Laying hen house intelligent drinking water scheduling and water-saving control method integrated with multi-source sensing

Abstract

The invention relates to the technical field of water consumption regulation and control, in particular to a laying hen house intelligent drinking water scheduling and water-saving control method integrated with multi-source sensing, which comprises the following steps of: acquiring a chicken flock space-time distribution map in real time through a positioning sensing network, and constructing a partition grid in combination with drinking water line pressure pulsation data, dividing the chicken house into a plurality of regulation and control units with independent hydraulic characteristics based on the partition grids; based on the triggering frequency sequence of the chicken beak touch sensor in each regulation and control unit, generating a water drinking behavior pattern matrix, and meanwhile, combining environment temperature and humidity gradient data to establish a water demand compensation function of each subarea; and according to the water drinking behavior mode matrix and the water demand compensation function, driving the micro-channel water-saving valve corresponding to each subarea to execute a differential water supply mode, and synchronously adjusting the on-off time sequence of the aeration device of the subarea. According to the invention, the real-time dynamic regulation of the flow is realized, the physiological water demand and the system water-saving target are effectively balanced, and the intelligence and robustness of the drinking water control strategy are obviously improved.

Description

Technical Field

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

[0002] Currently, in the process of large-scale laying hen farming, as an important infrastructure to ensure the life activities and production performance of poultry, the regulation strategies of the drinking water system generally have problems such as lagging response, rough regulation, and low water-saving efficiency. Traditional drinking water control methods are mostly based on timed water supply or fixed flow supply, lacking the ability to perceive the actual drinking behavior of the chicken flock and environmental changes in real time and dynamically adjust, which easily leads to over-supply of water, waste of water resources, or affects the health and egg production rate of poultry due to insufficient water supply.

[0003] Some water-saving controls have introduced local behavior detection devices, such as touch sensors, infrared monitoring, etc., to roughly evaluate the drinking activity, but usually fail to construct a high-resolution spatio-temporal behavior map for zoning control, nor can they be synergistically linked with environmental factors such as temperature and humidity. At the same time, the hydraulic state (such as pressure fluctuation, flow distribution) during the water supply process in the chicken house pipe network is not included in the regulation strategy, resulting in the regulation unit being unable to adaptively adjust the flow structure or respond to peak behavior, thereby increasing the risk of unstable operation.

[0004] In addition, in the current system, the drinking water device and the aeration system are generally controlled separately, lacking a gas-liquid cooperation mechanism and unable to cope with problems such as local water quality deterioration or biofilm accumulation. Under the condition of high-frequency pulsed water supply, without an effective hydraulic safety protection strategy, it is also easy to induce faults such as water hammer effect and pipeline oscillation, further restricting the deployment and expansion of the intelligent drinking water system. Summary of the Invention

[0005] The present invention provides an intelligent drinking water scheduling and water-saving control method for laying hen houses integrating multi-source sensing, which integrates multi-source sensing information, can dynamically perceive the drinking behavior characteristics of the chicken flock, realizes water demand compensation in combination with environmental parameters, and has an 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, self-adaptability, and operation 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: S1: Real-time obtain the spatio-temporal distribution map of the chicken flock through the positioning sensing network, construct a zoning grid in combination with the pressure pulsation data of the drinking water line, and divide the chicken house into multiple regulation units with independent hydraulic characteristics based on the zoning grid; S2: Generate a drinking behavior pattern matrix based on the trigger frequency sequence of the chicken beak touch sensors in each regulation unit, and establish a water demand compensation function for each zone by integrating the environmental temperature and humidity gradient data; S3: According to the drinking behavior pattern matrix output by S2 and the water demand compensation function, drive the micro-channel water-saving valves corresponding to each partition to execute a differential water supply mode, and synchronously adjust the opening and closing timing of the aeration device in this partition.

[0007] Optionally, the positioning and sensing network in S1 includes an infrared sensor array, RFID positioning beacons, and a ToF camera. Generate three-dimensional coordinate point cloud data of the chicken flock through the infrared sensor array, RFID positioning beacons, and ToF camera arranged on the chicken coop ceiling, and update and generate a spatio-temporal distribution map at a predetermined period. The spatio-temporal distribution map includes a heat map of chicken flock density in each area and a moving trajectory vector.

