Intelligent drinking water scheduling and water-saving control method for laying hen houses integrating multi-source sensing

Abstract

The present invention relates to the field of water control technology, specifically to a method for intelligent drinking water scheduling and water conservation control in laying hen houses that integrates multi-source sensing. The method comprises the following steps: obtaining a real-time spatiotemporal distribution map of a chicken flock through a positioning sensor network, constructing a partition grid based on drinking water line pressure pulsation data, and dividing the chicken house into multiple control units with independent hydraulic characteristics based on the partition grid; generating a drinking water behavior pattern matrix based on the trigger frequency sequence of the chicken beak touch sensors in each control unit, and simultaneously integrating environmental temperature and humidity gradient data to establish a water demand compensation function for each zone; and, based on the drinking water behavior pattern matrix and the water demand compensation function, driving the microchannel water-saving valve corresponding to each zone to implement a differentiated water supply mode, and synchronously adjusting the opening and closing timing of the zone aeration device. The present invention achieves real-time dynamic flow regulation, effectively balancing physiological water demand with system water conservation goals, and significantly improving the intelligence and robustness of the drinking water control strategy.

Description

Technical Field

[0001] The present invention relates to the technical field of water use regulation, and in particular to an intelligent drinking water scheduling and water-saving control method for a laying hen house integrating multi-source sensing. Background Art

[0002] In current large-scale laying hen farming, drinking water systems, as crucial infrastructure for ensuring the vital activities and production performance of poultry, are subject to widespread regulatory challenges, including delayed response, extensive regulation, and inefficient water conservation. Traditional drinking water control methods, often based on timed water supply or fixed flow rates, lack the ability to dynamically monitor and respond to actual drinking behavior and environmental changes. This can easily lead to oversupply and waste of water resources, or insufficient water supply that can compromise poultry health and egg production.

[0003] Some water-saving controls have introduced local behavior detection devices, such as touch sensors and infrared monitoring, to roughly assess drinking activity. However, they usually fail to construct high-resolution spatiotemporal behavior maps that can be used for zoning control, nor do they work in synergy with temperature and humidity environmental factors. At the same time, the hydraulic state of the chicken house pipe network during water supply (such as pressure fluctuations and flow distribution) is not included in the control strategy, resulting in the control unit being unable to adaptively adjust the flow structure or respond to peak behavior, thereby increasing the risk of operational instability.

[0004] Furthermore, current systems generally separate the control of drinking water devices from aeration systems, lacking a coordinated gas-liquid mechanism and unable to address issues such as localized water quality deterioration or biofilm accumulation. High-frequency pulsed water supply, without effective hydraulic safety protection strategies, can easily induce faults such as water hammer and pipeline oscillation, further limiting the deployment and expansion of intelligent drinking water systems. Summary of the Invention

[0005] The present invention provides an intelligent drinking water scheduling and water-saving control method for laying hen houses that integrates multi-source sensing. By integrating multi-source sensing information, the method can dynamically perceive the drinking behavior characteristics of chickens, realize water demand compensation in combination with environmental parameters, and has the intelligent drinking water scheduling and water-saving control method with high-frequency pulse response, gas-liquid linkage and hydraulic safety protection capabilities, so as to improve the accuracy, adaptability and operational reliability of the chicken house water supply system.

[0006] The intelligent drinking water scheduling and water-saving control method for laying hen houses integrating multi-source sensing includes the following steps:

[0007] S1: Using a positioning sensor network to obtain a real-time spatiotemporal distribution map of the chicken flock, a partition grid is constructed based on the drinking line pressure pulsation data. Based on the partition grid, the chicken house is divided into multiple control units with independent hydraulic characteristics.

