Ecological treatment method for livestock and poultry breeding waste

Abstract

The invention discloses an ecological treatment method for livestock and poultry breeding waste, and relates to the technical field of excretion waste treatment.The ecological treatment method comprises the steps that a diversion trench and a selective permeation grid device are arranged in a livestock and poultry house, the ground is divided into regular grid units, and accurate recognition of urination and defecation behaviors is achieved in combination with the acoustic and vibration monitoring technology; the flow resistance coefficient R and the urine flushing self-cleaning effect index Cclean of each grid unit are constructed, the flow guide qualification and the local blockage condition are dynamically judged, and negative pressure suction or flow division intervention can be triggered in a targeted mode; meanwhile, a long-term focused urine spot is recognized based on a urination position hot spot index HPI, a stable optimized urine spot is formed through the synergistic effect of smell induction, sound and light guide and the self-cleaning coating, the animals are guided to urinate to an area with qualified self-cleaning capacity, centralized collection and diversion of urine are achieved, the urine discharge uniformity and the waste self-cleaning efficiency are improved, and the method is suitable for large-scale popularization and application. The ecological treatment and resource utilization of the excrement are promoted.

Description

Technical Field

[0001] This invention relates to the field of waste treatment technology, specifically to an ecological treatment method for livestock and poultry breeding waste. Background Technology

[0002] Currently, while the livestock and poultry farming industry is developing rapidly, the problem of livestock and poultry manure waste disposal is becoming increasingly prominent. In traditional farming models, urination and defecation usually occur simultaneously in the same area of ​​the shed. Due to the lack of effective diversion and timely cleaning, urine and feces easily mix on the ground or in drainage ditches, leading to the rapid generation of harmful gases such as ammonia and hydrogen sulfide. This not only deteriorates the air quality inside the shed, affecting the health and growth performance of livestock and poultry, but also exacerbates the risk of nitrogen and phosphorus emissions and environmental pollution from farms. In addition, existing manure collection methods mostly rely on timed manual flushing or negative pressure suction, which has low cleaning efficiency and fails to provide targeted management based on the actual behavioral characteristics of animals, often resulting in incomplete cleaning, local blockages, or secondary pollution.

[0003] On the other hand, livestock and poultry exhibit certain urination and defecation behaviors, often establishing "preferred defecation points" within their enclosures. However, current technologies do not fully utilize this behavioral pattern for ecological guidance, leading to uneven waste distribution, insufficient self-cleaning effects, and long-term urine accumulation in localized areas, resulting in severe "ammonia hot spots," accelerating the corrosion and aging of floors and facilities, and increasing manual maintenance costs. While existing research has attempted to improve the environment using high-frequency flushing or overall ventilation, these methods suffer from limitations such as high energy consumption, unstable effectiveness, and disconnect from animal behavior. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the present invention provides an ecological treatment method for livestock and poultry breeding waste to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solution: an ecological treatment method for livestock and poultry breeding waste, comprising the following steps: S1, arranging a diversion channel and a selective infiltration grid device in the livestock and poultry house, dividing the diversion channel and the floor of the livestock house into several regular grid units, and collecting the acoustic sequence and vibration sequence of each grid unit; Extract time-frequency features from the voiceprint sequence to construct the short-time energy duration of each grid cell. Spectral centroid and bandpass energy ratio It determines whether the grid cell contains a "urination acoustic event" or a "defecation acoustic event". If it does, it generates a first urination event label or a first defecation event label for the grid cell. Extract vibration features from the vibration sequence and construct the short-time vibration energy duration for each grid cell. and peak vibration and shock It determines whether the grid cell has a "urination vibration event" or a "defecation vibration event". If it does, it generates a second urination event label or a second defecation event label for the grid cell. If a grid cell at the same time generates either a first urination event label and a second urination event label, or a first defecation event label and a second defecation event label, it is identified as an actual urination event or an actual defecation event. After triggering the second identification command, each grid cell is constructed to record porosity ϕ, average pore size davg, permeability coefficient K, surface energy γ, local slope α, and flow resistance coefficient. The overall flow resistance coefficient R of each grid cell is calculated to evaluate whether the flow guidance of each grid cell is qualified. When the flow guidance is not qualified, a first blockage mark or a second blockage mark is generated. If several grid cells on the same vertical or horizontal path are identified, and all grid cells on the same vertical or horizontal path are identified as first qualified markers, then the negative pressure suction strategy is not triggered. When any grid cell is identified as either a first blocking marker or a second blocking marker, a corresponding first intervention strategy and a second intervention strategy are generated. After the corresponding first intervention strategy and second intervention strategy are executed, an improvement rate ImpRate is constructed for verification.

[0006] Preferably, it also includes: S2, based on the actual urination event triggered in S1, collecting the time series data of ammonia concentration of the grid cell and calculating the hot spot index HPI of the urination location, setting a hot spot threshold, if the hot spot index HPI ≥ the hot spot threshold, it is determined that the grid cell has a long-term urination aggregation effect, forming a "long-term focused urine point", triggering the first urination location optimization sub-strategy; S3. While S1 and S2 are being executed, the scouring intensity parameters of urine impacting the ground are collected, and the self-cleaning scouring effect index Cclean for each "long-term focused urine point" is constructed. When Cclean≤0.7, it indicates that the self-cleaning ability of the urination point is unqualified, and a second urination position optimization sub-strategy is generated.

[0007] Preferably, condenser microphones are arranged at intervals of 1.0m to 1.5m on the ceiling or sidewall of each grid cell, at a height of 0.6m to 1.0m above the ground, to collect the voiceprint sequence of each grid cell. The voiceprint sequence is preprocessed, including DC removal, pre-emphasis, and Hamming windowing, before time-frequency features are extracted. The time-frequency features include: short-time energy. Spectral centroid Compared with bandpass energy ratio ; Among them, short-time energy It involves dividing a continuous sound signal into short time windows, ranging from 20ms to 25ms, and calculating the energy within each window to obtain the short-time energy. : ; Where N is the number of sample points per frame, and 25ms corresponds to a sampling rate of 16kHz, which is 400 points. This is the preprocessed signal; Perform FFT spectral transformation on the same frame to obtain the amplitude spectrum of the same frame, and extract the spectral centroid. The specific formula is as follows: ; in, For the k-th frequency point, The amplitude; For the amplitude spectrum of the same frame, first calculate the total energy of the full-band signal, then calculate the energy within the target frequency band, which is set to 300Hz to 1500Hz. The ratio of the two is the bandpass energy ratio. The specific formula is as follows: ; Set a short-time energy threshold, if If the energy level is higher than the short-time energy threshold, then for consecutive frames... The durations are summed to obtain the short-term energy duration. The preferred range for the short-time energy threshold is set to 1×10. 4 Up to 5×10 4 ; During the 2-second statistical window, if any two of the following conditions are met, the voiceprint event is determined to be a "urination acoustic event," and a first urination event label is generated for that grid cell, including: Condition 1: Duration of short-time energy ≥0.3s; the sound of flowing water is continuous and steady; Condition 2: Spectral mass center Within the 300Hz to 1500Hz range and with fluctuations ≤20% within the window; Condition 3: Bandpass energy ratio ≥0.45; During the 2-second statistical window, if any two of the following conditions are met, the voiceprint event is determined to be a "fecal acoustic event," and a first fecal event label is generated for that grid cell, including: Condition 1: Duration of short-time energy <0.2s; Condition 2: Spectral mass center <400Hz; the spectral centroid is concentrated in the low frequency range <400Hz; Condition 3: Bandpass energy ratio <0.45.

[0008] Preferably, piezoelectric or strain gauge vibration sensors are arranged at the edge of each grid cell to collect vibration sequences at a sampling rate of 5 kHz to 10 kHz, with the sequence length being the number of sampling points within the event time window; The vibration sequence was divided into frames with a frame length of 25ms or a frame shift of 10ms. Each frame was preprocessed, including DC removal and Hamming windowing, to generate a preprocessed vibration frame signal V[m]. Vibration features were extracted, and the short-time vibration energy was obtained using the following formula. and peak vibration and shock : ; ; Where N is the number of sample points per frame, and short-time vibration energy is... Reflecting the vibration intensity of the frame, the sound of urination corresponds to the impact of flowing water, which produces continuous vibration, while the sound of defecation corresponds to an instantaneous impact; the peak value of the vibration impact. The impact amplitude is instantaneous, and the impact at the point of defecation is significantly higher than that of urine flow; Set the urination vibration threshold Evth to 1×10. −3 Up to 5×10 −3 V 2 ; Set the fecal discharge vibration threshold Pvth to 1×10. −2 Up to 5×10 −2 V; Short-term vibration energy The duration exceeding the urination vibration threshold Evth is denoted as the short-time vibration energy duration. ; Within a 2-second statistical window, conditions are identified and corresponding vibration events are determined, including: When the duration of short-term vibration energy If the time is ≥0.3s, it is determined to be a "urination vibration event", and a second urination event label is generated for the grid cell; When the peak vibration and impact > The fecal vibration threshold Pvth, and the duration of short-term vibration energy If the time is less than 0.2s, it is determined to be a "fecal vibration event", and a second fecal event tag is generated for the grid cell; The grid cells at the same time simultaneously generate the first urination event label and the second urination event label, and are identified as actual urination events; The grid cells at the same time simultaneously generate the first and second defecation event labels, and are identified as actual defecation events. A second identification instruction is triggered when an actual urination or defecation event is identified.

