Breeding system capable of accurately managing breeding cycle of sows

Abstract

The invention relates to the technical field of livestock breeding, and discloses a breeding system capable of accurately managing the breeding cycle of sows, which comprises an active piezoelectric platform module, a quasi-static fringe electric field detection array and a multi-mode signal processing unit. The active piezoelectric platform module is located at the bottom of the limiting fence, provides vertical mechanical excitation for the sow and monitors the body weight; the quasi-static fringe electric field detection array is located on the side wall, and a deep penetration non-contact detection electric field is constructed by using an active shielding technology. And the multi-mode signal processing unit cooperatively drives the mechanical excitation signal and the electromagnetic carrier signal, and extracts an intermodulation component synchronous with the mechanical frequency from the feedback of the quasi-static fringe field detection array. And the multi-mode signal processing unit calculates a weighted reproductive response index according to a demodulation result, and judges the reproductive cycle state of the sow by using a double-threshold logic tree. On the basis of the mechanical dielectric intermodulation principle, non-intrusive high signal-to-noise ratio monitoring of the oestrus state of the sow is realized by detecting the mechanical compliance change of biological tissues.

Description

Technical Field

[0001] This invention relates to the field of animal husbandry technology, specifically to a breeding system for precise management of the sow reproductive cycle. Background Technology

[0002] In large-scale farming, precise management of the sow reproductive cycle determines the farm's production efficiency and economic benefits. Accurately identifying the sow's estrus state and seizing the optimal mating time are key to shortening the sow's non-productive days and increasing the number of weaned piglets produced by the sow annually. In intensive, high-density farming environments, building an automated system that can operate stably for a long time and achieve real-time monitoring and data-driven management of the physiological state of individual sows has become a core requirement for promoting the upgrading and development of smart farming technology.

[0003] In existing technologies, estrus monitoring in sows relies on manual back pressure observation, contact electronic probe detection, or passive behavior monitoring systems based on machine vision and accelerometers. Manual back pressure observation depends entirely on the subjective experience and responsibility of the handler, resulting in low efficiency and a high risk of missed or incorrect detections during large-scale inspections. While rectal resistance probes or B-mode ultrasound can obtain some physiological data, frequent contact operations can cause strong stress responses in sows, damaging their health and increasing the risk of cross-infection with pathogens. Non-contact solutions based on machine vision or accelerometers reduce stress interference, but can only record the sow's external movement or changes in body contour, failing to penetrate the body surface barrier to obtain deeper physiological parameters directly reflecting the ovulation cycle. When sows exhibit abnormal activity due to environmental noise, disease, or fighting within the pen, single-dimensional behavioral data can lead to logical misinterpretations and fail to accurately reflect the intrinsic physiological changes in the sow's reproductive system.

[0004] Existing single-modal monitoring methods struggle to simultaneously quantify the unique standing reflex muscle rigidity characteristics and dielectric response changes caused by pelvic reproductive tract congestion in sows during estrus, all under stress-free conditions. Therefore, this invention proposes a breeding system for precise management of the sow reproductive cycle to address the shortcomings of existing technologies. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a breeding system for precise management of the sow reproductive cycle, which solves the problems of low accuracy, high stress response, and inability to quantify physiological state caused by relying on manual observation or a single contact sensor for sow estrus monitoring in existing technologies.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solution: a breeding system for precise management of sow reproductive cycle mainly consists of an active piezoelectric platform module, a quasi-static edge electric field detection array, and a multimodal signal processing unit.

[0007] At the system architecture level, the active piezoelectric platform module is positioned at the bottom of the gestation pen, responsible for supporting the sow's body and transmitting mechanical waves. The quasi-static edge electric field detection array is positioned inside the side panels of the gestation pen, responsible for constructing a non-contact detection electric field within the pen's interior space. The multimodal signal processing unit, serving as the core control and computation hub, establishes electrical connections with both the active piezoelectric platform module and the quasi-static edge electric field detection array.

[0008] This invention employs an active intermodulation detection principle combining mechanical excitation and electromagnetic detection. A multimodal signal processing unit collaboratively controls an active piezoelectric platform module to emit a vertically oriented mechanical excitation signal and controls a quasi-static edge electric field detection array to emit an electromagnetic carrier signal. The mechanical waves generated by the active piezoelectric platform module are conducted through the sow's hooves and skeletal system to the pelvic region, causing microscopic mechanical displacement of biological tissues and periodic vibrations of tissue interfaces. Simultaneously, the quasi-static edge electric field detection array senses edge electric field signals penetrating the sow's torso. The multimodal signal processing unit receives the feedback signal output from the quasi-static edge electric field detection array and extracts a signal component whose frequency is synchronized with the mechanical excitation signal frequency from the feedback signal using a signal processing algorithm. This synchronized signal component characterizes the dielectric response properties of the sow's reproductive tract tissue under mechanical wave modulation. Based on this, the multimodal signal processing unit calculates characteristic parameters and determines the sow's reproductive cycle status.