[0008] S2: Based on the trigger frequency sequence of the chicken beak touch sensor in each control unit, a drinking behavior pattern matrix is ​​generated, and the ambient temperature and humidity gradient data are integrated to establish the water demand compensation function of each partition;

[0009] S3: According to the drinking water behavior pattern matrix and water demand compensation function output by S2, the microchannel water-saving valve corresponding to each partition is driven to execute a differentiated water supply mode, and the opening and closing timing of the aeration device of the partition is synchronously adjusted.

[0010] Optionally, the positioning sensing network in S1 includes an infrared sensor array, an RFID positioning beacon and a ToF camera. The infrared sensor array, RFID positioning beacon and ToF camera arranged on the ceiling of the chicken house generate three-dimensional coordinate point cloud data of the chicken flock, and the spatiotemporal distribution map is updated at a predetermined period. The spatiotemporal distribution map includes a heat map of chicken density in each area and a movement trajectory vector.

[0011] Optionally, S1 further includes installing a high-frequency pressure transmitter in the drinking water main pipeline, collecting pressure pulsation data, and extracting characteristic frequency components representing the topological structure of the pipe network through wavelet packet decomposition;

[0012] The chicken density heat map and the characteristic frequency components of the pipe network are input into the DBSCAN clustering algorithm to generate control units (i.e., partition grids) that meet the following constraints:

[0013] Constraint 1: The coefficient of variation in chicken density within the control unit does not exceed a predetermined percentage;

[0014] Constraint 2: The correlation coefficient of pipeline pressure fluctuations between adjacent control units is less than the predetermined coefficient threshold.

[0015] Optionally, S1 further includes a partition grid reconstruction mechanism: when any trigger condition is met, the partition grid is reconstructed:

[0016] Trigger condition 1: The chicken population density change rate in any sub-area exceeds 10% / minute;

[0017] Trigger condition two: The pressure pulsation main frequency offset is greater than the offset threshold.

[0018] Optionally, the S2 specifically includes:

[0019] S21: Install a chicken beak touch sensor at the drinking water terminal of each control unit and set a sampling period to record valid touch events. The valid touch events must meet both acceleration conditions and voiceprint conditions.

[0020] S22: Count the effective touch frequency sequence of each control unit according to the predetermined time window to construct a drinking behavior pattern matrix with extended time and space dimensions. , the matrix elements represents the normalized touch intensity of the i-th unit in the j-th time window;

[0021] S23, based on the current temperature and humidity environment, build a water demand compensation function , used to adjust the water supply target value of each control unit.

[0022] Optionally, the acceleration condition is expressed as: And the duration is <300ms, where The maximum acceleration value of the touch event;

[0023] The voiceprint condition is expressed as: ,in is the energy ratio of swallowing soundprint in the 500-800Hz frequency band, It is the total energy of the entire voiceprint signal in the full frequency band, that is, the sum of the energy of the sound signal in the entire recording time window (for example, 0–4 kHz).

[0024] Optionally, the water demand compensation function Expressed as:

[0025] ;

[0026] in, is the baseline water supply, is the current ambient temperature, is the current ambient relative humidity, 、 It is the temperature and humidity benchmark value for a comfortable environment for laying hens. is the temperature change response coefficient, is the humidity square response coefficient, Behavior influence gain coefficient, used to control the weight of drinking behavior in compensation, It is a hyperbolic tangent function, which is used to constrain the saturation effect of the behavioral factor and prevent surge compensation.

[0027] Optionally, in S3, according to the drinking behavior pattern matrix of each control unit Normalized touch strength in , determine the control mode level, including:

[0028] like If the high threshold is exceeded, the high-frequency pulse mode is activated, and the microchannel water-saving valve is set to a high-duty-cycle short-cycle action, and the single water supply flow is increased to the upper limit of the reference value;

[0029] like If it is in the middle range, it enters the adaptive pulse mode, determines the water-saving valve switch cycle through the logarithmic function, and uses the water demand compensation amount calculated in the early stage as the current pulse water supply benchmark;

[0030] like If the water level is lower than the low threshold, the water-saving monitoring mode is enabled, and water is only supplied tentatively for a short period of time within a predetermined period, and whether to shut down some drinking water terminals is determined by effective touch.