[0009] Preferably, when the second identification command is recognized, each grid cell is constructed to record the porosity ϕ, average pore size davg, permeability coefficient K, surface energy γ, local slope α, and flow resistance coefficient. ; Specifically as follows: S11. Calculate the proportion of pore volume within a grid cell to the total volume using 3D scanning to obtain the porosity ϕ: ; Where, represents the actual void volume within the grid cell. This represents the total volume of the grid cells; S12. Use laser scanning to obtain the aperture distribution, calculate the average equivalent diameter of the pores within the grid cell, and obtain the average aperture davg: ; in, Let be the equivalent diameter of the i-th hole, and n be the total number of holes in the unit. S13. Using Darcy's law, calculate the permeability of water or urine in the grid cell medium to obtain the permeability coefficient K: ; Where Q represents the penetration flow, Indicates the height of the grid cell, and A represents the cross-sectional area of ​​the grid cell. Indicates a negative pressure difference. Indicates the flow of time; S14. Measure the contact angle using the droplet method, calculate the wettability index of the grating material surface to the liquid, and obtain the surface energy γ: ; in, For liquid surface tension, Contact angle; S15. The tilt angle of the grid cell surface relative to the direction of liquid flow, obtaining the local slope α: ; Where Δh is the height difference of the grid cells, and Δx is the horizontal length; S16. The resistance of the grid cells to the fluid is used to obtain the flow resistance coefficient. : ; in, For unit pressure drop, Let ρ be the fluid density and v be the average flow velocity.

[0010] Preferably, each grid cell records the porosity ϕ, average pore size davg, permeability coefficient K, surface energy γ, local slope α, and flow resistance coefficient. Normalized to the [0,1] interval, and the overall flow resistance coefficient R of each grid cell is calculated using a weighted formula; The overall flow resistance coefficient R value of each grid cell is evaluated, including: When R < 0.3, it indicates that the grid cell is qualified for current conduction, and the first qualified mark is generated; When 0.3≤R≤0.6, it indicates that the flow guidance of the grid cell is unqualified, and the first blockage mark is generated; When R > 0.6, it indicates that the flow conduction of the grid cell is not qualified, and a second blockage mark is generated.

[0011] Preferably, several grid cells on the same vertical or horizontal path are identified. If all grid cells on the same vertical or horizontal path are identified as the first qualified marker, the negative pressure suction strategy is not triggered. If any grid cell is identified as the first blockage marker, a first intervention strategy is generated, including: triggering negative pressure suction, performing a single-round suction, with a maximum negative pressure of −2.5 kPa for the single-round pulse, and the duration of the single-round pulse is set to 3 to 6 seconds; If any grid cell is identified as a second blockage marker, a second intervention strategy is generated, including: triggering negative pressure suction, performing multiple rounds of suction, with a maximum negative pressure of -3.0 kPa to 6.0 kPa per round pulse, a pulse duration of 8 to 10 seconds per round pulse, and performing 2 to 5 pulses per round pulse with a pulse interval of 5 to 8 seconds.

[0012] Preferably, after implementing the first and second intervention strategies, a delay time trecheckt is set, and the new comprehensive flow resistance R2 of the grid cell is remeasured after 5 to 10 seconds, while the improvement rate ImpRate is calculated: ; Where R is the total flow resistance before intervention; If R2 < 0.3, mark it as "recovered", determine that the flow diversion of the grid cell is qualified, and stop subsequent intervention; If the improvement rate ImpRate ≥ 15% and R2 ≤ 0.6, it is judged as "relief", indicating that although the drainage is not qualified, there is improvement. The unit is included in the first priority follow-up observation queue, and the intervention is not upgraded for the time being. The first intervention strategy is repeated once, that is, a single round of negative pressure aspiration: −2.5kPa, duration 3s to 6s. If the improvement rate ImpRate < 15%, it is judged as "insufficient intervention", and a third intervention strategy is generated, including: increasing the number of the second intervention of the grid unit by 1, that is, multiple rounds of negative pressure suction: −3.0kPa to −6.0kPa, 8s to 10s, 2 to 5 cycles, with an interval of 5s to 8s. If the improvement rate ImpRate is still <15%, it is determined to be a continuous blockage. The grid cell is then directly added to the "manual cleaning and replacement sequence", and a manual replacement priority sequence list is generated according to the improvement rate ImpRate from small to large.

[0013] Preferably, S2 includes: S21. Based on the actual urination event triggered in S1, firstly, continuously collect time-series data of ammonia concentration changes over time within the target grid cell, and calculate the urination location hotspot index (HPI) of the grid cell according to the following formula: ; in, Let j represent the ammonia concentration at the j-th sampling point; j = 1, 2, ..., M, where j is the time series index. The time decay weighting coefficient is set as follows: ; S22. Set a hot spot threshold. When the hot spot index HPI at the urination location of the grid cell is greater than the hot spot threshold, it is determined that the grid cell has a long-term urination aggregation effect and is marked as a "long-term focused urine point". When the determination result is "long-term focus on urine point", the first urination position optimization sub-strategy is automatically triggered, including: A corrosion-resistant or hydrophobic protective film of 0.5mm to 2.0mm is laid in the flow channel within the grid unit, and an odor release device and fan are installed to drive odor diffusion, form a local odor gradient, and guide livestock and poultry to move towards the preset urine collection trough or flow channel; one to two odor release devices are arranged above or to the side of each preset urine point, with a height of 0.5m to 1m and a coverage radius of 1m to 2m.

[0014] Preferably, S3 includes: S31. Pressure sensors are installed on the inner wall of the guide channel at each "long-term focused urine point". The collected pressure signal P(t) is a continuous sequence that changes with time, thus obtaining impact signal data. Short-time energy analysis and peak statistics are performed on the collected impact signal data to obtain scouring intensity parameters, including the instantaneous impact peak value Fpeak, the average scouring force Favg, and the scouring duration. ; Instantaneous peak impact Fpeak, average scouring force Favg, and scouring duration The method of obtaining it is: The pressure signal is divided into frames with a frame length of 25ms or a frame shift of 10ms. Short-time energy is calculated for the signal P[m] in each frame to obtain the short-time pressure energy. : ; Where: N is the number of sampling points per frame, and P[m] is the preprocessed pressure signal; And identify short-term energy of pressure. The instantaneous peak value and sustained amplitude of the impact are calculated using the following formulas to obtain the instantaneous impact peak value Fpeak and the average scouring force Favg: ; ; Among them, the maximum value of the pressure signal multiplied by the effective area of ​​the sensor This yields the instantaneous impact force, i.e., the instantaneous impact peak value Fpeak; To determine the flushing duration, a pressure threshold Pth is set; when the pressure is short, energy is applied. > Pressure threshold Pth indicates the start of effective flushing; calculate the short-time energy of continuous pressure. The duration of flushing is determined by the time the pressure threshold Pth is exceeded. S32. Normalize the flushing intensity parameters and calculate the self-cleaning flushing effect index Cclean for each "long-term focused urine point" using the following formula: ; in, Indicates the preset maximum impact force. This indicates the preset maximum flushing time, used to set the instantaneous peak impact value Fpeak, average flushing force Favg, and flushing duration. Perform normalization processing; , and This represents the weight coefficients, and the sum of the weights is 1; S33. If the self-cleaning flushing effect index Cclean > 0.7 for each "long-term focused urine point", it means that the self-cleaning ability of the urine point is qualified. If the self-cleaning flushing effect index Cclean ≤ 0.7 for each "long-term focused urination point", it indicates that the self-cleaning ability of that urination point is unqualified. Therefore, a second urination location optimization sub-strategy is generated, including: Turn off the odor emitter and fan drive that are currently not self-cleaning enough, and set up a low-frequency sound wave generator and LED flexible light source in the grid unit of the nearest "long-term focused urine point" with good self-cleaning ability near the current "long-term focused urine point" with insufficient self-cleaning ability. When the animal approaches the preset urine point, the sound and light unit emits a soft low-frequency sound, set to 500Hz to 1000Hz and visible light signal, to form a directional cue to go to the "long-term focused urine point" with good self-cleaning ability, and assist the odor induction to the new long-term focused urine point, thus forming behavioral localization memory.

