cluster is marked as a subset of the flock exhibiting a localized clustering pattern. The system traverses the location coordinates of all cashmere goats in the pen, repeating the above calculation, filtering, and judgment logic to identify all subsets of the flock exhibiting a localized clustering pattern. During the traversal, cashmere goats already included in a certain flock subset can still participate in subsequent judgments as members of other flock subsets.

[0028] In some embodiments, determining the geographic coverage of the identified sheep flock subset is a preliminary step in obtaining associated environmental parameters. The geographic coverage is determined by calculating the extreme values ​​of the location coordinates of all members within the sheep flock subset in each dimension. Specifically, for a sheep flock subset containing k cashmere goats, the set of location coordinates of all its members {(x1,y1,z1),(x2,y2,z2),...,(x...} is extracted. k ,y k ,z kThe geographical coverage of this subset of sheep is defined as min(x1...x2) / x3. k ) and max(x1...x k ) is the x-axis boundary, with min(y1...y k ) and max(y1...y k ) is the boundary of the y-axis, with min(z1...z k ) and max(z1...z k The system defines a three-dimensional cuboid region bounded by the z-axis. Within this geographically covered region, the system searches for the physical address codes of all sensor nodes located within it. These sensor nodes include temperature and humidity sensors and environmental monitoring nodes. The system retrieves temperature, humidity, illuminance, and hazardous gas concentration values ​​collected by these sensor nodes at a time reference point that matches the current sheep flock subset identification time after timestamp alignment from the central processing unit's cache database. The set of temperature, humidity, illuminance, and hazardous gas concentration values ​​acquired from multiple sensor nodes is then labeled as the geographical location-associated environmental parameters for that sheep flock subset.

[0029] Optionally, the distance determination threshold and quantity threshold can be dynamically configured based on the actual area of ​​the breeding shed and the stocking density. In one configuration, the distance determination threshold... The calculation formula is: Where: symbol Represents the dynamically calculated distance threshold, symbol Represents the density adjustment coefficient, symbol Represents the effective surface area of ​​the breeding shed, symbol This represents the total number of cashmere goats in the barn. The system periodically updates the distance determination threshold based on a formula. The numerical values ​​and quantity thresholds are also adjusted proportionally. It can be understood that through traversal calculations and dynamic threshold determination, the system can identify spatially proximate subsets of sheep flocks from global location information, and by determining their geographical coverage, simultaneously acquire a set of data on temperature, humidity, illuminance, and harmful gas concentrations directly corresponding to the microenvironment of that subset.

[0030] In one embodiment of the invention, the rate of change in the consistency of head orientation and limb extension of cashmere goats within a subset of the flock is analyzed. The rate of change in head orientation consistency and limb extension is mapped to environmental parameters associated with the geographic location, including temperature gradient, humidity, light intensity, and concentration of harmful gases. By querying a pre-established behavior-environment mapping rule base, the expected behavior category of the flock subset under the current posture and environmental parameter combination is output. This category includes maintaining aggregation, slow diffusion, rapid collision, or migration to a specific area. The expected behavior category and its evolution path are defined as the expected behavioral evolution trend of the flock subset.

[0031] Based on the expected behavioral trends, the system determines whether a global intervention is needed for the distribution of all cashmere goats in the barn. Specifically, the logic is as follows: if the expected behavioral trend of any subset of goats is rapid collision, or if more than half of the goat subsets' expected behavioral trend is migration to the same specific area, then a global intervention is deemed necessary. If these conditions are not met, then a global intervention is deemed unnecessary, and the system only performs local observation and recording of individual goat subsets.

[0032] In practice, the consistency of head orientation and the rate of change of limb extension among cashmere goats within a subset of the flock are analyzed. The consistency of head orientation is quantified by calculating the standard deviation of the head orientation angles of all cashmere goats in the subset. The rate of change of limb extension is obtained by comparing the absolute difference between the current and previous limb extension values ​​and then normalizing it over time. The calculated values ​​of head orientation consistency and the rate of change of limb extension are then mapped to environmental parameters related to the geographic location, including temperature gradient, humidity, light intensity, and harmful gas concentration. This mapping is achieved by combining the quantified posture indicators with the environmental parameter indicators into a multi-dimensional feature vector.

