A method and system for monitoring and early warning of health of intensive breeding of Tibetan fragrant chicken

Abstract

The present application provides a kind of Tibetan chicken intensive breeding health monitoring early warning method and system, it is related to computer technology field, including: obtain the three-dimensional movement and spacing time series data of Tibetan chicken flock in breeding house, and original crowing audio time series data;According to three-dimensional movement and spacing time series data, motion pattern analysis is carried out, and vector field sequence is obtained;According to vector field sequence, stress level is inferred, and group implicit stress index is obtained;According to group implicit stress index and original crowing audio time series data, model construction is carried out, and crowing-stress correlation model is obtained;Real-time crowing audio data is acquired, and real-time crowing audio data is input into crowing-stress correlation model to carry out health state critical point prediction, and instability precursor signal is obtained;According to instability precursor signal, grading early warning is carried out, and health early warning grade is obtained.The present application realizes the health risk caused by group stress of Tibetan chicken in intensive breeding environment to be predicted and early warned.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and more specifically, to a method and system for health monitoring and early warning in the intensive farming of Tibetan chickens. Background Technology

[0002] In recent years, health monitoring in the intensive farming of specialty poultry has received increasing attention. Tibetan chickens, due to their high activity level, outstanding flying ability, and complex social structure, exhibit health abnormalities earlier in subtle changes in group behavior rather than significant abnormalities in individual physiological indicators; simultaneously, their unique call communication system contains rich information about group status. Existing poultry health monitoring technologies mostly focus on monitoring and thresholding early warnings for individual weight, feed intake, or fixed-location environmental parameters of common poultry such as broilers and laying hens. These methods face significant limitations when applied to Tibetan chicken farming: First, they struggle to effectively capture early stress precursors such as dynamic and rapid group spatial behavioral conflicts (e.g., perch competition, overlapping escape paths) caused by high-density rearing and their inherently active nature; second, existing sound analysis technologies are typically limited to identifying abnormal sounds such as coughs and sneezes or simply monitoring the frequency and volume of calls, failing to decode the complex nonlinear patterns in Tibetan chicken calls related to group stress levels (e.g., the chaos of pitch changes, the determinism of call sequences).

[0003] Based on the shortcomings of the existing technologies, there is an urgent need for a method and system for health monitoring and early warning in the intensive farming of Tibetan chickens. Summary of the Invention

[0004] The purpose of this invention is to provide a health monitoring and early warning method for intensive Tibetan chicken farming, in order to improve the aforementioned problems. To achieve the above objective, the technical solution adopted by this invention is as follows:

[0005] Firstly, this application provides a method for health monitoring and early warning in the intensive farming of Tibetan chickens, including:

[0006] Acquire three-dimensional motion and spacing time-series data of Tibetan chicken flocks in the breeding house, as well as the original call audio time-series data of the flocks;

[0007] Motion pattern analysis is performed based on the three-dimensional motion and spacing time series data. By quantifying the expansion, aggregation and vortex behavior of the group, a vector field sequence characterizing the spatial dynamics of the group is obtained.

[0008] Stress levels are inferred based on the vector field sequence, and the population latent stress index is obtained by analyzing the spatiotemporal coupling relationship between the vector field vortex intensity and the inter-individual distance compression rate.

[0009] A model is constructed based on the group latent stress index and the original chirping audio time series data. By calculating the recursive graph density of the audio signal and performing nonlinear regression with the stress index, a chirping-stress correlation model is obtained.

[0010] Real-time chirping audio data is acquired and input into the chirping-stress correlation model for health status critical point prediction. By monitoring the abrupt change points of the chirping signal recursion graph density and its projection on the correlation model, the instability precursor signals of the group's health status are obtained.

[0011] A graded early warning system is implemented based on the aforementioned instability precursor signals. The health warning level is obtained by evaluating the cumulative effect and diffusion rate of the instability precursor signals over time.

[0012] Secondly, this application also provides a health monitoring and early warning system for intensive Tibetan chicken farming, including:

[0013] The acquisition module is used to acquire the three-dimensional movement and spacing time-series data of Tibetan chicken flocks in the breeding house, as well as the original call audio time-series data of the flocks;

[0014] The analysis module is used to perform motion pattern analysis based on the three-dimensional motion and spacing time series data, and obtain a vector field sequence characterizing the spatial dynamics of the group by quantifying the expansion, aggregation and vortex behavior of the group.

[0015] The inference module is used to infer the stress level based on the vector field sequence. By analyzing the spatiotemporal coupling relationship between the vector field vortex intensity and the inter-individual distance compression rate, the group latent stress index is obtained.

[0016] The construction module is used to construct a model based on the group latent stress index and the original chirping audio time series data. By calculating the recursive graph density of the audio signal and performing nonlinear regression with the stress index, a chirping-stress correlation model is obtained.

[0017] The prediction module is used to acquire real-time chirping audio data and input the real-time chirping audio data into the chirping-stress correlation model to predict the critical point of health status. By monitoring the abrupt change points of the chirping signal recursion graph density and its projection on the correlation model, the unstable precursor signals of the group's health status are obtained.

[0018] The assessment module is used to provide graded early warnings based on the instability precursor signals. By assessing the cumulative effect and diffusion rate of the instability precursor signals over time, the health warning level is obtained.