[0015] This invention provides an ecological treatment method for livestock and poultry breeding waste. It has the following beneficial effects: (1) The ecological treatment method for livestock and poultry breeding waste can accurately identify urination and defecation events by arranging microphones and vibration sensors in each grid unit and combining acoustic characteristics (short-time energy, spectral centroid, bandpass energy ratio) and vibration characteristics (short-time vibration energy, impact peak) for dual judgment. This can effectively reduce misjudgment caused by environmental noise or other interference, improve the reliability of excretion behavior detection, and provide an accurate basis for subsequent grid unit management.

[0016] (2) The ecological treatment method for livestock and poultry breeding waste records the porosity, average pore size, permeability coefficient, surface energy, local slope and flow resistance coefficient for each grid unit, and calculates the comprehensive flow resistance coefficient R. By setting qualified and blocked markers and combining single or multiple rounds of negative pressure suction intervention strategy, timely treatment can be carried out for areas with poor flow. The intervention effect is judged by updating the improvement rate, ensuring uniform distribution of waste liquid and reducing retention, and improving the adaptive management capability of the grid system.

[0017] (3) This method of ecological treatment of livestock and poultry breeding waste identifies several grid units on the same vertical or horizontal path as a whole. When all units are qualified for diversion, the negative pressure suction strategy is not triggered. If some units are blocked, intervention is performed according to priority. This helps to dynamically regulate from local units to the whole path segment, avoid excessive intervention, improve resource utilization efficiency, and reduce system energy consumption and operating costs.

[0018] (4) This method of ecological treatment of livestock and poultry breeding waste identifies “long-term focused urine points” based on the urine location hot spot index HPI and combines it with the self-cleaning flushing effect index Cclean. It can scientifically judge the urine hot spots and areas with unqualified self-cleaning ability. By optimizing the urine location through odor, sound and light guidance and the protective film of the diversion channel, it can achieve the joint optimization of behavior guidance and self-cleaning, effectively reduce the accumulation of waste liquid, reduce the risk of ammonia hot spots, and improve the environmental hygiene and facility durability of livestock and poultry houses. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of step S1 of the method of the present invention.

[0020] Figure 2 This is a schematic diagram of steps S2-S3 of the method of the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Example 1 Please see Figure 1-2 This invention provides an ecological treatment method for livestock and poultry breeding waste, comprising the following steps: S1. Arrange diversion channels and selective infiltration grid devices in the livestock and poultry house, divide the diversion channels and the floor of the livestock house into several regular grid units, and collect the acoustic and vibration sequences of each grid unit. Extract time-frequency features from the voiceprint sequence to construct the short-time energy duration of each grid cell. Spectral centroid and bandpass energy ratio It determines whether the grid cell contains a "urination acoustic event" or a "defecation acoustic event". If it does, it generates a first urination event label or a first defecation event label for the grid cell. Extract vibration features from the vibration sequence and construct the short-time vibration energy duration for each grid cell. and peak vibration and shock It determines whether the grid cell has a "urination vibration event" or a "defecation vibration event". If it does, it generates a second urination event label or a second defecation event label for the grid cell. If a grid cell at the same time generates either a first urination event label and a second urination event label, or a first defecation event label and a second defecation event label, it is identified as an actual urination event or an actual defecation event. After triggering the second identification command, each grid cell is constructed to record porosity ϕ, average pore size davg, permeability coefficient K, surface energy γ, local slope α, and flow resistance coefficient. The overall flow resistance coefficient R of each grid cell is calculated to evaluate whether the flow guidance of each grid cell is qualified. When the flow guidance is not qualified, a first blockage mark or a second blockage mark is generated. If several grid cells on the same vertical or horizontal path are identified, and all grid cells on the same vertical or horizontal path are identified as first qualified markers, then the negative pressure suction strategy is not triggered. When any grid cell is identified as either a first blocking marker or a second blocking marker, a corresponding first intervention strategy and a second intervention strategy are generated. After the corresponding first intervention strategy and second intervention strategy are executed, an improvement rate ImpRate is constructed for verification.

[0023] S2. Based on the actual urination event triggered in S1, collect the time series data of ammonia concentration of the grid cell and calculate the hot spot index (HPI) of the urination location. Set a hot spot threshold. If the hot spot index (HPI) is greater than or equal to the hot spot threshold, it is determined that there is a long-term urination aggregation effect in the area, forming a "long-term focused urination point", and triggering the first urination location optimization sub-strategy. S3. While S1 and S2 are being executed, the scouring intensity parameters of urine impacting the ground are collected, and the self-cleaning scouring effect index Cclean for each "long-term focused urine point" is constructed. When Cclean≤0.7, it indicates that the self-cleaning ability of the urination point is unqualified, and a second urination position optimization sub-strategy is generated.

[0024] In this embodiment, the present invention divides the ground into regular grid units by arranging diversion channels and selective permeation grid devices in the livestock and poultry house. Combined with acoustic and vibration monitoring technology, it achieves accurate identification of urination and defecation behavior. Based on this, by constructing the flow resistance coefficient R and the urine flushing self-cleaning effect index Cclean for each grid unit, the invention dynamically judges the compliance of diversion and the situation of local blockage, and can trigger negative pressure suction or diversion intervention in a targeted manner, solving the problems of urine and feces mixing, local retention, and incomplete cleaning in traditional technologies. Simultaneously, the present invention identifies long-term focused urine points based on the urination location thermal spot index HPI, and through the synergistic effect of odor induction, sound and light guidance, and self-cleaning coating, forms stable preferred urine points, guiding animals to urinate in areas with qualified self-cleaning capabilities. This achieves centralized collection and diversion of urine, improves the uniformity of urine discharge and the self-cleaning efficiency of waste, effectively reduces the concentration of ammonia and harmful gases in the house, reduces ground corrosion and facility maintenance costs, and ensures the ecological treatment and resource utilization of excrement.

[0025] Example 2 This embodiment is an explanation of Embodiment 1. Specifically, a flow guide channel and a selective infiltration grid device are arranged inside the livestock and poultry house, dividing the flow guide channel and the floor of the livestock house into several regular grid units. Each grid cell has a length of 5cm to 15cm, a width of 5cm to 15cm, and a height of 2cm to 5cm. The initial spacing between grid units is set to 0mm to 5mm, and the width of the drainage guide gap between grid units is set to 2mm to 8mm. The initial inclination angle of the guide channel along the discharge flow direction is set to 1° to 3°.

[0026] By arranging diversion channels and selective infiltration grid devices in livestock and poultry houses, and dividing the diversion channels and the ground into several regular grid units, the size and height of each grid unit are set reasonably, the grid spacing and drainage diversion gap width are optimized, and the diversion channels have an appropriate inclination angle along the discharge flow direction. The grid height and spacing are moderate and do not affect the normal walking and activities of livestock and poultry. At the same time, the regular diversion guides the location of urination and defecation, which is conducive to forming stable excretion behavior habits and realizing ecological management.