[0009] Regarding the specific structure of the active piezoelectric platform module, it adopts a layered structure along the vertical direction, comprising a rigid load-bearing cover, a piezoelectric transducer array, and a shock-absorbing base from top to bottom. The rigid load-bearing cover uses insulating material to block electrical interference and directly supports the sow. The piezoelectric transducer array generates vertical mechanical displacement under the action of a drive signal through a matrix arrangement of piezoelectric ceramic transducer units. The shock-absorbing base fixes the piezoelectric transducer array and blocks the upward transmission of ground vibrations. The active piezoelectric platform module is internally equipped with a dual-mode interface circuit, divided into a static weighing signal path and a dynamic excitation drive path. The static weighing signal path collects the charge generated by the gravitational polarization of the piezoelectric transducer array, and outputs a static voltage signal representing the weight after conversion and filtering; the dynamic excitation drive path receives the control signal from the multi-mode signal processing unit, amplifies it, and drives the piezoelectric transducer array to generate periodic mechanical displacement, realizing the multiplexing of weight monitoring and active excitation functions in a single module.

[0010] Regarding the detection principle of the quasi-static edge electric field detection array, the first and second electrode plate assemblies employ a multi-layer composite shielding structure. This multi-layer composite shielding structure includes a dielectric isolation window, a signal electrode layer, an insulating substrate layer, an active shielding layer, and an electromagnetic grounding shell. The system improves detection depth through active shielding technology, driving the potential of the active shielding layer to follow the potential of the signal electrode layer in real time. This eliminates the parasitic capacitance effect between the signal electrode layer and the electromagnetic grounding shell, forcing the electric field lines to extend into the target area inside the restraining pen instead of flowing directly from the detection electrode to the ground wire, ensuring that the edge electric field effectively covers the reproductive organs deep within the sow's pelvic cavity.

[0011] In terms of signal processing and demodulation logic, the multimodal signal processing unit utilizes a dual-channel direct digital frequency synthesizer to generate co-origin mechanical drive signals and electromagnetic carrier signals, ensuring phase synchronization between the two. The weak signal analog front-end performs transimpedance amplification, bandpass filtering, and gain adjustment on the weak displacement current induced by the azimuth-static edge electric field detection array. The quadrature phase-locked loop demodulation circuit executes a two-stage demodulation process to extract physiological characteristics: the first stage demodulation uses the electromagnetic carrier frequency as a reference to downconvert the signal to baseband, obtaining an envelope signal containing mechanical modulation information; the second stage demodulation uses the mechanical wave frequency as a reference to synchronously detect the envelope signal, separating the in-phase and quadrature components.

[0012] This invention quantifies the physiological state of sows using a specific algorithm. A central microprocessor within a multimodal signal processing unit calculates the reproductive response index. The reproductive response index is defined as the ratio of the dynamic intermodulation vector magnitude to the static carrier vector magnitude. The dynamic intermodulation vector magnitude is determined by the square root of the sum of the squares of the in-phase and quadrature components, characterizing the coupling strength between tissue mechanical compliance and dielectric constant; the static carrier vector magnitude is determined by the magnitude of the DC component demodulated in the first stage, characterizing the fundamental dielectric properties. Furthermore, the central microprocessor calculates a phase weighting factor using mechanical conduction phase lag to correct the reproductive response index, obtaining a weighted reproductive response index to compensate for differences in conduction paths caused by different sow body sizes.

[0013] In terms of the judgment logic, the system uses a dual-threshold logic tree to classify and interpret the weighted reproductive response index. When the weighted reproductive response index is greater than or equal to the estrus judgment threshold and meets the preset number of consecutive times, it is judged as the peak of estrus, corresponding to the muscle rigidity caused by the standing reflex and the enhanced mechanoelectric coupling caused by pelvic tissue congestion during the estrus period; when the weighted reproductive response index is between the estrus judgment threshold and the noise judgment threshold, it is judged as the non-estrus quiescent period; when the weighted reproductive response index is lower than the noise judgment threshold, it is judged as an abnormal or invalid measurement.

[0014] The technical solution of this invention achieves non-invasive, high signal-to-noise ratio detection of the deep physiological state of the sow's reproductive system by actively applying low-frequency mechanical excitation far from the natural resonance frequency of the organism and utilizing the changes in mechanical coupling factors caused by the changes in muscle tension in sows during estrus to modulate penetrating edge electric field signals.