[0019] The beneficial effects of this invention are as follows:

[0020] This invention treats the Tibetan chicken flock as a whole dynamic system and constructs its three-dimensional spatial motion vector field sequence. Then, it analyzes the nonlinear correlation model between this sequence and the chaotic characteristics of the calling signal, and finally realizes early and accurate critical point prediction and early warning of health risks caused by group stress in intensive farming environments. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a schematic diagram of a health monitoring and early warning method for intensive Tibetan chicken farming as described in an embodiment of the present invention;

[0023] Figure 2 This is a schematic diagram of a health monitoring and early warning system for intensive Tibetan chicken farming as described in an embodiment of the present invention;

[0024] Figure 3 This is a schematic diagram of a health monitoring and early warning device for intensive Tibetan chicken farming, as described in an embodiment of the present invention.

[0025] The diagram is labeled as follows: 800, a health monitoring and early warning device for intensive Tibetan chicken farming; 801, processor; 802, memory; 803, multimedia component; 804, I / O interface; 805, communication component; 901, acquisition module; 902, analysis module; 903, inference module; 904, construction module; 905, prediction module; 906, evaluation module. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0027] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0028] Example 1

[0029] This embodiment provides a method for health monitoring and early warning in the intensive farming of Tibetan chickens.

[0030] See Figure 1 The figure shows that the method includes steps S100 to S600.

[0031] Step S100: Obtain the three-dimensional movement and spacing time-series data of the Tibetan chicken flock in the breeding house, as well as the original call audio time-series data of the flock;

[0032] The core of step S100 lies in constructing a multimodal data foundation that comprehensively reflects the behavioral characteristics of Tibetan chicken flocks. The three-dimensional motion data acquired in this step not only records the flock's movement on a plane, but more importantly, captures its activity trajectory in three-dimensional space. This is closely related to the Tibetan chicken's tendency to fly and perch at high altitudes. This type of data is typically collected using UWB (Ultra-Wideband) or similar three-dimensional positioning systems deployed in the poultry house, existing as a sequence of three-dimensional coordinates for each individual at continuous time points. Simultaneously, the real-time distance data between individuals directly reflects the dynamic changes in the social spatial structure of the flock under intensive farming conditions. This data can be derived in real-time from the raw coordinate data output by the aforementioned positioning system, existing as a time-series data of distance matrices between pairs of individuals. The continuously recorded raw call audio contains the complete acoustic characteristics of the flock's communication, usually recorded synchronously by microphone arrays placed in different areas of the house, existing as multi-channel continuous waveform data files. These three types of time-series data with precise timestamps are aligned and integrated through a synchronization protocol, providing necessary data support for assessing health status from the perspective of flock behavioral dynamics.

[0033] Step S200: Perform motion pattern analysis based on three-dimensional motion and spacing time series data, and obtain a vector field sequence characterizing the spatial dynamics of the group by quantifying the expansion, aggregation and vortex behavior of the group;

[0034] The core idea of ​​step S200 is to transform discrete individual motions into a continuous group dynamic field. By calculating the divergence of each individual's motion vector, the intensity of the flock's diffusion (exploration) or aggregation (herd behavior) in a specific area can be precisely quantified; while by calculating the curl of the motion vector, complex motion patterns such as vortices or circulations that may occur in the flock can be effectively identified. This introduction of concepts and methods from physical field theory into the analysis of animal group behavior elevates the description of group motion patterns from the traditional statistical level to the dynamic characterization level.

[0035] Step S300: Infer the stress level based on the vector field sequence. By analyzing the spatiotemporal coupling relationship between the vector field vortex intensity and the inter-individual distance compression rate, the population latent stress index is obtained.

[0036] Step S300 focuses on extracting potential stress information from group movement dynamics. In intensive, high-density rearing environments, abnormal compression of inter-individual distances is often accompanied by changes in movement patterns, particularly an increase in disordered movement. This step aims to reveal early stress states that have not yet led to obvious changes in physiological indicators but have been manifested through group spatial behavior by establishing a quantitative correlation between vortex intensity (characterizing the degree of movement disorder) and inter-individual distance compression rate (reflecting the sense of spatial oppression).

[0037] Step S400: Construct a model based on the group latent stress index and the original chirping audio time series data. Calculate the recursive graph density of the audio signal and perform nonlinear regression with the stress index to obtain the chirping-stress correlation model.

[0038] The core of step S400 is to establish a reliable mapping relationship between different modal behavioral indicators. The crowing of Tibetan pheasants is their primary means of social communication, and the chaotic characteristics of their crowing sounds (characterized by recursive graph density) are more sensitive to the psychological stress state of a group than traditional sound intensity or frequency. By using nonlinear regression modeling of this acoustic feature with a latent stress index based on spatial behavior analysis, a computational model is effectively constructed that can indirectly assess the stress level of a group through continuously monitored acoustic signals.

[0039] Step S500: Acquire real-time chirping audio data and input the real-time chirping audio data into the chirping-stress correlation model to predict the critical point of health status. By monitoring the abrupt change points of the chirping signal recursion graph density and its projection on the correlation model, the unstable precursor signals of the group's health status are obtained.

[0040] Step S500 facilitates the transition from a static model to dynamic prediction. This step identifies abrupt changes in the density of the chirping signal recurrence graph by analyzing the changes in the density in real time, and then evaluates these abrupt changes in the constructed chirping-stress correlation model, thereby enabling early detection and warning of potential critical shifts in the health status of the population.

[0041] Step S600: Based on the instability precursor signals, a graded early warning is issued. By evaluating the cumulative effect and diffusion rate of the instability precursor signals on the time axis, the health warning level is obtained.

[0042] Step S600 completes the transformation from theoretical early warning signals to practical application value. By continuously evaluating the instability precursor signals over time and analyzing their spatial propagation characteristics within a population, the final early warning level considers not only the instantaneous intensity of the signal but also its dynamic development characteristics, thus providing a more scientific and reliable basis for tiered intervention decisions in production management.

[0043] Further, step S200 includes steps S210 to S230.