[0031] Optionally, the S3 further includes executing sequential coupling control of the aeration device and the microchannel water-saving valve, specifically including:

[0032] After the micro-channel water-saving valve enters the open state, the Venturi aerator is started with a delay. The delay value is used to match the water flow propagation time to ensure synchronous diffusion of gas and liquid. The aeration duration is positively correlated with the current pulse flow rate.

[0033] When it is detected that the dissolved oxygen concentration decrease rate of the control unit exceeds the set decrease threshold, an additional aeration process is inserted during the water supply interval.

[0034] Optionally, the method further includes establishing a valve-aeration linkage protection mechanism, which automatically switches to a constant flow mode if abnormal pipeline pressure fluctuations are detected for multiple consecutive pulse cycles, or if the energy proportion of the pressure signal in the high-frequency band suddenly increases and exceeds a judgment threshold.

[0035] Beneficial effects of the present invention:

[0036] 1. This invention integrates multi-source spatial perception methods such as infrared arrays, RFID positioning, and Time of Flight (ToF) cameras to construct density heat maps and trajectory vector fields. Combined with acceleration and voiceprint sensors, this method extracts effective drinking events based on behavioral and acoustic dual-modal standard criteria. It then constructs a spatiotemporally expanded drinking behavior pattern matrix, categorizing the behavior into three levels of response mode based on intensity. Combined with logarithmic periodic regulation and water-saving detection strategies, this method achieves fine-grained hierarchical control of control units under different drinking conditions, improving the system's response accuracy and physiological adaptability to dynamic changes in drinking water demand.

[0037] 2. This invention introduces a nonlinear compensation function for temperature and humidity, constructs a dynamic water demand regulation model using real-time ambient temperature, relative humidity, and drinking behavior intensity as parameters, and uses a hyperbolic tangent function to limit the impact of extreme behavior, suppressing the risk of misjudgment due to abnormal amplification of drinking behavior in high temperature or high humidity scenarios. The compensation function is embedded in an adaptive pulse water supply mode to achieve real-time dynamic regulation of flow, effectively balancing physiological water demand and system water conservation goals, and significantly improving the intelligence and robustness of the drinking water control strategy.

[0038] 3. This invention achieves synchronous gas-liquid diffusion through an aeration delay start mechanism based on water flow propagation delay, adjusts aeration duration based on current pulse intensity and behavioral deviation, integrates dissolved oxygen change rate to drive intermittent aeration compensation, improves drinking water quality and device adaptability, and proposes a dual-domain pressure fluctuation determination mechanism. Based on the pulse cycle pressure peak change and the high-frequency energy mutation of the spectrum, it triggers the linkage protection program and constant flow switching logic in real time, and automatically reduces branch pressure to prevent water hammer impact and pipe network damage, enhancing system reliability and operational safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0040] Figure 1 Schematic diagram of the control method flow in an embodiment of the present invention;

[0041] Figure 2 Schematic diagram of a differentiated water supply mode according to an embodiment of the present invention. DETAILED DESCRIPTION

[0042] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. Those skilled in the art may also implement some known technologies in other alternative ways. The accompanying drawings are only for describing the embodiments in more detail and are not intended to limit the present invention in any specific way.

[0043] like Figure 1-Figure 2 As shown, the intelligent drinking water scheduling and water-saving control method for laying hen houses integrating multi-source sensing includes the following steps:

[0044] S1: The spatiotemporal distribution map of the chicken flock is obtained in real time through a positioning sensor network. The partition grid is constructed based on the pressure pulsation data of the drinking water line. The chicken house is divided into multiple control units with independent hydraulic characteristics based on the partition grid.