[0008] Optionally, S1 further includes installing a high-frequency pressure transmitter on the main drinking water pipeline to collect pressure pulsation data, and extracting characteristic frequency components representing the pipe network topology structure through wavelet packet decomposition; Input the chicken flock density heat map and the pipe network characteristic frequency components into the DBSCAN clustering algorithm to generate control units (i.e., partition grids) that meet the following constraints: Constraint 1: The coefficient of variation of chicken flock 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 a predetermined coefficient threshold.

[0009] Optionally, S1 further includes a partition grid reconstruction mechanism: when any trigger condition is met, reconstruct the partition grid: Trigger condition 1: The change rate of chicken flock density in any sub-region exceeds 10% / minute; Trigger condition 2: The offset of the main frequency of pressure pulsation is greater than the offset threshold.

[0010] Optionally, S2 specifically includes: S21, install a chicken beak touch sensor at the drinking water terminal of each control unit, set a sampling period to record valid touch events, and the valid touch events need to meet both the acceleration condition and the voiceprint condition; S22, statistically count the sequence of valid touch frequencies of each control unit according to a predetermined time window, and construct a drinking behavior pattern matrix with extended spatio-temporal dimensions , and the matrix element represents the normalized touch intensity of the i-th unit in the j-th time window; S23, construct a water demand compensation function based on the current temperature and humidity environment, for adjusting the water supply target value of each control unit.

[0011] Optionally, the acceleration condition is expressed as: and the duration < 300 ms, where is the maximum acceleration value of the touch event; The voiceprint condition is expressed as: , where is the energy proportion of the swallowing voiceprint in the frequency band of 500 - 800 Hz, is the total energy of the full frequency band of the entire voiceprint signal, that is, the sum of the energy of the sound signal within the entire recording time window (for example, 0 - 4 kHz).

[0012] Optionally, the water requirement compensation function is expressed as: ; where, is the reference water supply volume, is the current ambient temperature, is the current ambient relative humidity, , are the reference values of temperature and humidity in the comfortable environment of laying hens, is the temperature change response coefficient, is the humidity square response coefficient, is the behavior influence gain coefficient, which is used to control the weight of the drinking behavior in the compensation, is the hyperbolic tangent function, which is used to constrain the saturation influence of the behavior factor and prevent surge - type compensation.

[0013] Optionally, in S3, according to the standardized touch intensity in the drinking behavior pattern matrix of each regulation unit, the control mode level to which it belongs is determined, specifically including: If exceeds the high threshold, the high - frequency pulse mode is activated, and the micro - channel water - saving valve is set to operate with a high duty cycle and a short period, and the single - time water supply flow rate is increased to the upper limit of the increase amplitude of the reference value; If is in the medium range, the adaptive pulse mode is entered, the switch cycle of the water - saving valve is determined through a logarithmic function, and the water requirement compensation amount calculated previously is used as the current pulse water supply reference; If is lower than the low threshold, the water - saving monitoring mode is enabled, and water supply is only tentatively carried out for a short time within a predetermined period, and whether to close some drinking terminals is determined through effective touch judgment.

[0014] Optionally, S3 further includes the coupling control of the execution timing 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 the synchronous diffusion of gas and liquid, and the aeration duration is positively correlated with the current pulse flow rate; When it is detected that the rate of decrease in the dissolved oxygen concentration of the control unit exceeds the set decrease threshold, an additional aeration process is inserted during the water supply intermittent period.

[0015] Optionally, the method further includes establishing a valve-aeration linkage protection mechanism. If abnormal fluctuations in the pipeline pressure are detected in multiple consecutive pulse periods, or the energy ratio of the pressure signal in the high-frequency band suddenly increases beyond the determination threshold, it automatically switches to the constant flow mode.

[0016] Advantages of the present invention: 1. In the present invention, by integrating multi-source spatial perception means such as infrared arrays, RFID positioning, and ToF cameras, a density heat map and a trajectory vector field are constructed. Combining with acceleration and acoustic fingerprint sensors, effective drinking events are extracted based on dual-modal criteria of behavior and acoustics, and a spatio-temporal extended drinking behavior pattern matrix is constructed. It is classified into a three-level response mode according to the behavior intensity, and combined with logarithmic period adjustment and water-saving detection strategies, to achieve fine-grained hierarchical control of the control unit under different drinking states, and improve the response accuracy and physiological adaptability of the system to the dynamic changes in drinking water demand.