[0044] Step S210: Based on the three-dimensional motion and spacing time series data, construct the individual motion vector field. By calculating the displacement vector of each Tibetan chicken at continuous time points, obtain the individual motion vector field covering the entire flock.

[0045] Step S220: Based on the individual motion vector field and the spacing information in the three-dimensional motion and spacing time series data, perform group behavior modal analysis, calculate the divergence of the vector field to quantify expansion and aggregation behavior, and simultaneously introduce the instantaneous change rate of the distance between individuals to characterize the conflict potential under high density, so as to obtain the divergence field and conflict potential parameters of group motion.

[0046] Step S230: Based on the divergence field and conflict potential parameters, integrate vortex behavior and dynamic characteristics. Calculate the curl of the individual motion vector field and couple the conflict potential parameters as weights into the curl calculation to enhance the dynamic characteristics of dense regions, thus obtaining a vector field sequence.

[0047] First, step S210 transforms the discrete position observation data into a continuous individual motion vector field covering the breeding space by calculating the displacement vector of each Tibetan chicken at continuous time points, thus laying the mathematical foundation for subsequent field analysis.

[0048] Specifically, let the position of the i-th chicken in the flock at time t be... ,

[0049] Then its displacement vector within the time interval Δt is: ;

[0050] By using spatial interpolation methods, discrete individual displacement vectors are constructed into a vector field in continuous space: ;

[0051] In the formula, Let represent the three-dimensional position coordinates of the i-th chicken at time t; Let x, y, and height represent the x-coordinate, y-coordinate, and height coordinate of the i-th chicken at time t, respectively. The three-dimensional position coordinates of the i-th chicken at time (t+Δt); V(x,y,z,t) represents the displacement vector of individual i within the time interval Δt; V(x,y,z,t) is a vector function whose components are denoted as Vi. x V y V z , , respectively represent the components of the vector field in the x, y, z directions, and represent the average movement trend of the flock of chickens at the point (x, y, z); The weighting function represents the position of an empty location (x,y,z) relative to the position of individual i. It is usually based on distance decay (such as a Gaussian function) to ensure that the displacement of neighboring chickens contributes more to the vector field.

[0052] Next, step S220 performs group behavior modal analysis based on this vector field. On the one hand, by calculating the divergence of the vector field—a field theory concept—that is, the tendency of the vector field to "diverge" or "converge" at a certain point, the macroscopic behaviors such as the expansion (positive divergence) or aggregation (negative divergence) of the chicken flock are objectively quantified. On the other hand, closely combined with the characteristics of the high-density scenario of intensive farming, the parameter of the instantaneous change rate of the distance between individuals is introduced to characterize the potential for conflict that may be caused by the rapid compression of space. Thus, the divergence field and conflict potential parameters that simultaneously contain macroscopic movement patterns and microscopic spatial interaction pressure are obtained.

[0053] The formula for calculating the divergence field is: ;

[0054] The formula for calculating the conflict potential parameter is as follows (for each individual i, calculate the rate of change of its distance from its neighboring individual j): ;

[0055] In the formula, It represents the divergence field, characterizing the expansion (positive) or aggregation (negative) behavior of a group; Represents the Nabla operator; The sign for partial derivatives; Let N be the conflict potential parameter of chicken i, representing the sense of spatial oppression felt by chicken i; i Represents the set of neighboring chickens of chicken i; This represents the distance between chickens i and j at time t; This represents the distance between chickens i and j at time (t+Δt).

[0056] Finally, step S230 integrates the dynamic characteristics, not only calculating the curl of the vector field (used to quantify the intensity of rotation or vortex motion in a local area), but more importantly, coupling the conflict potential parameters obtained in step S220 as weights into the curl calculation. This process amplifies the calculation of any tiny vortex behavior in dense areas with small inter-individual spacing and high conflict potential, thereby significantly enhancing the sensitivity of this method to detecting early stress behaviors such as disorder and anxiety caused by spatial competition in high-density groups. The final output is a vector field sequence that can accurately reflect the spatial dynamic characteristics of Tibetan chicken groups in a specific breeding environment.

[0057] Here, curl is a vector measure of a vector field, representing the degree of rotation of the field. For a vector field... Its curl is defined as: ;

[0058] To enhance the dynamic characteristics of high-conflict regions, conflict potential parameters are coupled into curl calculations. First, the discrete conflict potential parameters are... Transformed into a continuous field through interpolation Then, the weighted curl field is defined as: In the formula, Let be the curl field, representing the intensity of the rotational motion of the chicken flock at point (x, y, z). It is the magnitude of the curl vector; A weighted curl field is used to highlight disordered motion in high-conflict regions and improve stress detection sensitivity. For continuous conflict potential field, it represents the spatial distribution of spatial compression risk.

[0059] Further, step S300 includes steps S310 to S330.

[0060] Step S310: Extract vortex intensity and distance compression ratio features based on the vector field sequence. Quantify vortex intensity by calculating the local curl value of each spatial point in the vector field sequence, and simultaneously calculate the instantaneous compression ratio of the distance between individuals in the three-dimensional motion and spacing time series data to obtain the spatiotemporally aligned vortex intensity field and distance compression ratio field.

[0061] Step S320: Based on the vortex intensity field and the distance compression rate field, perform spatiotemporal coupling relationship analysis. By calculating the time-delay cross-correlation function between the vortex intensity field and the distance compression rate field, and identifying the guiding relationship between vortex enhancement and the early occurrence of distance compression at a preset spatiotemporal scale, obtain the coupling strength coefficient characterizing the leading nature of group escape behavior.