[0045] Each control unit includes:

[0046] 1 set of controllable micro-channel water-saving valve;

[0047] 1 aeration device node;

[0048] Multiple drinking water terminals;

[0049] S2: Based on the trigger frequency sequence of the chicken beak touch sensor in each control unit, a drinking behavior pattern matrix is ​​generated, and the ambient temperature and humidity gradient data are integrated to establish the water demand compensation function of each partition;

[0050] S3: According to the drinking water behavior pattern matrix and water demand compensation function output by S2, the microchannel water-saving valve corresponding to each partition is driven to execute a differentiated water supply mode, and the opening and closing timing of the aeration device of the partition is synchronously adjusted.

[0051] The positioning sensor network includes an infrared sensor array, RFID positioning beacons, and a ToF (Time of Flight) camera. The S1 specifically includes:

[0052] S11, spatiotemporal distribution map construction: Using an infrared sensor array, RFID positioning beacons, and ToF (Time of Flight) cameras placed on the ceiling of the chicken house, we generate three-dimensional coordinate point cloud data of the chicken flock. This data is updated every 5 seconds to obtain a spatiotemporal distribution map of the chicken flock. This map includes:

[0053] Heat map of chicken density;

[0054] Movement trajectory vector field.

[0055] The fusion method is to use Kalman filtering to fuse infrared temperature hotspot data (accuracy ±0.5°C), RFID positioning data (resolution 10 cm) and ToF depth images to generate a dynamic distribution map.

[0056] S12, Hydraulic Characteristic Signal Acquisition and Processing: Install a high-frequency pressure transmitter on the drinking water main pipeline with a sampling rate set to 20–100 Hz to collect the original pressure pulsation signal. Wavelet packet decomposition is used to extract the following key frequency characteristic components that characterize the pipe network topology:

[0057] Main frequency band energy ratio : Used to identify the impedance characteristics of the pipe section;

[0058] Kurtosis coefficient : Characterizes the local turbulence intensity;

[0059] Band mutual information entropy : Determine the connectivity status of the pipeline network.

[0060] S13, clustering of control units: The flock density heat map and the hydraulic characteristic frequency components are input into the DBSCAN clustering algorithm, which generates control units that meet the following constraints:

[0061] The coefficient of variation of chicken density within a unit does not exceed 15%;

[0062] The correlation coefficient of pipeline pressure fluctuations between adjacent units is less than 0.3;

[0063] The core mechanism of DBSCAN is as follows:

[0064] S131, the distance metric function is defined as: ;in, is the pressure characteristic difference between nodes, is the Euclidean space distance between nodes, Represents the weight coefficient.

[0065] S132, adaptive neighborhood radius : ;in, is the regulating factor, is the impedance of the current pipe segment;

[0066] S133, dynamic adjustment rules for the minimum number of cluster points: ;in, is the total number of chickens in the current chicken house, The number of partitions currently divided.

[0067] S14, partition grid reconstruction mechanism: When any of the following trigger conditions is met, the dynamic partition grid is rebuilt:

[0068] S141. The rate of change of chicken population density in any sub-region satisfies: ;in, is the change in the number of chickens per unit time, is the area of ​​the region, is the sampling interval.

[0069] S142. Pressure pulsation main frequency offset is greater than the threshold: ; The center frequency of the main frequency band is obtained by performing STFT (short time Fourier transform) on the pressure signal The reconstruction process retains 30% of the original partition structure as the topological skeleton to ensure control continuity and response smoothness.

[0070] In the traditional DBSCAN algorithm, clustering is usually based on pure spatial coordinates or a certain location information. In this scheme, the clustering input is expanded to a high-dimensional feature vector containing multimodal features. Each point to be clustered represents a sampling unit grid area in the chicken house space. Its input features include:

[0071] ;in, For the The density of chickens in the area, is the density gradient, which characterizes the distribution trend of the chicken flock. is the proportion of the main frequency energy in this section of the pipeline, is the kurtosis of the waveform in this section, is the frequency band mutual information entropy, is the spatial coordinate center of the region. These data are used as the input of the clustering algorithm, reflecting the fusion of spatial density and hydraulic signals.