[0033] In some embodiments, the system outputs the expected behavior category by querying a pre-established behavior-environment mapping rule base. The pre-established behavior-environment mapping rule base contains a series of "if-then" type logical rules derived from expert experience and historical data. For example, one logical rule states: if the standard deviation of head orientation consistency is less than 15 degrees, the rate of change in limb extension is greater than 0.05 ppm, and the ammonia concentration in the associated environmental parameters exceeds 20 parts per million, then the expected behavior category is "rapid collision." Another logical rule states: if the standard deviation of head orientation consistency is less than 30 degrees, the temperature gradient in the associated environmental parameters is negative and its absolute value is greater than 0.5 degrees Celsius per hour, and the illuminance is less than 100 lux, then the expected behavior category is "migration to a specific area," where the specific area is by default the area illuminated by the heating lamps inside the breeding shed. The system inputs the multi-dimensional feature vectors corresponding to the sheep flock subsets into a pre-established behavior-environment mapping rule base for matching rule by rule. When a feature vector satisfies all the conditions of a certain logical rule, the system outputs the expected behavior category corresponding to that logical rule. The expected behavior categories include maintaining aggregation, slow diffusion, rapid collision, or migration to a specific region. The expected behavior category output by the matching and its subsequent behavior transformation path defined in the pre-established behavior-environment mapping rule base are collectively defined as the expected behavior evolution trend of the sheep flock subset.

[0034] Optional, a consistency coefficient for the degree of head orientation consistency. A more accurate calculation can be performed using the vector method. The calculation formula is as follows: Where: symbol The consistency coefficient represents the degree of consistency in head orientation, with the symbol […]. The symbol represents the number of cashmere goats within a subset of the flock. and ... The value ranges from -1 to 1, with a higher degree of consistency in head orientation as the value approaches 1. It can be understood that the implementation logic for determining whether global intervention is needed based on the expected behavioral evolution trend is deterministic. After each analysis cycle, the system checks the expected behavior category outputs of all identified sheep flock subsets. When the system detects that the expected behavior category of any sheep flock subset is rapid collision, a global intervention decision is immediately triggered. Alternatively, when the system detects that more than half of the sheep flock subsets output expected behavior categories of migrating to the same specific area, a global intervention decision is also triggered. As long as either of these two conditions is met, the system formally determines that a global intervention is needed for the distribution of all cashmere goats in the sheepfold. If no sheep flock subset is currently determined to have a rapid collision expected behavior category, and the number of sheep flock subsets migrating to the same specific area does not exceed half of the total, the system determines that a global intervention is not needed, and the subsequent processing flow enters a mode of local observation and data recording only for individual sheep flock subsets.

[0035] In one embodiment of the invention, for each cashmere goat, all its flock subsets are identified. The expected behavior category for each flock subset is obtained. Different urgency coefficients are assigned to these expected behavior categories: rapid charging, migrating to a specific area, slow diffusion, and maintaining agglomeration. The cashmere goat's movement guidance priority value is calculated, which is equal to the sum of the urgency coefficients of all its flock subsets, multiplied by the cashmere goat's average movement speed over the most recent time period. All cashmere goats are then sorted in descending order based on their movement guidance priority values; higher values ​​indicate higher movement guidance priority.

[0036] For the cashmere goat subset with the highest movement guidance priority, a straight or near-straight path is planned directly to the target area, which is an open area with low current aggregation. For the cashmere goat subset with medium movement guidance priority, a segmented, progressive path is planned, with intermediate transition points set in the path. For the cashmere goat subset with the lowest movement guidance priority, a fine-tuned path is planned, allowing it to make small detours around its current position. Each planned path must undergo spatial and temporal testing to ensure that it does not conflict with fixed facilities in the barn or the preset paths of other cashmere goats at the same time and space point.

[0037] In its implementation, the system first iterates through the identification list of all cashmere goats in the pen. For each cashmere goat in the list, the system retrieves the set of all flock subsets to which it belongs within the current identification period, all exhibiting localized aggregation. This is accomplished by querying the association mapping table between flock subsets and member cashmere goats. A cashmere goat may belong to zero, one, or more flock subsets simultaneously. The system then obtains the expected behavior category for each flock subset to which this cashmere goat belongs. The expected behavior category is the output of the previous derivation steps, with a value of one of "maintain aggregation," "slow spread," "rapid collision," or "migrate to a specific area." The system internally predefines an urgency coefficient assignment mapping table, assigning a fixed urgency coefficient value to each expected behavior category. The urgency coefficient value reflects the level of urgency required for the corresponding behavior category.