[0062] Step S330: Synthesize the latent stress index based on the coupling strength coefficient. By nonlinearly fusing the coupling strength coefficient with the spatial integral of the vortex intensity field and the time integral of the distance compressibility field, the degree of internal tension of the group caused by spatial competition is quantified, and the group latent stress index is obtained.

[0063] Understandably, the core of the above process lies in quantifying latent stress by analyzing the dynamic correlation between two key physical phenomena in flock movement—local disordered movement (vortex) and spatial pressure (distance compression). Step S310 first extracts the local curl value of each spatial point at each moment from the vector field sequence. The magnitude of this value directly reflects the disorder or rotation intensity of the flock's movement at that location, constituting a vortex intensity field. Simultaneously, it calculates the instantaneous rate of reduction in distance between individuals from the original spacing data, i.e., the distance compression rate, forming another spatiotemporally aligned data field. Step S320 is crucial for establishing causal inference. This process does not simply compare the intensity of the two fields, but rather detects whether the occurrence of the "distance compression" event systematically precedes the occurrence of the "vortex enhancement" event in time by calculating their time-delay cross-correlation function. If such a stable leading relationship is identified at a specific spatiotemporal scale, it indicates that a behavioral pattern of "escape-oriented disordered movement due to spatial compression" is prevalent in the flock, and its leading degree and correlation strength are quantified as a coupling strength coefficient. Step S330 ultimately involves index synthesis, where the coupling strength coefficient, representing the leading nature of behavioral patterns obtained in step S320, is nonlinearly fused with the overall performance of vortex intensity in global space (spatial integral) and the persistence of distance compression during the monitoring period (temporal integral). This fusion method assigns higher weight to the coupling strength coefficient, meaning that even if the overall vortex and compression intensities are not high, as long as the behavioral pattern of "compression leading to vortex" is very significant, it will be judged as having significant latent stress. This ultimately synthesizes a latent stress index that can more sensitively and earlier reflect the degree of internal tension in the group caused by spatial competition.

[0064] Further, step S400 includes steps S410 to S430.

[0065] Step S410: Extract chaotic dynamic features of the chirping signal based on the original chirping audio time series data. By embedding the audio signal into the phase space and constructing a recursive graph, calculate the density of the diagonal structure in the graph to quantify the determinism of the chirping signal and obtain the recursive graph density sequence of the chirping signal.

[0066] Step S420: Perform dynamic correlation analysis based on the recursive graph density sequence and the population latent stress index. By calculating the time-varying mutual information between the recursive graph density change rate and the stress index change rate, identify the warning lead time of the decrease in chirping determinism on the rise of population stress level, and obtain the dynamic correlation pattern parameters.

[0067] Step S430: Construct a nonlinear regression model based on the dynamic correlation mode parameters. By taking the recursive graph density and its rate of change as input and the stress index after introducing the early warning lead time compensation as output, fit a state-dependent autoregressive model to obtain the chirping-stress correlation model.

[0068] Specifically, step S410 employs a nonlinear time series analysis method to reconstruct the dynamic trajectory of a vocalization audio signal in a high-dimensional phase space. The determinism (or predictability) of the vocalization signal is quantified by calculating the density of the diagonal structure in the recurrence graph. The physical meaning is that vocalizations in healthy, stable states typically exhibit higher regularity and determinism (corresponding to higher recurrence graph density), while vocalizations in stressed, anxious states are more chaotic and random (corresponding to lower recurrence graph density). This transforms the continuous audio stream into a one-dimensional recurrence graph density time series. The core of step S420 is to dynamically mine the lead-lag relationship between acoustic features and stress levels. By calculating the time-varying mutual information between the rate of change of recurrence graph density and the rate of change of the group's latent stress index, it identifies at what time scale a decrease in vocalization determinism stably predicts a subsequent increase in group stress levels, thus determining a key "early warning lead time." Step S430 then uses the findings from the first two steps to construct the final prediction model. This model uses the current and recent recursive graph density and its changing trend (rate of change) as input features, and the group's latent stress index at a future time (i.e., after compensation for the "early warning lead time" determined in step S420) as the prediction target. By fitting a state-dependent autoregressive model, a "calling-stress association model" is finally obtained that can predict the future group stress level based on real-time calling signals, thus achieving a sublimation from correlation analysis to causal prediction.

[0069] Further, step S500 includes steps S510 to S530.

[0070] Step S510: Perform online calculation of recursion graph density and mutation identification based on real-time chirping audio data. By continuously calculating the recursion graph of the audio signal and detecting the continuous decay inflection point of its diagonal structure density, the deterministic mutation point of real-time chirping is obtained.

[0071] Step S520: Based on the real-time chirping deterministic mutation point and the chirping-stress correlation model, the mutation point is projected and located in the health state space. By inputting the recursive graph density and its first-order differential features at the mutation point into the model, the high-dimensional spatial distance between its output and the steady-state operating point of the model is calculated to obtain the local deviation of the individual's health state.

[0072] Step S530: Based on the local deviation of individual health status, the emergence judgment of the precursor to group instability is made. By analyzing the infectious chain reaction and temporal synchronicity of the local deviation in the spatial distribution of the flock, when the deviation signal is detected to show a nonlinear amplification and spatial diffusion pattern, the precursor signal of instability is obtained.