[0027] Condenser microphones are arranged at intervals of 1.0m to 1.5m on the ceiling or sidewall of each grid cell, at a height of 0.6m to 1.0m above the ground, to collect the voiceprint sequence of each grid cell. The voiceprint sequence is preprocessed, including DC removal, pre-emphasis, and Hamming windowing, before time-frequency features are extracted. The time-frequency features include: short-time energy. Spectral centroid Compared with bandpass energy ratio ; Among them, short-time energy It involves dividing a continuous sound signal into short time windows, ranging from 20ms to 25ms, and calculating the energy within each window to obtain the short-time energy. : ; Where N is the number of sample points per frame, and 25ms corresponds to a sampling rate of 16kHz, which is 400 points. For the preprocessed signal, the voiceprint sequence is first converted into an electrical signal, and then a digital signal is obtained through an analog-to-digital converter (ADC). The sequence length is the number of sampling points within the time window. The meaning is: sum the squared amplitude of each sampling point in this frame to get the total energy of the frame. The larger the sound amplitude, the higher the energy. The sound of urinating water is more continuous and stronger than the background environmental sound, so the energy is higher in a short time. Perform FFT spectral transformation on the same frame to obtain the amplitude spectrum of the same frame, and extract the spectral centroid. That is, the "centroid" of the amplitude spectrum: ; in, For the k-th frequency point, The amplitude; physical meaning: represents the frequency centroid of sound energy; urination sounds are concentrated in the 300Hz to 1500Hz range, and the spectral centroid fluctuation is ≤20% as the criterion; For the amplitude spectrum of the same frame, first calculate the total energy of the full-band signal, then calculate the energy within the target band, which is set to 300Hz to 1500Hz. The ratio of the two is the bandpass energy ratio. The specific formula is as follows: ; Among them, frequencies below 300Hz are easily mixed with low-frequency environmental noise (such as air conditioner noise, current noise). The frequency range above 1500Hz is susceptible to interference from splashes and high-frequency environmental noise. Therefore, 300Hz to 1500Hz was selected as the typical "main frequency band of the event", which is also the target frequency band. Set a short-time energy threshold, if If the energy level is higher than the short-time energy threshold, then for consecutive frames... The durations are summed to obtain the short-term energy duration. The preferred range for the short-time energy threshold is set to 1×10. 4 Up to 5×10 4 ; 1×10 4 5×10 represents a low sound intensity (weak sound of flowing water or small animals). 4 Indicates a high sound intensity (clearly the sound of running water or urination). Within this range, only frames exceeding the short-time energy threshold are considered valid signals of an excretion event; During the 2-second statistical window, if any two of the following conditions are met, the voiceprint event is determined to be a "urination acoustic event," and a first urination event label is generated for that grid cell, including: Condition 1: Duration of short-time energy ≥0.3s; the sound of flowing water is continuous and steady; Condition 2: Spectral mass center Within the 300Hz to 1500Hz range and with fluctuations ≤20% within the window; Condition 3: Bandpass energy ratio ≥0.45; During the 2-second statistical window, if any two of the following conditions are met, the voiceprint event is determined to be a "fecal acoustic event," and a first fecal event label is generated for that grid cell, including: Condition 1: Duration of short-time energy <0.2s; (instantaneous impact type).

[0028] Condition 2: Spectral mass center <400Hz; the spectral centroid is concentrated in the low frequency range <400Hz; Condition 3: Bandpass energy ratio <0.45.

[0029] In this embodiment, by deploying capacitive microphones in each grid cell, acoustic signature sequences are collected, and time-frequency features such as short-time energy, spectral centroid, and bandpass energy ratio are extracted. This allows for the differentiation between urination and defecation sounds, achieving precise excretion event identification at the grid cell level. Short-time energy duration and FFT spectral analysis enable the system to quickly determine urination or defecation events within a 2-second window, adapting to the instantaneous and irregular nature of animal excretion behavior and improving real-time monitoring. By selecting 300 to 1500 Hz as the main frequency band for urination, interference from low-frequency environmental noise and high-frequency splashing sounds is eliminated. Simultaneously, bandpass energy ratio and spectral centroid fluctuation conditions are set to ensure accurate and reliable signal determination. Precise location of urination and defecation events provides data support for subsequent diversion channel design, excrement diversion, and ecological intervention, achieving unified monitoring and intelligent management of urination and defecation. Automatic acoustic recognition reduces the number of manual inspections, improves waste collection efficiency, and, combined with grid management and diversion design, reduces the risk of secondary pollution and localized accumulation.

[0030] Example 3 This embodiment is an explanation of Embodiment 1. Specifically, piezoelectric or strain gauge vibration sensors are arranged at the edge of each grid cell to collect vibration sequences. The sampling rate is 5kHz to 10kHz, and the sequence length is the number of sampling points within the event time window. The vibration sequence was divided into frames with a frame length of 25ms or a frame shift of 10ms. Each frame was preprocessed, including DC removal and Hamming windowing, to generate a preprocessed vibration frame signal V[m]. Vibration features were extracted, and the short-time vibration energy was obtained using the following formula. and peak vibration and shock : ; ; Where N is the number of sample points per frame, and short-time vibration energy is... Reflecting the vibration intensity of the frame, the sound of urination corresponds to the impact of flowing water, which produces continuous vibration, while the sound of defecation corresponds to an instantaneous impact; the peak value of the vibration impact. The impact amplitude is instantaneous, and the impact at the point of defecation is significantly higher than that of urine flow; The preferred range for the urination vibration threshold Evth is set to 1×10. −3 V to 5×10 −3 V 2 During urination, urine continuously impacts the ground or a drainage channel from a certain height, producing a relatively continuous but moderate-amplitude vibration. The vibration amplitude is not as large as that of defecation, therefore the threshold is lower, 1×10⁻⁶. −3 Up to 5×10 −3 V 2 It is sufficient to capture continuous water vibration signals without misinterpreting slight environmental vibrations; The preferred range for the fecal discharge vibration threshold Pvth is set to 1×10. −2 V to 5×10 −2 V; During defecation, the impact of feces hitting the ground is usually instantaneous, with a vibration amplitude significantly greater than that of flowing water, but the duration is very short. The instantaneous impact vibration amplitude generated by defecation is large, therefore the threshold is relatively high, ranging from 1×10⁻⁶. −2 V to 5×10 −2 V; Short-term vibration energy The duration exceeding the urination vibration threshold Evth is denoted as the short-time vibration energy duration. ; Within a 2-second statistical window, conditions are identified and corresponding vibration events are determined, including: When the duration of short-term vibration energy If the time is ≥0.3s, it is determined to be a "urination vibration event", and a second urination event label is generated for the grid cell; When the peak vibration and impact > The fecal vibration threshold Pvth, and the duration of short-term vibration energy If the time is less than 0.2s, it is determined to be a "fecal vibration event", and a second fecal event tag is generated for the grid cell; Urinating is a continuous event, typically lasting ≥0.3s, so the cumulative duration of consecutive frames must reach 0.3s to be considered a valid urination vibration event. Low amplitude + long duration → corresponds to continuous vibration caused by the sound of flowing water during urination. Fecal impact is very short, typically <0.2s, therefore short cumulative time is a criterion for judgment; The grid cells at the same time simultaneously generate the first urination event label and the second urination event label, and are identified as actual urination events; The grid cells at the same time simultaneously generate the first and second defecation event labels, and are identified as actual defecation events. A second identification instruction is triggered when an actual urination or defecation event is identified.

[0031] In this embodiment, by arranging piezoelectric or strain gauge vibration sensors at the edge of each grid cell, high-sampling-rate vibration sequences are collected and combined with acoustic fingerprint sequences for dual-modal determination, enabling accurate identification of urination and defecation behaviors. Urination events are determined by a short-term vibration energy duration ≥ 0.3s, while defecation events are determined by a vibration impact peak value higher than a threshold and a duration < 0.2s. Furthermore, acoustic short-term energy, spectral centroid, and bandpass energy ratio are combined to further confirm the excretion type, thereby significantly reducing misjudgments and omissions caused by environmental noise, mechanical vibration, or animal movement, achieving high-precision event localization. This method can distinguish between the continuous low-amplitude vibration of urination and the instantaneous high-amplitude impact vibration of defecation landing points, providing accurate data support for grid cell-level waste management. It supports ecological intervention measures such as diversion channel control, negative pressure suction, and urination location optimization, and can capture transient characteristics of excretion behavior in real time, reducing the frequency of manual inspections and cleaning costs.