[0017] 2. In the present invention, a temperature and humidity non-linear compensation function is introduced. With the real-time ambient temperature, relative humidity, and drinking behavior intensity as parameters, a dynamic water demand adjustment model is constructed, and the influence of extreme behaviors is restricted by the hyperbolic tangent function to suppress the misjudgment risk of abnormal amplification of drinking behaviors in high-temperature or high-humidity scenarios. The compensation function is embedded in the adaptive pulse water supply mode to achieve real-time dynamic adjustment of the flow rate, effectively balance the physiological water demand and the system water-saving goal, and significantly improve the intelligence and robustness of the drinking water control strategy.

[0018] 3. In the present invention, through an aeration delay start mechanism based on the water flow propagation delay, gas-liquid synchronous diffusion is achieved, and the aeration duration is adjusted according to the current pulse intensity and behavior deviation. The dissolved oxygen change rate is integrated to drive the aeration compensation during the intermittent period, improving the drinking water quality and the adaptability of the device. A pressure fluctuation dual-domain determination mechanism is proposed. Based on the change in the pressure peak value of the pulse period and the mutation of the high-frequency energy of the spectrum, the linkage protection program and the constant flow switching logic are triggered in real time, and the branch pressure is automatically reduced to prevent water hammer impact and pipeline network damage, enhancing the reliability and operation safety of the system. Description of the drawings

[0019] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings in the following description are only those of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0020] Figure 1 It is a schematic flow chart of the control method for the embodiments of the present invention; Figure 2 Schematic diagram of the differential water supply mode according to an embodiment of the present invention. Specific embodiments

[0021] The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments. For some well-known technologies, those skilled in the art can also adopt other alternative methods for implementation; moreover, the accompanying drawings are only for more specifically describing the embodiments and are not intended to specifically limit the present invention.

[0022] As Figure 1 - Figure 2 shown, the intelligent drinking water scheduling and water-saving control method for laying hen houses integrating multi-source sensing includes the following steps: S1: Obtain the spatio-temporal distribution map of the chicken flock in real time through the positioning sensing network, construct a partition grid in combination with the pressure pulsation data of the drinking line, and divide the chicken house into multiple regulation units with independent hydraulic characteristics based on the partition grid; Each regulation unit includes: 1 set of controllable micro-channel water-saving valves; 1 aeration device node; Multiple drinking terminals; S2: Generate a drinking behavior pattern matrix based on the trigger frequency sequence of the chicken beak touch sensors in each regulation unit, and establish a water demand compensation function for each partition by integrating the environmental temperature and humidity gradient data; S3: According to the drinking behavior pattern matrix and the water demand compensation function output by S2, drive the micro-channel water-saving valves corresponding to each partition to execute a differential water supply mode, and synchronously adjust the opening and closing time sequence of the aeration device in this partition.

[0023] The positioning sensing network includes an infrared sensor array, an RFID positioning beacon, and a ToF (Time of Flight) camera. S1 specifically includes: S11, Spatio-temporal distribution map construction: Generate the three-dimensional coordinate point cloud data of the chicken flock through the infrared sensor array, RFID positioning beacon, and ToF (Time of Flight) camera arranged on the chicken house ceiling, and update it with a period of 5 seconds to obtain the spatio-temporal distribution map of the chicken flock, which includes: Chicken flock density heat map; Moving trajectory vector field.

[0024] The fusion method is: Use Kalman filtering to fuse the infrared temperature hot spot data (accuracy ±0.5°C), RFID positioning data (resolution 10 cm), and ToF depth image to generate a dynamic distribution map.

[0025] S12, Hydraulic Feature Signal Acquisition and Processing: Install a high-frequency pressure transmitter on the main drinking water pipeline, set the sampling rate to 20–100 Hz, collect the original pressure pulsation signal, and extract the following key frequency feature components characterizing the pipe network topology through wavelet packet decomposition: Proportion of energy in the main frequency band : Used to identify the impedance characteristics of the pipe section; Waveform kurtosis coefficient : Characterize the local turbulence intensity; Band mutual information entropy : Judge the connectivity status of the pipe network.

[0026] S13, Regulation Unit Clustering and Partitioning: Input the chicken flock density heat map and the hydraulic feature frequency components into the DBSCAN clustering algorithm together. This algorithm generates regulation units that meet the following constraint conditions: The difference coefficient of chicken flock density within the unit does not exceed 15%; The correlation coefficient of pipeline pressure fluctuations between adjacent units is less than 0.3; The core mechanism of DBSCAN is as follows: S131, The distance metric function is defined as: ; where, is the pressure feature difference between nodes, is the Euclidean space distance between nodes, represents the weight coefficient.