[0072] A custom distance metric function D is proposed in the algorithm:

[0073] ; Indicates the 、 The difference in hydraulic characteristics between two areas is the combined difference in the following quantities:

[0074] ;

[0075] is the spatial distance, using Euclidean distance:

[0076] ;

[0077] Weight coefficient ,allowing to adjust the degree of hydraulic and spatial influence.,Through this distance function, DBSCAN not only considers spatial distance when determining which points are “neighboring”, but also,introduces the hydraulic difference dimension under the influence of the flock distribution,,achieving dual-constrained clustering of spatial density thermal features and hydraulic frequency,features.

[0078] The further scheme for constructing the spatiotemporal distribution map is as follows:

[0079] 1. Sensor data acquisition and fusion:

[0080] Infrared sensor array (IR): Installed on the ceiling of the chicken house, it collects 2D thermal imaging data to detect temperature hotspots and reflect the distribution of chickens. The accuracy is approximately ±0.5°C and can distinguish between heat source concentrations of individuals or small groups.

[0081] RFID positioning beacons: Chickens wear active RFID tags, and receiving anchor points (analyzing distance + angle) are arranged on the ceiling or fence to provide 2D or 3D coordinate information of each chicken with a resolution of about 10cm.

[0082] ToF camera: Captures depth maps of the chickens, fills in blind spots caused by spatial occlusion, and outputs real-time point cloud data of the chicken house, including Z-axis depth information.

[0083] The three types of data are synchronized and aligned according to timestamps, and the extended Kalman filter (EKF) is used to fuse the heat source center point, RFID positioning point, and ToF spatial contour point to obtain a three-dimensional coordinate point cloud of the chicken flock with millimeter-level accuracy that is updated every 5 seconds.

[0084] 2. Construction of Chicken Density Heat Map:

[0085] Spatial grid division:

[0086] 1. Spatial grid division: Divide the chicken house plane into regular grids (0.5m×0.5m), and each grid unit is recorded as , contains a point set , indicating the The three-dimensional positioning coordinates of each chicken, Indicates the first Rank The spatial grid cells of the columns;

[0087] 2. Density calculation: Count the number of chickens that appear in each grid per unit time , using Gaussian kernel density estimation to generate a smooth density map: , Indicates at time Grid Cell The number of chickens in Represents a grid cell At the moment The estimated density value of Represents a grid cell The center coordinates of Represents the bandwidth parameter of the Gaussian kernel, which is used to control the smoothness of the density estimation. Represents the natural exponential function, which is used to construct the Gaussian kernel function weight;

[0088] 3. Visualize as a heat map: Mapped as color intensity, higher heat indicates denser flocks.

[0089] 3. Generation of chicken flock movement trajectory vector field:

[0090] 1. Track point pairing: for each chicken In continuous moments Anchor point:

[0091] ; Get its instantaneous velocity vector, It is The position of a chicken at time t, Indicates the A chicken at the moment location, It is The instantaneous velocity vector of each chicken at time t;

[0092] 2. Vector projection and aggregation: Project the velocity vectors of all chickens to the current grid, average or fit the velocity vectors in the same grid to generate the grid-level average velocity vector :

[0093] ;in, Represents a grid cell At the moment The average velocity vector, Indicates at time Falling into a grid cell The number of chickens, Represents chicken The instantaneous velocity vector of

[0094] 3. Form a vector field map: Use arrows to indicate the main direction and movement speed within each grid, and identify the flock's movement trends, avoidance paths, concentrated movement and other behavioral patterns.