[0032] Example 4 This embodiment is an explanation of Embodiment 1. Specifically, when the second identification command is recognized, each grid cell is constructed to record the porosity ϕ, average pore size davg, permeability coefficient K, surface energy γ, local slope α, and flow resistance coefficient. ; Specifically as follows: S11. Calculate the proportion of pore volume within a grid cell to the total volume using 3D scanning to obtain the porosity ϕ: ; Where, represents the actual void volume within the grid cell. The total volume of the grid cell; the higher the porosity ϕ, the easier it is for the liquid to permeate; used to evaluate the grid cell's ability to store and conduct urine / feces. S12. Use laser scanning to obtain the aperture distribution, calculate the average equivalent diameter of the pores within the grid cell, and obtain the average aperture davg: ; in, The equivalent diameter of the i-th hole is denoted by , and n is the total number of holes in the cell. The average equivalent diameter of the pores inside the grid cell reflects the influence of pore size on the resistance to liquid flow. S13. Using Darcy's law, calculate the permeability of water or urine in the grid cell medium to obtain the permeability coefficient K: ; Where Q represents the penetration flow, Indicates the height of the grid cell, and A represents the cross-sectional area of ​​the grid cell. Indicates a negative pressure difference. The flow time is represented by the permeability coefficient K, which describes the ability of a liquid to pass through the grid cell medium. A higher permeability coefficient indicates easier flow; it is used to predict flow velocity and negative pressure suction efficiency. S14. Measure the contact angle using the droplet method, calculate the wettability index of the grating material surface to the liquid, and obtain the surface energy γ: ; in, For liquid surface tension, The contact angle is γ; surface energy γ describes the wettability of the grid material surface to the liquid. The greater the surface energy, the easier it is for the liquid to spread; it affects urine drainage and self-cleaning effect. S15. The tilt angle of the grid cell surface relative to the direction of liquid flow, obtaining the local slope α: ; Where Δh is the height difference of the grid cells, Δx is the horizontal length, and the slope angle of the grid cell surface relative to the horizontal plane is used to predict the direction and velocity of the liquid flow along the grid cells. S16. The resistance of the grid cells to the fluid is used to obtain the flow resistance coefficient. : ; in, For unit pressure drop, Where is the fluid density, v is the average flow velocity, and the flow resistance coefficient is... It indicates the resistance of the grid cell to the liquid flow. The greater the resistance to the liquid surface, the slower the liquid flow rate; it is used to dynamically adjust the negative pressure suction and diversion efficiency.

[0033] In this embodiment, by constructing the porosity ϕ, average pore size davg, permeability coefficient K, surface energy γ, local slope α, and flow resistance coefficient of each grid cell, the storage, permeation, and flow conduction characteristics of the grid cell for liquids can be comprehensively quantified, enabling accurate prediction and control of urine and fecal flow behavior. Porosity ϕ reflects the liquid's permeability within the grid cell; the average pore size davg reflects the magnitude of liquid flow resistance; permeability coefficient K is used to assess the speed of liquid flow through the medium and the efficiency of negative pressure suction; surface energy γ represents the wettability of the grid material and affects liquid spreading and self-cleaning effects; local slope α determines the direction and speed of liquid flow along the grid surface; and the flow resistance coefficient comprehensively describes the ease with which liquid flows due to the adhesion resistance of the grid surface. By measuring and calculating these parameters, not only can the flow conduction and storage capacity of each grid cell be dynamically evaluated, but a scientific basis can also be provided for negative pressure suction strategies, diversion control, and self-cleaning design, thereby reducing local liquid retention and accumulation and improving the uniformity of urine and fecal distribution.

[0034] Example 5 This embodiment is an explanation of Embodiment 1. Specifically, each grid cell records the porosity ϕ, average pore size davg, permeability coefficient K, surface energy γ, local slope α, and flow resistance coefficient. Normalize to the [0,1] interval, and calculate the overall flow resistance coefficient R of each grid cell using the following formula; ; in, , , and The larger the value, the greater the resistance. , , , , and This represents the weight coefficients, and the sum of the weights is 1; Weighting coefficient Porosity has a significant impact on liquid flow; the lower the porosity, the greater the increase in resistance, hence its higher weight. Weighting coefficient The permeability coefficient directly reflects the flow capacity of a liquid in a grid medium and has the greatest impact on the smoothness of flow, therefore it is given the highest weight. Weighting coefficient Local slope affects the velocity of liquid flow along the surface, but its range of variation is limited, therefore its weight is moderate. Weighting coefficient 0.15; The average pore size reflects the channel size and has some influence on flow resistance, but its effect is slightly less than that of the overall porosity. Weighting coefficient Surface energy affects liquid spreading and adhesion, but its effect is relatively small and its weight is low. Weighting coefficient The local flow resistance of the grid cell directly affects the liquid flow rate, but considering the redundancy with other parameters, the weight is appropriate. R∈[0,1], the larger the value, the more difficult the grid flow is; the smaller the value, the smoother the flow is, and negative pressure suction or diversion can be triggered. The overall flow resistance coefficient R value for each grid cell is evaluated, including: When R < 0.3, it indicates that the grid cell is qualified for current conduction, and the first qualified mark is generated; A qualified flow indicates that the migration of fecal sewage or waste liquid between grid units is smooth, and it can naturally infiltrate or flow through its own gravity, capillary action and slope. In this case, no additional negative pressure suction or artificial diversion is required. When 0.3≤R≤0.6, it indicates that the flow guidance of the grid cell is unqualified, and the first blockage mark is generated; When R > 0.6, it indicates that the flow conduction of the grid cell is unqualified, and a second blockage mark is generated; Inadequate flow guidance indicates high pore resistance and poor channel flow, which can easily lead to local stagnation or accumulation, resulting in uneven distribution of waste liquid. Only in this case is it necessary to introduce external force intervention. If several grid cells on the same vertical or horizontal path are identified, and all grid cells on the same vertical or horizontal path are identified as first qualified markers, then the negative pressure suction strategy is not triggered. If any grid cell is identified as the first blockage marker, a first intervention strategy is generated, including: triggering negative pressure suction, performing a single-round suction, with a maximum negative pressure of −2.5 kPa for the single-round pulse and a single-round pulse duration of 3 to 6 seconds; If any grid cell is identified as a second blockage marker, a second intervention strategy is generated, including: triggering negative pressure suction, performing multiple rounds of suction, with a maximum negative pressure of -3.0 kPa to 6.0 kPa per round pulse, a pulse duration of 8 to 10 seconds per round pulse, and performing 2 to 5 pulses per round pulse with a pulse interval of 5 to 8 seconds.

[0035] In this embodiment, the overall flow resistance status of several grid units on the same vertical or horizontal path is integrated through path-level judgment. If all grid units in the path are marked as first qualified, the path is smooth and the negative pressure suction strategy is not triggered. If any grid unit in the path is marked as blocked, the corresponding intervention strategy is triggered according to the level of blockage. The first blockage mark triggers a single round of suction (pulse maximum negative pressure -2.5 kPa, lasting 3 to 6 seconds), and the second blockage mark triggers multiple rounds of suction (pulse maximum negative pressure -3.0 to -6.0 kPa, single round pulse 8 to 10 seconds, 2 to 5 times per round, pulse interval 5 to 8 seconds). The beneficial effects of this method are that by determining the flow from individual grid cells to the path level, it achieves accurate differentiation between local and overall flow states, making the waste liquid distribution uniform and reducing the risk of local liquid accumulation. At the same time, it can reasonably trigger external force intervention strategies, improve cleaning efficiency and reduce energy consumption, ensure the coordination between the self-cleaning ability of the excretion area and animal behavior habits, effectively reduce blockage and secondary pollution, and improve the level of ecological management of livestock and poultry houses and the efficiency of waste treatment.

[0036] Example 6 This embodiment is an explanation of Embodiment 1. Specifically, after executing the first and second intervention strategies, a delay time trecheckt is set, and the new comprehensive flow resistance R2 of the grid cell is remeasured after 5 to 10 seconds, while the improvement rate ImpRate is calculated. ; Where R is the total flow resistance before intervention; If R2 < 0.3, mark it as "recovered", determine that the flow diversion of the grid cell is qualified, and stop subsequent intervention; If the improvement rate ImpRate ≥ 15% and R2 ≤ 0.6, it is judged as "relief", indicating that although the drainage is not qualified, there is improvement. The unit is included in the first priority follow-up observation queue, and the intervention is not upgraded for the time being. The first intervention strategy is repeated once, that is, a single round of negative pressure aspiration: -2.5kPa, duration 3s to 6s. If the improvement rate ImpRate < 15%, it is judged as "insufficient intervention", and a third intervention strategy is generated, including: increasing the number of the second intervention of the grid unit by 1, that is, multiple rounds of negative pressure suction: −3.0kPa to −6.0kPa, 8s to 10s, 2 to 5 cycles, with an interval of 5s to 8s. If the improvement rate ImpRate is still <15%, it is determined to be a continuous blockage. The grid cell is then directly added to the "manual cleaning and replacement sequence", and a manual replacement priority sequence list is generated according to the improvement rate ImpRate from small to large.

[0037] In this embodiment, dynamic retesting and improvement rate evaluation enable hierarchical adjustment and precise management of intervention strategies. This not only avoids ineffective or excessive intervention and reduces energy consumption and equipment wear, but also ensures the continuity and stability of waste liquid diversion.