[0027] S132, Adaptive neighborhood radius : ; where, is the adjustment factor, is the impedance of the current pipe section; S133, Dynamic adjustment rule for the minimum number of clustering points: ; where, is the total number of chickens in the current chicken coop, is the number of partitions that have been divided currently.

[0028] S14, Reconstruction Mechanism of the Partition Grid: When any of the following trigger conditions is met, reconstruct the dynamic partition grid: S141. The change rate of chicken flock density in any sub-region satisfies: ; where, is the change in the number of chickens per unit time, is the area of this region, is the sampling interval time.

[0029] S142. The main frequency shift of the pressure pulsation 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 offset. During the reconstruction process, 30% of the original partition structure is retained as the topological skeleton to ensure control continuity and response smoothness.

[0030] In the traditional DBSCAN algorithm, clustering is usually based on pure spatial coordinates or a certain location information. In this solution, the clustering input is extended to a high-dimensional feature vector containing multi-modal features. Each point to be clustered represents a sampling unit grid area in the chicken coop space, and its input features include: ; where is the flock density of the th area, is the density gradient, representing the change trend of the flock distribution, is the proportion of the main frequency energy of this section of the pipeline, is the waveform kurtosis of this section, is the mutual information entropy of the frequency band, is the spatial coordinate center of the area. All these data are used as the input of the clustering algorithm, reflecting the fusion of spatial density and hydraulic signals.

[0031] A custom distance metric function D is proposed in the algorithm: ; represents the hydraulic feature difference between the st and nd areas, which is a combined difference of the following quantities: ; is the spatial distance, using the Euclidean distance: ; The weight coefficient , allowing adjustment of the influence degree of hydraulics / spatial. Through this distance function, when DBSCAN determines which points are "adjacent", it not only refers to the spatial distance but also introduces the dimension of hydraulic difference under the influence of flock distribution, realizing the dual-constraint clustering of spatial density thermal characteristics and hydraulic frequency characteristics.

[0032] The further solution for constructing the spatio-temporal distribution map is as follows: I. Sensing data acquisition and fusion: Infrared sensor array (IR): Installed on the chicken coop ceiling, it collects 2D thermal imaging data for detecting temperature hotspots, reflecting the chicken distribution. The accuracy is about ±0.5°C, and it can distinguish the heat source concentration areas of individuals or small groups.

[0033] RFID positioning beacon: Active RFID tags are worn by chickens, and receiving anchors (resolution distance + angle) are arranged on the ceiling or fence to provide 2D or 3D coordinate information of each chicken, with a resolution of approximately 10 cm.

[0034] ToF camera: Captures the depth map of chickens, compensates for the blind spots under spatial occlusion, and outputs real-time point cloud data in the chicken coop, including Z-axis depth information.

[0035] Synchronize and align the three types of data according to the timestamp, and use the Extended Kalman Filter (EKF) to fuse the heat source center point, RFID positioning point, and ToF space contour point to obtain a three-dimensional coordinate point cloud of the chicken flock with millimeter-level accuracy updated every 5 seconds.

[0036] II. Construction of chicken flock density heat map: Spatial grid division: 1. Spatial grid division: Divide the chicken coop plane into regular grids (0.5m × 0.5m), and each grid cell is denoted as , containing a point set , representing the three-dimensional positioning coordinates of the th chicken, indicating the rd row and th column of the spatial grid cell after the chicken coop plane is divided;

[0037] 2. Density calculation: Count the number of chickens appearing in each grid within a unit time , and use Gaussian kernel density estimation to generate a smooth density map: , represents the number of chickens in the grid cell at time , represents the estimated density value of the grid cell at time , represents the central coordinates of the grid cell , represents the bandwidth parameter of the Gaussian kernel, used to control the smoothness of density estimation, represents the natural exponential function, used to construct the Gaussian kernel function weight; 3. Visualize as a heat map: Map to color intensity, and the higher the heat, the denser the chicken flock.

[0038] III. Generation of chicken flock movement trajectory vector field: 1. Trajectory point pairing: For each chicken at consecutive times of the positioning points: ; Obtain its instantaneous velocity vector, is the position of the th chicken at time t, represents the th chicken at time position; is the instantaneous velocity vector of the th chicken at time t; 2. Vector projection and aggregation: Project the velocity vectors of all chickens onto the current grid, average or fit the main direction of the velocity vectors within the same grid to generate a grid-level average velocity vector : ; Among them, represents the average velocity vector of grid cell at time , represents the number of chickens that fall into grid cell at time , represents the instantaneous velocity vector of chicken ; 3. Form a vector field diagram: Use arrows to represent the main direction and moving speed within each grid to identify behavior patterns such as the movement trend of the chicken flock, avoidance paths, and concentrated movement.