[0095] S2 specifically includes:

[0096] S21, valid touch event detection mechanism: A three-axis acceleration chicken beak touch sensor is installed at the drinking water terminal of each control unit. The raw touch data is recorded with a sampling period of 200ms. A touch event is only considered a "valid touch" when the following two conditions are met simultaneously:

[0097] Acceleration conditions: And the duration is <300ms, where The maximum acceleration value of the touch event;

[0098] Voiceprint conditions: ,in The energy ratio of swallowing voiceprint in the 500-800Hz frequency band is collected synchronously by the microphone array, and the touch and acoustic signals are fused through the VGG-9 compression model, with an error rate of <3%.

[0099] S22, drinking behavior pattern matrix construction: every 5 minutes is a statistical time window, and the number of effective touch events accumulated by each control unit is counted to form a two-dimensional matrix , indicating the Unit in the Normalized touch intensity within a time window: ,in, Indicates the Unit in the The effective touch frequency in a time window, Indicates the mean and standard deviation of the touch frequency of the unit in the historical time window, For the The temperature increase rate within the time window, is the touch intensity after standardization and temperature correction, reflecting the actual tendency of drinking water demand, and the exponential decay term To suppress the falsely high touch frequency caused by high ambient temperature, the behavior matrix supports two extended dimensions:

[0100] Time expansion (vertical): use the moving average of the first three time windows ;

[0101] Spatial expansion (horizontal): considers the covariance coupling terms of adjacent units.

[0102] S33, compensation function construction: based on the current temperature and humidity environment, construct a water demand compensation function , used to adjust the water supply target value of each unit:

[0103] ;

[0104] in, is the base water supply (the default water supply per unit per cycle), is the current ambient temperature, Current ambient relative humidity, 、 It is the temperature and humidity benchmark value for a comfortable environment for laying hens. is the temperature change response coefficient, and its initial value is , is the humidity square response coefficient, the initial value is , Behavior influence gain coefficient, used to control the weight of drinking behavior in compensation, It is a hyperbolic tangent function, which is used to constrain the saturation effect of behavioral factors and prevent surge compensation. Temperature linearly affects drinking water demand, and humidity uses a nonlinear superposition model to reflect the inhibitory or enhancing effect in a high humidity environment. The function can prevent behavioral outliers from triggering a surge in water volume, which has physiological constraint implications.

[0105] S3 specifically includes:

[0106] S31, Construction of three-level response water supply model: Based on the matrix elements of drinking behavior pattern The value range of each control unit in the current time window The water supply methods are classified as follows:

[0107] S311, high-frequency pulse mode (high-intensity drinking response): ;

[0108] The microchannel water-saving valve operates at a fixed duty cycle: open for 0.5 seconds and closed for 1.0 seconds;

[0109] The single pulse flow rate is set to 120% of the baseline water supply value: ;

[0110] S312, Adaptive Pulse Mode (Normal Drinking Response): ;

[0111] Pulse water supply cycle Dynamic Adjustment: ;

[0112] In the adaptive pulse water supply mode, the water supply cycle is determined by the drinking behavior pattern matrix element Control, and the actual water supply per time is based on the water demand compensation function defined by S2 Real-time calculation drive enables the pulse execution process to simultaneously integrate behavioral intensity and environmental compensation parameters to achieve precise water-saving control;

[0113] S313, water-saving monitoring mode (low-intensity drinking water response): ;

[0114] The "probe-type water supply" is triggered every 3 minutes: each water supply lasts for 2 seconds. If no effective touch event is detected for 3 consecutive times, 50% of the drinking water terminals in the unit will be automatically shut down and enter a low-activity monitoring state.

[0115] S32, spatiotemporal coupling control of aeration device and water-saving valve:

[0116] S321, start delay matching hydraulic propagation time: after each pulse water supply is turned on, the Venturi aerator is started with a delay of 0.3 seconds; the aeration duration is nonlinearly positively correlated with the current pulse water supply flow rate:

[0117] ;in, is the duration of this aeration, is the aeration efficiency coefficient (obtained through historical training regression), is the current pulse water supply flow, The intensity of the current drinking behavior pattern.