[0038] The design of a priority list for manual replacement helps to scientifically prioritize the cleaning or replacement of various grid units within the livestock shed, thereby improving maintenance efficiency and resource utilization. By arranging them from lowest to highest improvement rate, the most clogged and least recoverable grid units are prioritized, while units with relatively better flow characteristics are addressed later, avoiding blind or repetitive manual intervention. This prioritization ensures that limited manpower and time are first applied to the most critical blockage points, thereby quickly restoring overall flow performance and reducing the risk of wastewater retention and secondary pollution in the livestock shed. Overall, the priority list implements a systematic maintenance strategy from high-risk to low-risk, improving the scientific, continuous, and operable nature of grid unit management, while balancing intervention efficiency and ecological management effects.

[0039] Example 7 This embodiment is an explanation of Embodiment 1; please refer to it. Figure 2 Specifically, S2 includes: S21. Based on the actual urination event triggered in S1, firstly, continuously collect time-series data of ammonia concentration changes over time within the target grid cell, and calculate the urination location hotspot index (HPI) of the grid cell according to the following formula: ; in, Let j represent the ammonia concentration at the j-th sampling point; j = 1, 2, ..., M, where j is the time series index. The time decay weighting coefficient is set as follows: M represents the total number of sampling points; the earliest sampling point has the lowest weight; assuming a grid cell previously had urine output but recently has had little or no urine output → that location is no longer active; if average ammonia concentration is used without distinguishing time, the impact of historical urine output may be overestimated; giving the most recent sampling point a higher weight allows HPI to more sensitively reflect recent urine output; the earliest urine output events may have been cleared or diluted by new urine, and this historical information has less impact on current urine hotspots, so it has a lower weight, making HPI more representative of the current actual state; simply put: the more recent the urine output event, the more important it is for determining "whether this place is a hotspot".

[0040] S22. Set a hot spot threshold. When the hot spot index HPI of the urination location of the grid cell is greater than the hot spot threshold, it is determined that the grid cell has a long-term urination aggregation effect, that is, the individual has a habit of repeated urination in this area, and it is marked as a "long-term focused urine point". When the determination result is "long-term focus on urine point", the first urination position optimization sub-strategy is automatically triggered, including: A corrosion-resistant or hydrophobic protective film of 0.5mm to 2.0mm is laid in the flow channel within the grid unit, and an odor release device and fan are installed to drive odor diffusion, form a local odor gradient, and guide livestock and poultry to move towards the preset urine collection trough or flow channel; one to two odor release devices are arranged above or to the side of each preset urine point, with a height of 0.5m to 1m and a coverage radius of 1m to 2m.

[0041] This first urine location optimization sub-strategy effectively reduces urine corrosion and adhesion to the diversion channel material by laying a corrosion-resistant or hydrophobic protective membrane at the "long-term focused urine point," extending the facility's service life. The membrane thickness is controlled between 0.5mm and 2.0mm to balance protective effectiveness and smooth liquid flow. Simultaneously, by arranging odor releasers and fan equipment, a local odor gradient is created, utilizing livestock's olfactory behavior to guide their urine towards the pre-set collection or diversion channel, thereby achieving targeted wastewater management and reducing urine accumulation and ammonia hot spots caused by random urination.

[0042] In this embodiment, the system fully utilizes the behavioral characteristics of livestock and poultry having fixed urination and defecation habits. By arranging diversion channels, selective permeation grids, and odor release devices and fans, the system scientifically guides livestock and poultry to concentrate their excretion behavior at preset collection points or diversion channels. Specifically, by combining real-time monitoring of the urination hotspot index (HPI) and the identification of long-term focused urine points, the system can perform targeted optimization of urination hotspots: a corrosion-resistant or hydrophobic protective film is laid within the target grid unit, and odor gradients guide animals to prefer urination areas, thereby reducing urine accumulation and ammonia hotspot formation caused by random urination. This method conforms to the natural behavioral habits of animals and, through both physical and behavioral interventions, achieves centralized collection and targeted diversion of waste liquid. This not only improves diversion efficiency and self-cleaning effect but also reduces the risk of local pollution, improves air quality and animal welfare in the livestock shed, and realizes ecological and precise management of urination and defecation.

[0043] Example 8 This embodiment is an explanation of Embodiment 1; please refer to it. Figure 2 Specifically, S3 includes: S31. Pressure sensors are installed on the inner wall of the guide channel at each "long-term focused urine point". The collected pressure signal P(t) is a continuous sequence that changes with time, thus obtaining impact signal data. Short-time energy analysis and peak statistics are performed on the collected impact signal data to obtain scouring intensity parameters, including the instantaneous impact peak value Fpeak, the average scouring force Favg, and the scouring duration. ; Instantaneous peak impact Fpeak, average scouring force Favg, and scouring duration The method of obtaining it is: The pressure signal is divided into frames with a frame length of 25ms or a frame shift of 10ms. Short-time energy is calculated for the signal P[m] in each frame to obtain the short-time pressure energy. : ; Where: N is the number of sampling points per frame, and P[m] is the preprocessed pressure signal; And identify short-term energy of pressure. The instantaneous peak value and sustained amplitude of the impact are calculated using the following formulas to obtain the instantaneous impact peak value Fpeak and the average scouring force Favg: ; ; Among them, the maximum value of the pressure signal multiplied by the effective area of ​​the sensor This yields the instantaneous impact force, i.e., the instantaneous impact peak value Fpeak; To determine the flushing duration, a pressure threshold Pth is set; when the pressure is short, energy is applied. > Pressure threshold Pth indicates the start of effective flushing; calculate the short-time energy of continuous pressure. The duration of flushing is determined by the time the pressure threshold Pth is exceeded. S32. Normalize the flushing intensity parameters and calculate the self-cleaning flushing effect index Cclean for each "long-term focused urine point" using the following formula: ; in, Indicates the preset maximum impact force. This indicates the preset maximum flushing time, used to set the instantaneous peak impact value Fpeak, average flushing force Favg, and flushing duration. Perform normalization processing; , and This represents the weight coefficients, and the sum of the weights is 1; The reason is that the instantaneous impact force directly reflects the cleaning ability of urine or liquid when it hits the ground. The greater the impact, the more obvious the flushing effect on residual liquid, so it has a higher weight. Average scouring force reflects the overall scouring capacity of the liquid under continuous action and plays an important role in maintaining the self-cleaning of the grid. It is second only to the instantaneous impact peak value, so its weight is slightly lower. The duration of rinsing affects the cumulative cleaning effect, but it is less dependent on the instantaneous rinsing and average rinsing force, so it has the lowest weight.

[0044] S33. If the self-cleaning flushing effect index Cclean > 0.7 for each "long-term focused urine point", it means that the self-cleaning ability of the urine point is qualified. If the self-cleaning flushing effect index Cclean ≤ 0.7 for each "long-term focused urination point", it indicates that the self-cleaning ability of that urination point is unqualified. Therefore, a second urination location optimization sub-strategy is generated, including: Turn off the odor emitter and fan drive that are currently not self-cleaning enough, and set up a low-frequency sound wave generator and LED flexible light source in the grid unit of the nearest "long-term focused urine point" with good self-cleaning ability near the current "long-term focused urine point" with insufficient self-cleaning ability. When the animal approaches the preset urine point, the sound and light unit emits a soft low-frequency sound, set to 500Hz to 1000Hz and visible light signal, to form a directional cue to go to the "long-term focused urine point" with good self-cleaning ability, and assist the odor induction to the new long-term focused urine point, thus forming behavioral localization memory.

[0045] The self-flushing effect occurs because when livestock urinate or defecate, the liquid impacts the ground or drainage channels, creating a continuous or instantaneous fluid scouring force in a localized area. This scouring force, combined with the pore structure, slope, permeability, and surface properties of the grid cells, allows the liquid to flow naturally along the grid cells, infiltrate, or be carried away by the drainage channels, thus cleaning away residual liquid or impurities without additional manual intervention or external suction. Therefore, "self-flushing cleaning" can be achieved by utilizing the animals' own excretory behavior and the characteristics of the grid structure. By quantifying the scouring intensity of each "long-term focused urine point" and calculating the self-cleaning scouring effect index, the self-cleaning capacity of the urine point can be accurately determined, ensuring timely infiltration or diversion of urine, reducing local accumulation and ammonia hot spots, decreasing the frequency of manual cleaning, and improving the hygiene and management efficiency of the breeding environment.