[0039] S2 specifically includes: S21, effective touch event detection mechanism: Install a three-axis acceleration chicken beak touch sensor at the drinking water terminal of each regulation unit, record the original touch data with a sampling period of 200 ms, and a touch event is determined to be "effective" only when the following two conditions are simultaneously met: Acceleration condition: and the duration < 300 ms, where is the maximum acceleration value of the touch event; Voiceprint condition: , where is the energy ratio of the swallowing voiceprint in the 500 - 800 Hz frequency band. The voiceprint data is synchronously collected by a microphone array, and the touch and acoustic signals are fused through a VGG-9 compression model, with a misjudgment rate < 3%.

[0040] S22, construction of the drinking behavior pattern matrix: Use every 5 minutes as a statistical time window, count the cumulative number of effective touch events for each regulation unit to form a two-dimensional matrix , representing the normalized touch intensity of the rd unit in the th time window: , among which, represents the rd unit in the The effective touch frequency of a time window represents the mean and standard deviation of the touch frequency of the unit under the historical time window is the temperature rise rate within the th time window, and is the touch intensity after standardization and temperature correction, reflecting the actual drinking water demand tendency. The exponential decay term is used to suppress the inflated touch frequency caused by high environmental temperature. The behavior matrix supports two extended dimensions: ; Spatial expansion (lateral): Consider the covariance coupling term of adjacent units.

[0041] 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: ; Among them, is the benchmark water supply (the default water supply per unit per cycle), is the current environmental temperature, the current environmental relative humidity, , are the benchmark values of temperature and humidity in the comfortable environment of laying hens, is the temperature change response coefficient, and the initial value is , is the humidity square response coefficient, and the initial value is , is the behavior influence gain coefficient, used to control the weight of drinking behavior in compensation, is the hyperbolic tangent function, used to constrain the saturation influence of behavior factors and prevent surge-like compensation. Temperature linearly affects drinking water demand, and humidity uses a non-linear superposition model to reflect the inhibitory or enhancing effect in a high-humidity environment; introducing function can prevent abnormal behavior values from causing a surge in water volume and has the meaning of physiological constraint.

[0042] S3 specifically includes: S31, Construction of three-level response water supply mode: According to the numerical range of the elements of the drinking behavior pattern matrix , classify the water supply methods of each regulation unit in the current time window as follows: S311, High-frequency pulse mode (high-intensity drinking water response): ; The microchannel water-saving valve works at a fixed duty cycle: it opens for 0.5 seconds and closes for 1.0 second; The single-pulse flow rate is set to 120% of the reference water supply value: ; S312, Adaptive pulse mode (normal drinking water response): ; Pulse water supply period Dynamically adjusted: ; In the adaptive pulse water supply mode, the water supply period is controlled by the elements of the drinking behavior pattern matrix while the actual water supply volume per time is driven by real-time calculation according to the water demand compensation function defined in S2 to make the pulse execution process integrate the behavior intensity and environmental compensation parameters at the same time, realizing precise water-saving control; S313, Water-saving monitoring mode (low-intensity drinking water response): ; "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 of this unit will be automatically shut down and enter the low-activity monitoring state.

[0043] S32, Spatiotemporal coupling control of the aeration device and the water-saving valve: S321, Start-up delay matching the hydraulic propagation time: The Venturi aerator is started 0.3 seconds after each pulse water supply is turned on; The aeration duration is non-linearly positively correlated with the current pulse water supply flow rate: ; where is the current aeration duration, is the aeration efficiency coefficient (obtained through historical training regression), is the current pulse water supply flow rate, is the current drinking behavior pattern intensity.

[0044] S322, Aeration enhancement mechanism: When it is detected that the dissolved oxygen concentration decline rate in the regulation unit exceeds 0.2 mg / L·s, additional aeration pulses are inserted during the water supply interval, and variable-cycle PWM control is adopted, and its duty cycle is adaptively adjusted according to the dissolved oxygen recovery rate.