[0073] It should be noted that the above steps constitute a progressive analysis process from individual anomaly detection to group risk warning. Step S510 is responsible for online monitoring of the real-time audio stream. Its core is to identify abrupt changes in the deterministic characteristics of the call signal. By continuously calculating the recursive graph density and using statistical process control methods to detect whether there is an inflection point of continuous decay, it can determine whether a significant degradation event has occurred in the acoustic environment of the chicken flock. Step S520 then places the detected acoustic mutation event into the health state space defined by the established call-stress correlation model for precise localization. By inputting the recursive graph density value of the mutation point and its changing trend (first derivative) into the model, the distance between the predicted stress level corresponding to the event and the baseline value of the model under healthy steady state in high-dimensional space is calculated. This distance is quantified as the local deviation, realizing the transformation from "detecting an anomaly" to "assessing the severity of the anomaly". Step S530 ultimately completes the assessment of the risk of group instability. The key is to identify whether local deviation events are signs of group risk. This is achieved by analyzing whether the local deviations generated by multiple individuals exhibit chain-like transmission characteristics from near to far in the spatial distribution of the flock, and whether they show high synchronicity in time. Only when such deviation signals meet the conditions of both nonlinear amplification in intensity and continuous diffusion in space can they be confirmed as true precursor signals of group instability, thereby effectively distinguishing between local accidental events and the early outbreak of systemic risks.

[0074] Further, step S600 includes steps S610 to S630.

[0075] Step S610: Based on the instability precursor signal, evaluate the time accumulation effect of the signal strength. By calculating the integral value of the instability precursor signal within the sliding time window and comparing it with the adaptive threshold, the time accumulation strength index of the signal is obtained.

[0076] Step S620: Based on the time cumulative intensity index and the instability precursor signal, perform spatial diffusion rate analysis of the signal in the population. By tracking the propagation speed and range of the individual with a significant time cumulative intensity index that induces other individuals to generate instability precursor signals in its spatial neighborhood, the spatiotemporal diffusion intensity index of the signal is obtained.

[0077] Step S630: Based on the time cumulative intensity index and the spatiotemporal diffusion intensity index obtained in step S620, perform a nonlinear mapping of the warning level. By inputting the two indices into a nonlinear function based on a population dynamic propagation threshold model for fusion calculation, the health warning level is finally obtained.

[0078] Understandably, this process constructs a hierarchical early warning logic that ranges from judging the duration of instability and analyzing its spatial propagation to quantifying the final risk level. Step S610 first assesses the cumulative effect of instability precursor signals in the time dimension. Its core lies in assessing the severity of the signal's persistence by calculating the integral value of the signal within a sliding time window, rather than relying on the signal strength at a single instant. This is then compared with an adaptive threshold dynamically adjusted based on historical normal state data, resulting in a "time-cumulative intensity index" that reflects the duration of the abnormal state. This effectively filters out instantaneous, isolated interference signals. Step S620 introduces spatial dimension analysis based on this. The key technical operation is tracking "core individuals" whose time-cumulative intensity index has exceeded the threshold and analyzing whether these individuals have induced a chain reaction within their spatial neighborhood (e.g., chickens within a specific radius). This involves monitoring whether surrounding individuals subsequently exhibit instability precursor signals at later time points. By calculating the speed of this signal propagation and the range of affected individuals, a "spatiotemporal diffusion intensity index" is obtained. This index directly reflects the likelihood and scale of the potential risk spreading within the flock. Step S630 is the final decision fusion. It takes the indices representing "duration" and "spread range" obtained in the first two steps as inputs. Instead of performing a simple linear weighting, it inputs them into a nonlinear function based on a group dynamic propagation threshold model for calculation. The core mechanism of this function is to simulate the propagation dynamics of behavior or stress state in a high-density group. It sets a critical point or threshold effect, that is, when the cumulative intensity and spread range of the signal reach a certain nonlinear critical point, the overall risk level will jump, thus ultimately outputting a health warning level that is more in line with the propagation law of group behavior and can accurately reflect the level of systemic risk.

[0079] Example 2

[0080] like Figure 2 As shown in the figure, this embodiment provides a health monitoring and early warning system for intensive Tibetan chicken farming. The system includes:

[0081] The acquisition module 901 is used to acquire the three-dimensional movement and spacing time-series data of the Tibetan chicken flock in the breeding house, as well as the original call audio time-series data of the flock.

[0082] Analysis module 902 is used to perform motion pattern analysis based on three-dimensional motion and spacing time series data. By quantifying the expansion, aggregation and vortex behavior of the group, a vector field sequence characterizing the spatial dynamics of the group is obtained.

[0083] The inference module 903 is used to infer the stress level based on the vector field sequence. By analyzing the spatiotemporal coupling relationship between the vector field vortex intensity and the inter-individual distance compression rate, the population latent stress index is obtained.

[0084] Module 904 is used to build a model based on the group latent stress index and the original chirping audio time series data. By calculating the recursive graph density of the audio signal and performing nonlinear regression with the stress index, a chirping-stress correlation model is obtained.

[0085] The prediction module 905 is used to acquire real-time chirping audio data and input the real-time chirping audio data into the chirping-stress correlation model to predict the critical point of health status. By monitoring the abrupt change points of the chirping signal recursion graph density and its projection on the correlation model, the unstable precursor signals of the group's health status are obtained.

[0086] The assessment module 906 is used to provide graded early warning based on the instability precursor signals. By assessing the cumulative effect and diffusion rate of the instability precursor signals on the time axis, the health warning level is obtained.

[0087] In one specific embodiment of this application, the analysis module 902 includes:

[0088] The first analysis unit is used to construct the individual motion vector field based on the three-dimensional motion and spacing time series data. By calculating the displacement vector of each Tibetan chicken at continuous time points, the individual motion vector field covering the entire flock is obtained.

[0089] The second analysis unit is used to perform group behavior modal analysis based on the individual motion vector field and the spacing information in the three-dimensional motion and spacing time series data. It calculates the divergence of the vector field to quantify expansion and aggregation behavior, and simultaneously introduces the instantaneous change rate of the distance between individuals to characterize the conflict potential under high density, thereby obtaining the divergence field and conflict potential parameters of group motion.