[0046] In this embodiment, the flushing intensity of the guide channel at each "long-term focused urine point" is monitored and analyzed in real time to construct the self-cleaning flushing effect index Cclean, thereby achieving a quantitative assessment of the self-cleaning ability of the urine point. Using the instantaneous impact peak value Fpeak, average flushing force Favg, and flushing duration, the flushing effect of the guide channel during actual urination can be accurately determined, thus distinguishing areas with adequate and insufficient self-cleaning ability. For urine points with insufficient self-cleaning ability, the system automatically triggers a second urination position optimization sub-strategy. This involves shutting down the currently inadequate odor emitter and fan, and using a low-frequency sound wave generator and LED flexible light source to deploy behavioral induction signals at nearby urine points with adequate self-cleaning ability, guiding the animal to prefer urine points with good self-cleaning ability, thus achieving the scientific transfer and reshaping of the animal's urination behavior. This method fully integrates animal urination habits and physical environmental conditions, which not only improves the problems of local urine accumulation and ammonia heat spots, but also enhances the self-cleaning efficiency of the diversion channel, reduces manual intervention and maintenance costs, and at the same time improves the ecological management level of the breeding house and animal welfare by forming a long-term stable urination pattern through behavioral positioning memory.

[0047] The threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by those skilled in the art for each set of sample data; as long as it does not affect the ratio between the parameter and the quantized value, it is acceptable.

[0048] The above formulas are all derived from software simulation using a large amount of data and are selected to be close to the actual values. The coefficients in the formulas are set by those skilled in the art according to the actual situation. The above description is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any equivalent substitutions or changes made by those skilled in the art within the technical scope disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the protection scope of the present invention.

Claims

1. An ecological treatment method for livestock and poultry breeding waste, characterized in that, Includes the following steps: S1. Arrange diversion channels and selective infiltration grid devices in the livestock and poultry house, divide the diversion channels and the floor of the livestock house into several regular grid units, and collect the acoustic and vibration sequences of each grid unit. Extract time-frequency features from the voiceprint sequence to construct the short-time energy duration of each grid cell. Spectral centroid and bandpass energy ratio It determines whether the grid cell has a "urination acoustic event" or a "defecation acoustic event". If it does, it generates a first urination event label or a first defecation event label for the grid cell. Extract vibration features from the vibration sequence and construct the short-time vibration energy duration for each grid cell. and peak vibration and shock It determines whether the grid cell has a "urination vibration event" or a "defecation vibration event". If it does, it generates a second urination event label or a second defecation event label for the grid cell. If a grid cell at the same time generates either a first urination event label and a second urination event label, or a first defecation event label and a second defecation event label, it is identified as an actual urination event or an actual defecation event. After triggering the second identification command, each grid cell is constructed to record porosity ϕ, average pore size davg, permeability coefficient K, surface energy γ, local slope α, and flow resistance coefficient. The overall flow resistance coefficient R of each grid cell is calculated to evaluate whether the flow guidance of each grid cell is qualified. When the flow guidance is not qualified, a first blockage mark or a second blockage mark is generated. If several grid cells on the same vertical or horizontal path are identified, and all grid cells on the same vertical or horizontal path are identified as first qualified markers, then the negative pressure suction strategy is not triggered. When any grid cell is identified as either a first blocking marker or a second blocking marker, a corresponding first intervention strategy and a second intervention strategy are generated. After the corresponding first intervention strategy and second intervention strategy are executed, an improvement rate ImpRate is constructed for verification.

2. The method for ecological treatment of livestock and poultry breeding waste according to claim 1, characterized in that, Also includes: S2. Based on the actual urination event triggered in S1, collect the time series data of ammonia concentration of the grid cell and calculate the hot spot index (HPI) of the urination location. Set a hot spot threshold. If the hot spot index (HPI) is greater than or equal to the hot spot threshold, it is determined that the grid cell has a long-term urination aggregation effect, forming a "long-term focused urine point", and triggering the first urination location optimization sub-strategy. S3. While S1 and S2 are being executed, the scouring intensity parameters of urine impacting the ground are collected, and the self-cleaning scouring effect index Cclean for each "long-term focused urine point" is constructed. When Cclean≤0.7, it indicates that the self-cleaning ability of the urination point is unqualified, and a second urination position optimization sub-strategy is generated.

3. The method for ecological treatment of livestock and poultry breeding waste according to claim 1, characterized in that, Condenser microphones are arranged at intervals of 1.0 to 1.5 m on the ceiling or sidewall of each grid cell, at a height of 0.6 to 1.0 m above the ground, to collect the acoustic signature sequence of each grid cell. The acoustic signature sequence is preprocessed, including DC removal, pre-emphasis, and Hamming windowing, before time-frequency features are extracted. The time-frequency features include: short-time energy. Spectral centroid Compared with bandpass energy ratio ; Among them, short-time energy It involves dividing a continuous sound signal into short time windows, ranging from 20ms to 25ms, and calculating the energy within each window to obtain the short-time energy. : ； Where N is the number of sample points per frame, and 25ms corresponds to a sampling rate of 16kHz, which is 400 points. This is the preprocessed signal; Perform FFT spectral transformation on the same frame to obtain the amplitude spectrum of the same frame, and extract the spectral centroid. The specific formula is as follows: ； in, For the k-th frequency point, The amplitude; For the amplitude spectrum of the same frame, first calculate the total energy of the full-band signal, then calculate the energy within the target frequency band, which is set to 300Hz to 1500Hz. The ratio of the two is the bandpass energy ratio. The specific formula is as follows: ； Set a short-time energy threshold, if If the energy level is higher than the short-time energy threshold, then for consecutive frames... The durations are summed to obtain the short-term energy duration. The preferred range for the short-time energy threshold is set to 1×10. 4 Up to 5×10 4 ; During the 2-second statistical window, if any two of the following conditions are met, the voiceprint event is determined to be a "urination acoustic event," and a first urination event label is generated for that grid cell, including: Condition 1: Duration of short-time energy ≥0.3s; the sound of flowing water is continuous and steady; Condition 2: Spectral mass center Within the 300Hz to 1500Hz range and with fluctuations within the window ≤20%; Condition 3: Bandpass energy ratio ≥0.45; During the 2-second statistical window, if any two of the following conditions are met, the voiceprint event is determined to be a "fecal acoustic event," and a first fecal event label is generated for that grid cell, including: Condition 1: Duration of short-time energy <0.2s; Condition 2: Spectral mass center <400Hz; the spectral centroid is concentrated in the low frequency range <400Hz; Condition 3: Bandpass energy ratio <0.

45.

4. The method for ecological treatment of livestock and poultry breeding waste according to claim 1, characterized in that, Piezoelectric or strain gauge vibration sensors are arranged at the edge of each grid cell to collect vibration sequences at a sampling rate of 5kHz to 10kHz, with the sequence length being the number of sampling points within the event time window. The vibration sequence was divided into frames with a frame length of 25ms or a frame shift of 10ms. Each frame was preprocessed, including DC removal and Hamming windowing, to generate a preprocessed vibration frame signal V[m]. Vibration features were extracted, and the short-time vibration energy was obtained using the following formula. and peak vibration and shock : ； ； Where N is the number of sample points per frame, and short-time vibration energy is... Reflecting the vibration intensity of the frame, the sound of urination corresponds to the impact of flowing water, which produces continuous vibration, while the sound of defecation corresponds to an instantaneous impact; the peak value of the vibration impact. The impact amplitude is instantaneous, and the impact at the point of defecation is significantly higher than that of urine flow; Set the urination vibration threshold Evth to 1×10. −3 Up to 5×10 −3 V 2 ; Set the fecal discharge vibration threshold Pvth to 1×10. −2 Up to 5×10 −2 V; Short-term vibration energy The duration exceeding the urination vibration threshold Evth is denoted as the short-time vibration energy duration. ; Within a 2-second statistical window, conditions are identified and corresponding vibration events are determined, including: When the duration of short-term vibration energy If the time is ≥0.3s, it is determined to be a "urination vibration event", and a second urination event label is generated for the grid cell; When the peak vibration and impact > The fecal vibration threshold Pvth, and the duration of short-term vibration energy If the time is less than 0.2s, it is determined to be a "fecal vibration event", and a second fecal event tag is generated for the grid cell; The grid cells at the same time simultaneously generate the first urination event label and the second urination event label, and are identified as actual urination events; The grid cells at the same time simultaneously generate the first and second defecation event labels, and are identified as actual defecation events. A second identification instruction is triggered when an actual urination or defecation event is identified.