[0045] S33, Valve-aeration linkage protection mechanism: To ensure the safety of the hydraulic system, anomaly detection and response at the pulse period level are established: Double conditions for anomaly detection: Time-domain determination: In 5 consecutive pulse cycles, if the pressure peak difference between adjacent pulses satisfies: , represents the pressure peak difference between the nth pulse period and the previous period, and represents the pipeline pressure peak of the current period and the previous period; Frequency domain determination: Perform spectral analysis on the pressure signal through Fourier transform FFT. If the energy ratio of the high-frequency component (>500 Hz) suddenly increases by more than 30%: , represents the energy value of the pressure signal in the frequency range >500 Hz in the nth period, represents the energy value of the pressure signal in the frequency range >500 Hz in the nth period, represents the total energy value of the pressure signal in the nth period, represents the total energy value of the pressure signal in the nth period; If any of the above conditions is satisfied, trigger the following protection mechanism: Immediately switch to the constant flow mode (cancel pulse control); At the same time, reduce the pressure of the water supply branch by 20%.

[0046] The present invention covers any alternatives, modifications, equivalent methods, and solutions made within the spirit and scope of the present invention. For the public to have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention. However, those skilled in the art can fully understand the present invention without these detailed descriptions. Additionally, well-known methods, processes, procedures, components, and circuits are not described in detail to avoid unnecessary confusion to the essence of the present invention.

[0047] The above description is only a preferred embodiment of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. Intelligent drinking water scheduling and water-saving control method for laying hen houses integrating multi-source sensing, characterized in that, It includes the following steps: S1: Obtain the spatio-temporal distribution map of the chicken flock in real time through a positioning sensor network, construct a partition grid by combining the pressure pulsation data of the drinking water line, and divide the chicken house into multiple regulation units with independent hydraulic characteristics based on the partition grid; S2: Generate a drinking behavior pattern matrix based on the trigger frequency sequence of the chicken beak touch sensors in each regulation unit, and establish a water demand compensation function for each partition by fusing the environmental temperature and humidity gradient data; S3: According to the drinking behavior pattern matrix and the water demand compensation function output by S2, drive the microchannel water-saving valves corresponding to each partition to execute a differential water supply mode, and synchronously adjust the opening and closing timing of the aeration device in this partition.

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

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

4. The intelligent drinking water scheduling and water-saving control method for laying hen houses integrating multi-source sensing according to claim 3, characterized in that S1 also includes a partition grid reconstruction mechanism: when any trigger condition is met, reconstruct the partition grid: Trigger condition 1: The change rate of the chicken flock density in any sub-region exceeds 10% / minute; Trigger condition 2: The offset of the main frequency of the pressure pulsation is greater than the offset threshold.

5. The intelligent drinking water scheduling and water-saving control method for laying hen houses integrating multi-source sensing according to claim 1, characterized in that, S2 specifically includes: S21, install a chicken beak touch sensor at the drinking water terminal of each regulation unit, set the sampling period to record effective touch events, and the effective touch events need to meet both the acceleration condition and the voiceprint condition; S22. Statistically analyze the effective touch frequency sequences of each regulation unit according to a predetermined time window, and construct a drinking behavior pattern matrix with extended spatio-temporal dimensions. , where the matrix element represents the normalized touch intensity of the i-th unit within the j-th time window. S23. Construct a water demand compensation function based on the current temperature and humidity environment , which is used to adjust the water supply target values of each control unit.

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

7. The intelligent drinking water scheduling and water-saving control method for laying hen houses integrating multi-source sensing according to claim 6, characterized in that, The water demand compensation function is expressed as: ； Among them, is the reference water supply volume, is the current ambient temperature, is the current relative humidity of the environment, 、 are the temperature and humidity reference values in the comfortable environment of laying hens, is the temperature change response coefficient, is the humidity square response coefficient, The behavior influence gain coefficient is used to control the weight of the drinking behavior in the compensation, is the hyperbolic tangent function, which is used to constrain the saturation influence of the behavior factor and prevent surge compensation.

8. The intelligent drinking water scheduling and water-saving control method for laying hen houses integrating multi-source sensing according to claim 5, characterized in that, In S3, according to the standardized touch intensity in the drinking behavior pattern matrix of each regulation unit , determine the level of the control mode to which it belongs, specifically including: ​ If exceeds the high threshold, the high-frequency pulse mode is activated, and the microchannel water-saving valve is set to operate with a high duty cycle and a short period, and the increase in the single-supply water flow is increased to the upper limit of the increase relative to the reference value; If it is in the medium range, it enters the adaptive pulse mode, determines the water-saving valve switching period through a logarithmic function, and uses the water demand compensation amount calculated previously as the current pulse water supply benchmark; If below the low threshold, the water-saving monitoring mode is enabled, and water supply is tentatively provided only for a short time within a predetermined period, and it is determined whether to close some drinking terminals by effective touch control.

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

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

Images