[0090] The third analysis unit is used to integrate vortex behavior and dynamic characteristics based on the divergence field and conflict potential parameters. It calculates the curl of the individual motion vector field and couples the conflict potential parameters as weights into the curl calculation to enhance the dynamic characteristics of dense regions, thereby obtaining a vector field sequence.

[0091] In one specific embodiment of this application, the inference module 903 includes:

[0092] The first inference unit is used to extract vortex intensity and distance compression rate features based on the vector field sequence. It quantifies the vortex intensity by calculating the local curl value of each spatial point in the vector field sequence, and simultaneously calculates the instantaneous compression rate of the distance between individuals in the three-dimensional motion and spacing time series data to obtain the spatiotemporally aligned vortex intensity field and distance compression rate field.

[0093] The second inference unit is used to perform spatiotemporal coupling relationship analysis based on the vortex intensity field and the distance compression rate field. By calculating the time-delay cross-correlation function between the vortex intensity field and the distance compression rate field, and identifying the guiding relationship between vortex enhancement and the early occurrence of distance compression at a preset spatiotemporal scale, the coupling strength coefficient characterizing the leading nature of group escape behavior is obtained.

[0094] The third inference unit is used to synthesize the latent stress index based on the coupling strength coefficient. By nonlinearly fusing the coupling strength coefficient with the spatial integral of the vortex intensity field and the time integral of the distance compressibility field, the intrinsic tension of the group caused by spatial competition is quantified, and the group latent stress index is obtained.

[0095] In one specific embodiment of this application, the construction module 904 includes:

[0096] The first construction unit is used to extract the chaotic dynamic features of the chirping signal based on the original chirping audio time series data. By embedding the audio signal into the phase space and constructing a recursive graph, the density of the diagonal structure in the graph is calculated to quantify the determinism of the chirping signal, and the recursive graph density sequence of the chirping signal is obtained.

[0097] The second building unit is used to perform dynamic correlation analysis based on the recursive graph density sequence and the population latent stress index. By calculating the time-varying mutual information between the recursive graph density change rate and the stress index change rate, it identifies the warning lead time of the decrease in chirping determinism on the rise in population stress level and obtains dynamic correlation pattern parameters.

[0098] The third building unit is used to construct a nonlinear regression model based on the parameters of the dynamic correlation pattern. By taking the recursive graph density and its rate of change as input and the stress index after introducing the early warning lead time compensation as output, a state-dependent autoregressive model is fitted to obtain the chirping-stress correlation model.

[0099] In one specific embodiment of this application, the prediction module 905 includes:

[0100] The first prediction unit is used to perform online calculation of recursion graph density and mutation identification based on real-time chirping audio data. By continuously calculating the recursion graph of the audio signal and detecting the continuous decay inflection point of its diagonal structure density, the deterministic mutation point of the real-time chirping is obtained.

[0101] The second prediction unit is used to project and locate the mutation point in the health state space based on the real-time chirping deterministic mutation point and the chirping-stress association model. By inputting the recursive graph density and its first-order differential features at the mutation point into the model, the high-dimensional spatial distance between its output and the steady-state operating point of the model is calculated to obtain the local deviation of the individual's health state.

[0102] The third prediction unit is used to make an emergence judgment on the precursors of group instability based on the local deviation of individual health status. By analyzing the infectious chain reaction and temporal synchronicity of the local deviation in the spatial distribution of the flock, when the deviation signal is detected to show a nonlinear amplification and spatial diffusion pattern, the precursor signal of instability is obtained.

[0103] Example 3

[0104] Corresponding to the above method embodiments, this embodiment also provides a health monitoring and early warning device for intensive Tibetan chicken farming. The health monitoring and early warning device for intensive Tibetan chicken farming described below can be referred to in correspondence with the health monitoring and early warning method for intensive Tibetan chicken farming described above.

[0105] Figure 3 This is a block diagram illustrating a health monitoring and early warning device 800 for intensive Tibetan chicken farming, according to an exemplary embodiment. Figure 3 As shown, the Tibetan chicken intensive farming health monitoring and early warning device 800 may include: a processor 801 and a memory 802. The Tibetan chicken intensive farming health monitoring and early warning device 800 may also include one or more of the following: a multimedia component 803, an I / O interface 804, and a communication component 805.

[0106] The processor 801 controls the overall operation of the Tibetan chicken intensive farming health monitoring and early warning device 800 to complete all or part of the steps in the aforementioned Tibetan chicken intensive farming health monitoring and early warning method. The memory 802 stores various types of data to support the operation of the Tibetan chicken intensive farming health monitoring and early warning device 800. This data may include, for example, instructions for any application or method operating on the Tibetan chicken intensive farming health monitoring and early warning device 800, as well as application-related data, such as contact data, sent and received messages, images, audio, video, etc. The memory 802 can be implemented using any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The multimedia component 803 may include a screen and an audio component. The screen may be, for example, a touchscreen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signals may be further stored in the memory 802 or transmitted via the communication component 805. The audio component also includes at least one speaker for outputting audio signals. I / O interface 804 provides an interface between processor 801 and other interface modules, such as keyboards, mice, and buttons. These buttons can be virtual or physical. Communication component 805 is used for wired or wireless communication between the Tibetan chicken intensive farming health monitoring and early warning device 800 and other devices. Wireless communication includes Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, or 4G, or a combination thereof. Therefore, the corresponding communication component 805 may include a Wi-Fi module, a Bluetooth module, and an NFC module.

[0107] In an exemplary embodiment, a Tibetan chicken intensive farming health monitoring and early warning device 800 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to execute the aforementioned Tibetan chicken intensive farming health monitoring and early warning method.