5. The method for ecological treatment of livestock and poultry breeding waste according to claim 1, characterized in that, When the second identification command is recognized, each grid cell is constructed to record the porosity ϕ, average pore size davg, permeability coefficient K, surface energy γ, local slope α, and flow resistance coefficient. ; Specifically as follows: S11. Calculate the proportion of pore volume within a grid cell to the total volume using 3D scanning to obtain the porosity ϕ: ； in, This represents the actual void volume within the grid cell. This represents the total volume of the grid cells; S12. Use laser scanning to obtain the aperture distribution, calculate the average equivalent diameter of the pores within the grid cell, and obtain the average aperture davg: ； in, Let be the equivalent diameter of the i-th hole, and n be the total number of holes in the unit. S13. Using Darcy's law, calculate the permeability of water or urine in the grid cell medium to obtain the permeability coefficient K: ； Where Q represents the penetration flow, Indicates the height of the grid cell, and A represents the cross-sectional area of ​​the grid cell. Indicates a negative pressure difference. Indicates the flow of time; S14. Measure the contact angle using the droplet method, calculate the wettability index of the grating material surface to the liquid, and obtain the surface energy γ: ； in, For liquid surface tension, Contact angle; S15. The tilt angle of the grid cell surface relative to the direction of liquid flow, obtaining the local slope α: ； Where Δh is the height difference of the grid cells, and Δx is the horizontal length; S16. The resistance of the grid cells to the fluid is used to obtain the flow resistance coefficient. : ； in, For unit pressure drop, Let ρ be the fluid density and v be the average flow velocity.

6. The method for ecological treatment of livestock and poultry breeding waste according to claim 5, characterized in that, For each grid cell, record the porosity ϕ, average pore size davg, permeability coefficient K, surface energy γ, local slope α, and flow resistance coefficient. Normalized to the [0,1] interval, and the overall flow resistance coefficient R of each grid cell is calculated using a weighted formula; The overall flow resistance coefficient R value of each grid cell is evaluated, including: When R < 0.3, it indicates that the grid cell is qualified for current conduction, and the first qualified mark is generated; When 0.3≤R≤0.6, it indicates that the flow guidance of the grid cell is unqualified, and the first blockage mark is generated; When R > 0.6, it indicates that the flow conduction of the grid cell is not qualified, and a second blockage mark is generated.

7. The method for ecological treatment of livestock and poultry breeding waste according to claim 6, characterized in that, If several grid cells on the same vertical or horizontal path are identified, and all grid cells on the same vertical or horizontal path are identified as first qualified markers, then the negative pressure suction strategy is not triggered. If any grid cell is identified as the first blockage marker, a first intervention strategy is generated, including: triggering negative pressure suction, performing a single-round suction, with a maximum negative pressure of −2.5 kPa for the single-round pulse, and the duration of the single-round pulse is set to 3 to 6 seconds; If any grid cell is identified as a second blockage marker, a second intervention strategy is generated, including: triggering negative pressure suction, performing multiple rounds of suction, with a maximum negative pressure of -3.0 kPa to 6.0 kPa per round pulse, a pulse duration of 8 to 10 seconds per round pulse, and performing 2 to 5 pulses per round pulse, with a pulse interval of 5 to 8 seconds.

8. The method for ecological treatment of livestock and poultry breeding waste according to claim 7, characterized in that, After implementing the first and second intervention strategies, a delay time trecheckt is set, and the new comprehensive flow resistance R2 of the grid cell is remeasured after 5 to 10 seconds, while the improvement rate ImpRate is calculated: ； Where R is the total flow resistance before intervention; If R2 < 0.3, mark it as "recovered", determine that the grid cell is qualified for flow diversion, and stop subsequent intervention; If the improvement rate ImpRate ≥ 15% and R2 ≤ 0.6, it is judged as "relief", indicating that although the drainage is not qualified, there is improvement. The unit is included in the first priority follow-up observation queue, and the intervention is not upgraded for the time being. The first intervention strategy is repeated once, that is, a single round of negative pressure aspiration: −2.5kPa, duration 3s to 6s. If the improvement rate ImpRate < 15%, it is judged as "insufficient intervention", and a third intervention strategy is generated, including: increasing the number of the second intervention of the grid unit by 1, that is, multiple rounds of negative pressure suction: −3.0kPa to −6.0kPa, 8s to 10s, 2 to 5 cycles, with an interval of 5s to 8s. If the improvement rate ImpRate is still <15%, it is determined to be a continuous blockage. The grid cell is then directly added to the "manual cleaning and replacement sequence", and a manual replacement priority sequence list is generated according to the improvement rate ImpRate from small to large.

9. A method for ecological treatment of livestock and poultry breeding waste according to claim 2, characterized in that, S2 include: S21. Based on the actual urination event triggered in S1, firstly, continuously collect time-series data of ammonia concentration changes over time within the target grid cell, and calculate the urination location hotspot index (HPI) of the grid cell according to the following formula: ； in, Let j represent the ammonia concentration at the j-th sampling point; j = 1, 2, ..., M, where j is the time series index. The time decay weighting coefficient is set as follows: ; S22. Set a hot spot threshold. When the hot spot index HPI at the urination location of the grid cell is greater than the hot spot threshold, it is determined that the grid cell has a long-term urination aggregation effect and is marked as a "long-term focused urine point". When the determination result is "long-term focus on urine point", the first urination position optimization sub-strategy is automatically triggered, including: A corrosion-resistant or hydrophobic protective film of 0.5mm to 2.0mm is laid in the flow channel within the grid unit, and an odor release device and fan are installed to drive odor diffusion, form a local odor gradient, and guide livestock and poultry to move towards the preset urine collection trough or flow channel; one to two odor release devices are arranged above or to the side of each preset urine point, with a height of 0.5m to 1m and a coverage radius of 1m to 2m.

10. The method for ecological treatment of livestock and poultry breeding waste according to claim 2, characterized in that, S3 include: S31. Pressure sensors are installed on the inner wall of the guide channel at each "long-term focused urine point". The collected pressure signal P(t) is a continuous sequence that changes with time, thus obtaining impact signal data. Short-time energy analysis and peak statistics are performed on the collected impact signal data to obtain scouring intensity parameters, including the instantaneous impact peak value Fpeak, the average scouring force Favg, and the scouring duration. ; Instantaneous peak impact Fpeak, average scouring force Favg, and scouring duration The method of obtaining it is: The pressure signal is divided into frames with a frame length of 25ms or a frame shift of 10ms. Short-time energy is calculated for the signal P[m] in each frame to obtain the short-time pressure energy. : ； Where: N is the number of sampling points per frame, and P[m] is the preprocessed pressure signal; And identify short-term energy of pressure. The instantaneous peak value and sustained amplitude of the impact are calculated using the following formulas to obtain the instantaneous impact peak value Fpeak and the average scouring force Favg: ； ； Among them, the maximum value of the pressure signal multiplied by the effective area of ​​the sensor This yields the instantaneous impact force, i.e., the instantaneous impact peak value Fpeak; To determine the flushing duration, a pressure threshold Pth is set; when the pressure is short, energy is applied. > Pressure threshold Pth indicates the start of effective flushing; calculate the short-time energy of continuous pressure. The duration of flushing is determined by the time the pressure threshold Pth is exceeded. S32. Normalize the flushing intensity parameters and calculate the self-cleaning flushing effect index Cclean for each "long-term focused urine point" using the following formula: ； in, Indicates the preset maximum impact force. This indicates the preset maximum flushing time, used to set the instantaneous peak impact value Fpeak, average flushing force Favg, and flushing duration. Perform normalization processing; , and This represents the weight coefficients, and the sum of the weights is 1; S33. If the self-cleaning flushing effect index Cclean > 0.7 for each "long-term focused urine point", it means that the self-cleaning ability of the urine point is qualified. If the self-cleaning flushing effect index Cclean ≤ 0.7 for each "long-term focused urination point", it indicates that the self-cleaning ability of that urination point is unqualified. Therefore, a second urination location optimization sub-strategy is generated, including: Turn off the odor emitter and fan drive that are currently not self-cleaning enough, and set up a low-frequency sound wave generator and LED flexible light source in the grid unit of the nearest "long-term focused urine point" with good self-cleaning ability near the current "long-term focused urine point" with insufficient self-cleaning ability. When the animal approaches the preset urine point, the sound and light unit emits a soft low-frequency sound, set to 500Hz to 1000Hz and visible light signal, to form a directional prompt to go to the "long-term focused urine point" with good self-cleaning ability, and assist the odor in guiding to the new long-term focused urine point, thus forming behavioral localization memory.

Images