[0038] In some embodiments, the urgency coefficient assignment logic is implemented through a static mapping table, as shown in Table 1, which is loaded during system initialization. Table 1 illustrates a mapping relationship between expected behavior categories and urgency coefficients. The system quickly retrieves the urgency coefficient value mapped to the expected behavior category for each subset of the flock to which a cashmere goat belongs by looking up the table.

[0039] Table 1: Mapping Table of Expected Behavioral Categories and Urgency Levels Optional, move boot priority value The calculation formula is: Where: symbol Represents the calculated mobile boot priority value, symbol Represents the total number of the current cashmere goat's flock subset, symbol This represents the urgency coefficient corresponding to the q-th subset of the cashmere goat's flock, obtained through a table lookup; the symbol is... This represents the average movement speed of the cashmere goat over a recent time period. The average movement speed is obtained by dividing the sum of the differential distances of the position coordinate sequence within that time period by the length of the time period. Based on this formula, the system calculates a movement guidance priority value for each cashmere goat in the barn. It's understandable that if a cashmere goat's flock subset exhibits anticipated behaviors with high urgency, such as "rapid collisions," the sum of their urgency coefficients will increase, along with their average movement speed. If the value is higher, the calculated movement guidance priority value will be higher. This will increase significantly. The system will calculate the movement guidance priority values ​​for all cashmere goats after... Then, based on the movement guidance priority value All cashmere goats are sorted in descending order, with the cashmere goat with the highest movement guidance priority value at the front of the list and the cashmere goat with the lowest movement guidance priority value at the back of the list. The order of movement guidance priority values ​​directly determines the differentiated generation order of subsequent path guidance strategies.

[0040] In practice, the process of generating differentiated path guidance strategies based on movement guidance priority unfolds according to a sorted list. The system divides the sorted list of cashmere goats into three subsets: the top 30% of cashmere goats with the highest movement guidance priority are assigned to the subset with the highest priority, the middle 40% to the subset with the medium priority, and the bottom 30% to the subset with the lowest priority. For the subset with the highest priority, a straight or near-straight path is planned directly to the target area. The target area is determined by real-time analysis of the global location density map of all cashmere goats in the shed. Open areas with low current density are selected as the target area. The path planning algorithm calculates a collision-free straight or near-straight trajectory from the current position to a specific point within the target area for each cashmere goat in the subset. For the subset with the medium priority, a segmented progressive path is planned. This segmented progressive path is decomposed into two or three consecutive straight segments, with an intermediate transition point set at the junction of each segment. The intermediate transition point is located in a low-density area in the direction leading to the target area. For the cashmere goats with the lowest priority for movement guidance, a fine-tuning path is planned. The fine-tuning path consists of a series of short-distance, low-curvature arcs or polylines, which allows the cashmere goats to make small-range detours within a circular area with a radius of 2 to 3 meters centered on their current position.

[0041] When planning each path, whether it's a straight path, a segmented progressive path, or a fine-tuned path, the system invokes the path conflict detection module. This module compares the planned path's spatial position with the 3D model of the fixed facilities within the goat shed, and performs a spatiotemporal occupancy simulation with paths already planned for other cashmere goats. This ensures the planned path doesn't conflict with the fixed facilities or other cashmere goat's preset paths at the same spatiotemporal point. If a conflict is detected, the path's direction or movement time is automatically adjusted until all paths pass the conflict detection. In practice, when comparing the spatial position of the 3D model, the path conflict detection module first loads pre-stored 3D model data of the fixed facilities within the goat shed. This model is constructed based on the actual layout of the goat shed and includes the geometric shapes and spatial coordinates of facilities such as feed troughs, water troughs, and pillars. The module discretizes the planned path into a series of continuous time-space point sequences. For each path point, it calculates the Euclidean distance between it and each vertex on the surface of the fixed facility's 3D model and finds the minimum distance value. The system compares the minimum distance value with an internally set tolerance parameter based on preset conflict criteria. This tolerance parameter defines the acceptable minimum safety interval. If the minimum distance value is less than the tolerance parameter, it is determined that there is a spatial conflict between the path point and the fixed facility. Simultaneously, the module performs spatiotemporal overlay analysis on the time-space point sequence of the currently planned path and the path point sequences already planned for other cashmere goats, checking for overlapping timestamps and similar location coordinates to ensure that different paths do not intersect at the same spatiotemporal point. If a conflict is detected, the module automatically adjusts the path direction, for example, by inserting detour points, modifying the path curvature, or adjusting the movement time arrangement, such as delaying the start time, until the recalculated path passes the conflict detection, ensuring that all paths are executed without conflict.