[0108] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided. When executed by a processor, these program instructions implement the steps of the above-described method for health monitoring and early warning in intensive Tibetan chicken farming. For example, the computer-readable storage medium may be the memory 802 including the program instructions, which may be executed by the processor 801 of a Tibetan chicken intensive farming health monitoring and early warning device 800 to complete the above-described method for health monitoring and early warning in intensive Tibetan chicken farming.

[0109] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for health monitoring and early warning in intensive Tibetan chicken farming, characterized in that, include: Acquire three-dimensional motion and spacing time-series data of Tibetan chicken flocks in the breeding house, as well as the original call audio time-series data of the flocks; Motion pattern analysis is performed based on the three-dimensional motion and spacing time series data. By quantifying the expansion, aggregation and vortex behavior of the group, a vector field sequence characterizing the spatial dynamics of the group is obtained. Stress levels are inferred based on the vector field sequence, and the population latent stress index is obtained by analyzing the spatiotemporal coupling relationship between the vector field vortex intensity and the inter-individual distance compression rate. A model is constructed based on the group latent stress index and the original chirping audio time series data. By calculating the recursive graph density of the audio signal and performing nonlinear regression with the stress index, a chirping-stress correlation model is obtained. Real-time chirping audio data is acquired and input into the chirping-stress correlation model for health status critical point prediction. By monitoring the abrupt change points of the chirping signal recursion graph density and its projection on the correlation model, the instability precursor signals of the group's health status are obtained. A graded early warning system is implemented based on the aforementioned instability precursor signals. The health warning level is obtained by evaluating the cumulative effect and diffusion rate of the instability precursor signals over time. The stress level inference based on the vector field sequence includes: Based on the vector field sequence, vortex intensity and distance compression ratio features are extracted. The vortex intensity is quantified by calculating the local curl value of each spatial point in the vector field sequence, and the instantaneous compression ratio of the distance between individuals in the three-dimensional motion and spacing time series data is calculated simultaneously to obtain the spatiotemporally aligned vortex intensity field and distance compression ratio field. Based on the vortex intensity field and the distance compression rate field, a spatiotemporal coupling relationship analysis is performed. By calculating the time-delay cross-correlation function between the vortex intensity field and the distance compression rate field, and identifying the guiding relationship between vortex enhancement and the early occurrence of distance compression at a preset spatiotemporal scale, a coupling strength coefficient characterizing the leading nature of group escape behavior is obtained. The latent stress index is synthesized based on the coupling strength coefficient. By nonlinearly fusing the coupling strength coefficient with the spatial integral of the vortex intensity field and the time integral of the distance compressibility field, the degree of intrinsic tension of the group caused by spatial competition is quantified, and the group latent stress index is obtained.

2. The method for health monitoring and early warning in intensive Tibetan chicken farming according to claim 1, characterized in that, Motion pattern analysis is performed based on the aforementioned three-dimensional motion and spacing time-series data, including: Based on the three-dimensional motion and spacing time series data, an individual motion vector field is constructed. By calculating the displacement vector of each Tibetan chicken at continuous time points, an individual motion vector field covering the entire flock is obtained. Based on the individual motion vector field and the spacing information in the three-dimensional motion and spacing time series data, group behavior modality analysis is performed. The divergence of the vector field is calculated to quantify expansion and aggregation behavior, and the instantaneous rate of change of distance between individuals is introduced to characterize the conflict potential under high density, so as to obtain the divergence field and conflict potential parameters of group motion. Based on the divergence field and the conflict potential parameters, vortex behavior and dynamic characteristics are integrated. By calculating the curl of the individual motion vector field and coupling the conflict potential parameters as weights into the curl calculation to enhance the dynamic characteristics of dense regions, a vector field sequence is obtained.

3. The method for health monitoring and early warning in intensive Tibetan chicken farming according to claim 1, characterized in that, Model construction is performed based on the aforementioned group latent stress index and the original chirping audio time-series data, including: Based on the original chirping audio time-series data, the chaotic dynamic features of the chirping signal are extracted. By embedding the audio signal into the phase space and constructing a recursive graph, the density of the diagonal structure in the graph is calculated to quantify the determinism of the chirping signal, and the recursive graph density sequence of the chirping signal is obtained. Dynamic correlation analysis is performed based on the recursive graph density sequence and the population latent stress index. By calculating the time-varying mutual information between the recursive graph density change rate and the stress index change rate, the warning lead time of the decrease in chirping determinism on the rise in population stress level is identified, and dynamic correlation pattern parameters are obtained. A nonlinear regression model is constructed based on the parameters of the dynamic correlation pattern. By taking the recursive graph density and its rate of change as input and the stress index after introducing the early warning lead time compensation as output, a state-dependent autoregressive model is fitted to obtain the chirping-stress correlation model.

4. The method for health monitoring and early warning in intensive Tibetan chicken farming according to claim 1, characterized in that, Acquire real-time chirping audio data and input the real-time chirping audio data into the chirping-stress correlation model for predicting health status thresholds, including: Based on the real-time chirping audio data, the recurrence graph density is calculated online and abrupt changes are identified. By continuously calculating the recurrence graph of the audio signal and detecting the continuous decay inflection point of its diagonal structure density, the deterministic abrupt change point of the real-time chirping is obtained. Based on the real-time chirping deterministic mutation point and the chirping-stress correlation model, the mutation point is projected and located in the health state space. By inputting the recursive graph density and its first-order differential features at the mutation point into the model, the high-dimensional spatial distance between its output and the steady-state operating point of the model is calculated to obtain the local deviation of the individual's health state. Based on the local deviation of the individual's health status, the emergence of precursors to group instability is judged. By analyzing the infectious chain reaction and temporal synchronicity of the local deviation in the spatial distribution of the flock, when the deviation signal is detected to exhibit a nonlinear amplification and spatial diffusion pattern, an instability precursor signal is obtained.