[0118] S322, aeration enhancement mechanism: When the dissolved oxygen concentration in the control unit is monitored to decrease at a rate exceeding 0.2 mg / L·s, additional aeration pulses are inserted during the water supply interval, using variable-cycle PWM control, with the duty cycle adaptively adjusted according to the dissolved oxygen recovery rate.

[0119] S33, valve-aeration linkage protection mechanism: To ensure the safety of the hydraulic system, establish pulse cycle-level abnormality detection and response:

[0120] Anomaly detection dual conditions:

[0121] Time domain judgment: In 5 consecutive pulse cycles, if the pressure peak difference between adjacent pulses meets the following conditions:

[0122] , Indicates the The pressure peak difference between a pulse cycle and the previous cycle, Indicates the pipeline pressure peak value of the current cycle and the previous cycle;

[0123] Frequency domain judgment: Perform spectrum analysis on the pressure signal through Fourier transform (FFT). If the energy proportion of high-frequency components (>500Hz) suddenly increases by more than 30%, the following is the result:

[0124] , Indicates the The energy value of the pressure signal in the frequency range > 500Hz during the cycle, Indicates the The energy value of the pressure signal in the frequency range > 500 Hz during the cycle, Indicates the The total energy value of the pressure signal in the cycle, Indicates the The total energy value of the pressure signal in the cycle;

[0125] If any of the above conditions are met, the following protection mechanism will be triggered:

[0126] Immediately switch to constant flow mode (cancel pulse control);

[0127] At the same time, the water supply branch pressure will be reduced by 20%.

[0128] The present invention encompasses any alternatives, modifications, equivalents, and solutions that fall within the spirit and scope of the present invention. To provide a thorough understanding of the present invention, specific details are described in detail below in connection with the preferred embodiments of the present invention, but those skilled in the art will be able to fully understand the present invention without these detailed descriptions. Furthermore, to avoid unnecessary confusion regarding the essence of the present invention, well-known methods, processes, procedures, components, and circuits have not been described in detail.

[0129] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. An intelligent drinking water scheduling and water-saving control method for laying hens integrating multi-source sensing is characterized by: The following steps are involved: S1: Using a positioning sensor network to obtain a real-time spatiotemporal distribution map of the chicken flock, a partition grid is constructed based on the drinking line pressure pulsation data. Based on the partition grid, the chicken house is divided into multiple control units with independent hydraulic characteristics. S2: Generate a drinking behavior pattern matrix based on the trigger frequency sequence of the chicken beak touch sensor in each control unit. At the same time, integrate the ambient temperature and humidity gradient data to establish the water demand compensation function for each zone. Specifically, it includes: S21: Install a chicken beak touch sensor at the drinking water terminal of each control unit and set a sampling period to record valid touch events. The valid touch events must meet both acceleration conditions and voiceprint conditions. S22: Count the effective touch frequency sequence of each control unit according to the predetermined time window to construct a drinking behavior pattern matrix with extended time and space dimensions. , the matrix elements represents the normalized touch intensity of the i-th unit in the j-th time window; S23, based on the current temperature and humidity environment, build a water demand compensation function , used to adjust the water supply target value of each control unit; The water demand compensation function Expressed as: ； in, is the baseline water supply, is the current ambient temperature, is the current ambient relative humidity, It is the temperature and humidity benchmark value for a comfortable environment for laying hens. is the temperature change response coefficient, is the humidity square response coefficient, Behavior influence gain coefficient, used to control the weight of drinking behavior in compensation, It is a hyperbolic tangent function, which is used to constrain the saturation effect of the behavioral factor and prevent surge compensation; S3: According to the drinking water behavior pattern matrix and water demand compensation function output by S2, the microchannel water-saving valve corresponding to each partition is driven to execute a differentiated water supply mode, and the opening and closing timing of the aeration device of the partition is synchronously adjusted.