[0042] See Figure 4 This is a grouped comparison chart of path planning parameters for cashmere goats with different priorities. The higher the priority, the farther the target area; this aligns with the requirement of "highest priority requiring rapid migration to open areas." The number of path segments increases with lower priority, reflecting a differentiated strategy of "direct access for high priority, fine-tuning for low priority." This type of chart serves as a strategy verification tool for intelligent cashmere goat farming path planning, visually displaying the differences in path parameters corresponding to different priorities, ensuring the strategy matches priority requirements, verifying the rationality of path planning, and providing data support for subsequent path optimization.

[0043] In one embodiment of the invention, the planned straight or near-straight path, segmented progressive path, and fine-tuned path are converted into a sequence of specific control commands for adjustable feeders, controllable lighting, directional sound devices, and automatic access control within the goat shed. Control of the adjustable feeders, controllable lighting, directional sound devices, and automatic access control is initiated with time delays, in descending order of movement guidance priority. During execution, the deviation between the actual position of the cashmere goats and the planned path is continuously monitored. If the deviation exceeds a preset tolerance, a real-time correction process for the cashmere goat path guidance strategy is triggered, and subsequent control command sequences are updated based on the correction results.

[0044] In practice, each path in the path set is associated with a cashmere goat identifier. Path types include straight or near-straight paths, segmented progressive paths, and fine-tuned paths. The system converts the planned straight or near-straight paths, segmented progressive paths, and fine-tuned paths into specific control command sequences for adjustable feeders, controllable lights, directional sound devices, and automatic access control within the goat shed. The conversion process follows a predefined equipment action mapping rule library. For example, for a straight path pointing to area A, the control command sequence includes: adjusting an adjustable feeder located in area A to open the concentrate feeding port; adjusting the controllable light above area A to switch to green and increase its brightness; controlling the directional sound device installed along the path to play a specific frequency call sound; and controlling the automatic access control leading to area A to remain open. For segmented progressive paths, the control command sequence triggers controllable lights and directional sound devices at different locations along the path in stages. For fine-tuned paths, the control command sequence only triggers a slight flashing of controllable lights or the directional sound device to play a low-frequency sound within a small area.

[0045] In some embodiments, the control of adjustable feeders, controllable lights, directional sound devices, and automatic access control is initiated in a time-delayed manner according to the movement guidance priority from high to low. The system generates a time-delayed start schedule table based on the cashmere goat movement guidance priority ranking list. The control command sequence corresponding to the cashmere goat subset with the highest movement guidance priority is executed immediately at time T0. The control command sequence corresponding to the cashmere goat subset with the medium movement guidance priority is delayed by ΔT1 seconds, i.e., it begins execution at time T0+ΔT1. The control command sequence corresponding to the cashmere goat subset with the lowest movement guidance priority is delayed by ΔT2 seconds, i.e., it begins execution at time T0+ΔT2, where ΔT2 is greater than ΔT1. This time-delayed start mechanism ensures that high-priority cashmere goats can respond to the guidance signal first and begin moving, thereby making room for the movement of subsequent low-priority cashmere goats and reducing path crossing interference. For example, ΔT1 is set to 5 seconds and ΔT2 to 10 seconds.