5. A health monitoring and early warning system for intensive Tibetan chicken farming, characterized in that, include: The acquisition module is used to acquire the three-dimensional movement and spacing time-series data of Tibetan chicken flocks in the breeding house, as well as the original call audio time-series data of the flocks; The analysis module is used to perform motion pattern analysis based on the three-dimensional motion and spacing time series data, and obtain a vector field sequence characterizing the spatial dynamics of the group by quantifying the expansion, aggregation and vortex behavior of the group. The inference module is used to infer the stress level based on the vector field sequence. By analyzing the spatiotemporal coupling relationship between the vector field vortex intensity and the inter-individual distance compression rate, the group latent stress index is obtained. The construction module is used to construct a model based on the group latent stress index and the original chirping audio time series data. By calculating the recursive graph density of the audio signal and performing nonlinear regression with the stress index, a chirping-stress correlation model is obtained. The prediction module is used to acquire real-time chirping audio data and input the real-time chirping audio data into the chirping-stress correlation model to predict the critical point of health status. By monitoring the abrupt change points of the chirping signal recursion graph density and its projection on the correlation model, the unstable precursor signals of the group's health status are obtained. The assessment module is used to provide graded early warning based on the instability precursor signals. By assessing the cumulative effect and diffusion rate of the instability precursor signals on the time axis, the health warning level is obtained. The inference module includes: The first inference unit is used to extract vortex intensity and distance compression rate features based on the vector field sequence. It quantifies the vortex intensity by calculating the local curl value of each spatial point in the vector field sequence, and simultaneously calculates the instantaneous compression rate of the distance between individuals in the three-dimensional motion and spacing time series data to obtain the spatiotemporally aligned vortex intensity field and distance compression rate field. The second inference unit is used to perform spatiotemporal coupling relationship analysis based on the vortex intensity field and the distance compression rate field. By calculating the time-delay cross-correlation function between the vortex intensity field and the distance compression rate field, and identifying the guiding relationship between vortex enhancement and the early occurrence of distance compression at a preset spatiotemporal scale, the coupling strength coefficient characterizing the leading nature of group escape behavior is obtained. The third inference unit is used to synthesize a latent stress index based on the coupling strength coefficient. By nonlinearly fusing the coupling strength coefficient with the spatial integral of the vortex intensity field and the time integral of the distance compressibility field, the unit quantifies the degree of internal tension of the group caused by spatial competition and obtains the group latent stress index.

6. The Tibetan chicken intensive farming health monitoring and early warning system according to claim 5, characterized in that, The analysis module includes: The first analysis unit is used to construct an individual motion vector field based on the three-dimensional motion and spacing time series data. By calculating the displacement vector of each Tibetan chicken at continuous time points, an individual motion vector field covering the entire flock is obtained. The second analysis unit is used to perform group behavior modal analysis based on the individual motion vector field and the spacing information in the three-dimensional motion and spacing time series data. It calculates the divergence of the vector field to quantify expansion and aggregation behavior, and simultaneously introduces the instantaneous change rate of the distance between individuals to characterize the conflict potential under high density, thereby obtaining the divergence field and conflict potential parameters of group motion. The third analysis unit is used to integrate vortex behavior and dynamic characteristics based on the divergence field and the conflict potential parameters. It calculates the curl of the individual motion vector field and couples the conflict potential parameters as weights into the curl calculation to enhance the dynamic characteristics of the dense region, thereby obtaining a vector field sequence.

7. The Tibetan chicken intensive farming health monitoring and early warning system according to claim 5, characterized in that, The building module includes: The first construction unit is used to extract the chaotic dynamic features of the chirping signal based on the original chirping audio time-series data. By embedding the audio signal into the phase space and constructing a recursive graph, the density of the diagonal structure in the graph is calculated to quantify the determinism of the chirping signal, and a recursive graph density sequence of the chirping signal is obtained. The second construction unit is used to perform dynamic correlation analysis based on the recursive graph density sequence and the population latent stress index. By calculating the time-varying mutual information between the recursive graph density change rate and the stress index change rate, it identifies the warning lead time of the decrease in chirping determinism on the rise of population stress level and obtains dynamic correlation pattern parameters. The third construction unit is used to construct a nonlinear regression model based on the dynamic correlation mode parameters. By taking the recursive graph density and its rate of change as input and the stress index after introducing early warning lead time compensation as output, a state-dependent autoregressive model is fitted to obtain the chirping-stress correlation model.

8. The Tibetan chicken intensive farming health monitoring and early warning system according to claim 5, characterized in that, The prediction module includes: The first prediction unit is used to perform online calculation of recursion graph density and mutation identification based on the real-time chirping audio data. By continuously calculating the recursion graph of the audio signal and detecting the continuous decay inflection point of its diagonal structure density, the deterministic mutation point of the real-time chirping is obtained. The second prediction unit is used to project and locate the mutation point in the health state space based on the real-time chirping deterministic mutation point and the chirping-stress correlation model. By inputting the recursive graph density and its first-order differential features at the mutation point into the model, the high-dimensional spatial distance between its output and the steady-state operating point of the model is calculated to obtain the local deviation of the individual's health state. The third prediction unit is used to make an emergence judgment on the precursor of group instability based on the local deviation of the individual's health status. By analyzing the infectious chain reaction and temporal synchronicity of the local deviation in the spatial distribution of the flock, when the deviation signal is detected to show a nonlinear amplification and spatial diffusion pattern, an instability precursor signal is obtained.

Images