2. The method for intelligent drinking water scheduling and water-saving control of a laying hen house integrating multi-source sensing according to claim 1 is characterized in that: The positioning sensor network in S1 includes an infrared sensor array, an RFID positioning beacon and a ToF camera. The infrared sensor array, RFID positioning beacon and ToF camera arranged on the ceiling of the chicken house generate three-dimensional coordinate point cloud data of the chicken flock, and the spatiotemporal distribution map is updated at a predetermined period. The spatiotemporal distribution map includes a heat map of chicken density in each area and a movement trajectory vector.

3. The method for intelligent drinking water scheduling and water-saving control of a laying hen house integrating multi-source sensing according to claim 2 is characterized in that: Said S1 also includes installing a high-frequency pressure transmitter in the main drinking water pipe, collecting pressure pulsation data, and extracting characteristic frequency components representing the topological structure of the pipe network through wavelet packet decomposition; The heat map of chicken density and the characteristic frequency components of the pipe network topology are input into the DBSCAN clustering algorithm to generate control units that meet the following constraints: Constraint 1: The coefficient of variation in chicken density within the control unit does not exceed a predetermined percentage; Constraint 2: The correlation coefficient of pipeline pressure fluctuations between adjacent control units is less than the predetermined coefficient threshold.

4. The method for intelligent drinking water scheduling and water-saving control of a laying hen house integrating multi-source sensing according to claim 3 is characterized in that: S1 also includes a partition grid reconstruction mechanism: when any trigger condition is met, the partition grid is reconstructed: Trigger condition 1: The chicken population density change rate in any sub-area exceeds 10% / minute; Trigger condition two: The pressure pulsation main frequency offset is greater than the offset threshold.

5. The method for intelligent drinking water scheduling and water-saving control of a laying hen house integrating multi-source sensing according to claim 1 is characterized in that: The acceleration condition is expressed as: And the duration is <300ms, where The maximum acceleration value of the touch event; The voiceprint condition is expressed as: ,in The energy of the swallowing soundprint in the 500-800Hz frequency band, It is the total energy of the entire voiceprint signal in the full frequency band.

6. The method for intelligent drinking water scheduling and water-saving control of a laying hen house integrating multi-source sensing according to claim 1 is characterized in that: In S3, according to the drinking behavior pattern matrix of each control unit Normalized touch strength in , determine the control mode level, including: like If the high threshold is exceeded, the high-frequency pulse mode is activated, and the microchannel water-saving valve is set to a high-duty-cycle short-cycle action, and the single water supply flow is increased to the upper limit of the reference value; like If it is in the middle range, it enters the adaptive pulse mode, determines the water-saving valve switch cycle through the logarithmic function, and uses the water demand compensation amount calculated in the early stage as the current pulse water supply benchmark; like If the water level is lower than the low threshold, the water-saving monitoring mode is enabled, and water is only supplied tentatively for a short period of time within a predetermined period, and whether to shut down some drinking water terminals is determined by effective touch.

7. The method for intelligent drinking water scheduling and water-saving control of a laying hen house integrating multi-source sensing according to claim 1 is characterized in that: Said S3 also includes the timing coupling control of the aeration device and the micro-channel water-saving valve, specifically including: After the micro-channel water-saving valve enters the open state, the Venturi aerator is started with a delay. The delay value is used to match the water flow propagation time to ensure synchronous diffusion of gas and liquid. The aeration duration is positively correlated with the current pulse flow rate. When it is detected that the dissolved oxygen concentration decrease rate of the control unit exceeds the set decrease threshold, an additional aeration process is inserted during the water supply interval.

8. The method for intelligent drinking water scheduling and water-saving control of a laying hen house integrating multi-source sensing according to claim 1 is characterized in that: The method also includes establishing a valve-aeration linkage protection mechanism. If abnormal fluctuations in pipeline pressure are detected in multiple consecutive pulse cycles, or the energy proportion of the pressure signal in the high-frequency band suddenly increases and exceeds the judgment threshold, it will automatically switch to a constant flow mode.

Images