[0046] During execution, the system continuously monitors the deviation between the cashmere goats' actual positions and the planned path, with the monitoring period synchronized with the image capture period of the visual sensor. The cashmere goats' actual positions are obtained through real-time image analysis, while the planned path is defined as a sequence of expected position points that changes over time. For each cashmere goat, at each time t, the deviation between its actual position and the planned path is obtained by calculating the Euclidean distance between the actual position coordinates and the coordinates of the target point on the planned path at time t. Deviation tolerance value. This is a preset distance threshold, for example, set to 0.8 meters. The formula for determining if the deviation exceeds the tolerance is: Where: symbol Represents the calculated Euclidean distance between the actual location and the target point, with the sign... and Represents the actual planar coordinates of the cashmere goat at time t, with the symbol... and The planar coordinates of the target point on the planned path at time t are represented by the symbol. This represents the tolerance value for deviation. It can be understood that when the system detects a deviation in a particular cashmere goat... Exceeding the tolerance limit If the path guidance strategy for the cashmere goat is not immediately triggered, a real-time correction process is initiated. This process calls the path planning module, using the cashmere goat's latest actual location as the starting point and either the original target area or a new alternative target area as the ending point, to recalculate a path that avoids current obstacles and the latest locations of other cashmere goats. Based on the recalculated path, the system generates new control command sequences for the adjustable feeder, controllable lights, directional sound devices, and automatic access control. These new sequences update any unexecuted suffix control command sequences from the original plan. Subsequent updates to the control command sequences are dynamic, continuing until the distribution of the cashmere goats is adjusted to meet the global intervention objective.

[0047] See Figure 5 This is a bar chart showing the correlation between the priority of cashmere goat movement guidance and the expected behavior category. Priority is strongly correlated with behavior; cashmere goats exhibiting the expected behavior of "rapid collision" have a significantly higher priority than other behavior categories; cashmere goats exhibiting the expected behavior of "maintaining aggregation" have the lowest priority. A clear priority gradient is observed, with priority decreasing in a stepwise manner from "rapid collision" to "maintaining aggregation," consistent with the logic of assigning urgency coefficients. This type of chart serves as a decision-making basis for cashmere goat aggregation behavior intervention strategies, quickly identifying high-priority intervention targets, ensuring that risky behaviors such as "rapid collision" are addressed first; verifying the rationality of priority calculations; and helping to differentiate intervention levels, avoiding resource waste.

[0048] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0049] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A fully indoor, intelligent method for raising cashmere goats, characterized in that, The method includes: The location coordinates and posture information of cashmere goats were collected inside the breeding shed, and environmental parameters were collected at multiple locations within the breeding shed. Based on the collected location coordinates, a subset of sheep flocks exhibiting local clustering patterns was identified; For the identified subset of sheep flocks exhibiting local clustering patterns, environmental parameters associated with the geographical location of the sheep flock subset are acquired simultaneously. By combining the posture information of a subset of sheep with synchronously acquired associated environmental parameters, the expected behavioral evolution trend of the subset of sheep under the influence of the associated environmental parameters is deduced. Based on the expected behavioral evolution trend, determine whether a global intervention is needed to control the distribution of all cashmere goats in the breeding shed; When it is determined that a global intervention is required, the movement guidance priority for each cashmere goat in the breeding shed is calculated. Based on movement guidance priority, generate differentiated path guidance strategies for each cashmere goat; Implement the differentiated path guidance strategy to adjust the distribution of cashmere goats in the barn.

2. The intelligent farming method for fully indoor cashmere goats according to claim 1, characterized in that, The location coordinates and posture information of cashmere goats were collected inside the goat pen, and environmental parameters were collected at multiple locations within the pen, specifically including: Images are periodically captured by visual sensors deployed on the top of the breeding shed, and the real-time position coordinates of each cashmere goat, as well as the posture information such as head orientation and limb extension, are extracted from the images. Through a sensor network distributed on the ground of the breeding shed and at different heights, environmental parameters such as temperature, humidity, light intensity and concentration of harmful gases are collected synchronously at corresponding locations. Establish a timestamp alignment mechanism to ensure that the location coordinates and posture information of each cashmere goat collected have the same time reference as the environmental parameters of each location point.

3. The intelligent farming method for fully indoor cashmere goats according to claim 2, characterized in that, Based on the collected location coordinates, subsets of sheep flocks exhibiting localized clustering patterns were identified, specifically including: Set a distance determination threshold; For each cashmere goat in the barn, calculate the Euclidean distance between its location coordinates and the location coordinates of all other cashmere goats in the barn; Filter out all other cashmere goats whose Euclidean distance from the current cashmere goat is less than the distance determination threshold; If the number of selected cashmere goats exceeds the set threshold, the current cashmere goat and all other selected cashmere goats will be defined together as a subset of the flock with a local clustering pattern. Traverse all cashmere goats in the breeding shed to identify all subsets of the flock that exhibit localized clustering patterns.

4. The intelligent farming method for fully indoor cashmere goats according to claim 3, characterized in that, For the identified subset of sheep exhibiting localized clustering patterns, environmental parameters associated with the geographical location of the sheep subset are simultaneously acquired, specifically: Determine the geographical coverage of the identified sheep flock subset; Within the geographical coverage area, locate the deployment locations of all sensor nodes; The system acquires environmental parameters such as temperature, humidity, illuminance, and concentration of harmful gases collected by the sensor nodes after timestamp alignment. The acquired set of environmental parameters is labeled as the geographical location-associated environmental parameters of the sheep flock subset.

5. The intelligent farming method for fully indoor cashmere goats according to claim 4, characterized in that, By combining the posture information of a sheep subset with synchronously acquired associated environmental parameters, the expected behavioral evolution trend of the sheep subset under the influence of the associated environmental parameters is derived, specifically including: Analyze the rate of change in head orientation consistency and limb extension of cashmere goats within a subset of the flock; The rate of change of head orientation consistency and limb extension is correlated and mapped with the temperature change gradient, humidity value, light intensity and harmful gas concentration in the geographical location-related environmental parameters. By querying a pre-established behavior-environment mapping rule base, the expected behavior category of a subset of sheep under the current posture and environmental parameter combination is output; the expected behavior category includes maintaining aggregation, slow diffusion, rapid collision, or migration to a specific area. The expected behavior categories and their possible evolution paths are defined as the expected behavior evolution trends of the sheep subset.

6. The intelligent farming method for fully indoor cashmere goats according to claim 5, characterized in that, Based on the expected behavioral evolution trend, determine whether a global intervention is needed to control the distribution of all cashmere goats in the breeding shed, specifically: When the expected behavioral trend of any subset of sheep flocks is rapid collision, or when the expected behavioral trend of more than half of the sheep flocks is migration to the same specific area, it is determined that a global intervention is needed on the distribution of all cashmere goats in the breeding shed. Otherwise, it is determined that no global intervention is needed, and only local observation of individual sheep subsets is required.

7. The intelligent farming method for fully indoor cashmere goats according to claim 6, characterized in that, When it is determined that a global intervention is necessary, the movement guidance priority for each cashmere goat in the barn is calculated, and the specific process is as follows: For each cashmere goat, identify all the subsets of the flock to which it belongs; Obtain the expected behavior category for each sheep subset to which it belongs; Different urgency coefficients are assigned to the expected behavior categories of rapid collision, migration to a specific area, slow spread, and maintaining aggregation. Calculate the movement guidance priority value of the cashmere goat, which is equal to the sum of the urgency coefficients of all the flock subsets to which it belongs, and then multiply it by the average movement speed of the cashmere goat in the most recent time period. All cashmere goats are sorted in descending order based on their movement guidance priority value; the higher the value, the higher the movement guidance priority.

8. The intelligent farming method for fully indoor cashmere goats according to claim 7, characterized in that, Based on movement guidance priorities, a differentiated path guidance strategy is generated for each cashmere goat, specifically including: Plan a straight or near-straight path to the target area for the cashmere goat subset with the highest priority for movement guidance, where the target area is an open area with low current aggregation. Plan segmented, progressive paths for a subset of cashmere goats with medium priority for movement guidance, and set intermediate transition points in the paths; Plan a fine-tuned path for the cashmere goat subset with the lowest priority for movement, so that it makes a small detour around its current position; Each planned path must ensure that it does not conflict with the fixed facilities in the breeding shed or the pre-set paths of other cashmere goats at the same time and space point.

9. The intelligent farming method for fully indoor cashmere goats according to claim 8, characterized in that, Implementing the differentiated path guidance strategy to adjust the distribution of cashmere goats within the barn includes: The planned straight or near-straight paths, segmented progressive paths, and fine-tuning paths are converted into specific control command sequences for adjustable feeders, controllable lights, directional sound devices, and automatic access control in the breeding sheds. Control of the adjustable feeder, controllable lights, directional sound devices, and automatic access control is initiated in descending order of movement guidance priority with time delays. During execution, the deviation between the actual location of the cashmere goats and the planned path is continuously monitored. If the deviation exceeds the tolerance, a real-time correction process for the cashmere goat path guidance strategy is triggered, and the subsequent control command sequence is updated.

10. An intelligent cashmere goat farming system for fully indoor feeding, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the intelligent breeding method for fully house-fed cashmere goats as described in any one of claims 1 to 9.

Images