[0015] This invention provides a breeding system for precise management of the sow reproductive cycle. It has the following beneficial effects: 1. This invention constructs an active detection mode based on the principle of mechanical dielectric intermodulation by coordinating the active piezoelectric platform module and the quasi-static edge electric field detection array through multi-modal signal processing unit. Using the mechanical excitation signal as the modulation source, the quasi-static edge electric field detection array can extract the signal component synchronized with the mechanical excitation frequency. This coherent detection mechanism greatly improves the signal-to-noise ratio of the system in complex breeding environments. The system utilizes the changes in mechanical coupling factors caused by muscle rigidity and pelvic tissue congestion due to the standing reflex phenomenon generated during estrus in sows to monitor the physiological state of the deep reproductive system in sows, overcoming the shortcomings of traditional body surface temperature monitoring or simple behavioral monitoring that cannot accurately reflect the internal ovulation cycle.

[0016] 2. This invention applies a multi-layer composite shielding structure combined with active equipotential driving technology in a quasi-static edge electric field detection array. By driving the active shielding layer to follow the potential of the signal electrode layer in real time, the parasitic capacitance between the signal electrode layer and the electromagnetic grounding shell is eliminated, forcing the edge electric field lines to extend into the target area inside the limiting fence. This increases the penetration depth of the detection electric field in the vertical direction, ensuring that the detection field can cover the reproductive organs deep in the sow's pelvic cavity. This avoids the measurement distortion caused by the electric field lines being concentrated only on the superficial skin or air path in non-contact detection, and ensures that the collected feedback signal can truly characterize the dielectric properties of the internal organs.

[0017] 3. This invention achieves the quantification and identification of reproductive status through a weighted reproductive response index algorithm and a dual-threshold logic tree determination strategy set within a multimodal signal processing unit. The system obtains basic indicators by calculating the ratio of dynamic intermodulation vector magnitude to static carrier vector magnitude, and introduces mechanical transmission phase lag to calculate phase weighting factors to correct the reproductive response index, thereby compensating for mechanical wave transmission errors caused by differences in body size among different sows. Combined with the dual-threshold logic tree, the weighted reproductive response index is divided into peak estrus, non-estrus quiescent period, and abnormal state. This multi-dimensional quantitative determination method improves the automation level and accuracy of sow reproductive cycle management. Attached Figure Description

[0018] Figure 1 This is a block diagram illustrating the overall system architecture of the present invention. Figure 2 This is a block diagram of the signal demodulation and feature extraction logic of the present invention; Figure 3 This is a schematic diagram of the reproductive status determination logic flow of the present invention.

[0019] 100. Active piezoelectric platform module; 102. Piezoelectric transducer array; 102a. Piezoelectric stack unit; 200. Quasi-static edge electric field detection array; 202a. Receiving electrode; 300. Multimodal signal processing unit; 310. Central main control microprocessor; 320. Dual-channel direct digital frequency synthesizer; 330. High-voltage power drive circuit; 340. Weak signal analog front end; 350. Quadrature phase-locked demodulation circuit; 400. Host computer management terminal. Detailed Implementation

[0020] The technical solutions in 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.

[0021] See attached document Figure 1 This invention provides a breeding system for precise management of the sow reproductive cycle, which is applied in a controlled environment such as a sow gestation pen or electronic feeding station. The system mainly includes: an active piezoelectric platform module 100, a quasi-static edge electric field detection array 200, and a multimodal signal processing unit 300. The active piezoelectric platform module 100 is disposed at the bottom of the gestation pen to support the sow and provide vertical mechanical excitation. The quasi-static edge electric field detection array 200 is disposed within the side panels of the gestation pen to construct a non-contact detection electric field covering the internal space of the gestation pen. The multimodal signal processing unit 300 is electrically connected to both the active piezoelectric platform module 100 and the quasi-static edge electric field detection array 200, and is used to control the transmission of excitation signals and process the collected feedback signals.

[0022] To clearly describe the spatial relationship and physical field direction of each module, a three-dimensional Cartesian coordinate system is established. The direction perpendicular to the ground plane is defined as the z-axis, with upward as positive; the direction parallel to the surface of the active piezoelectric platform module 100 and along the body length of the sow is defined as the x-axis; and the direction parallel to the surface of the active piezoelectric platform module 100 and perpendicular to the body length of the sow (i.e., the direction of the line connecting the two side panels) is defined as the y-axis.

[0023] The active piezoelectric platform module 100 includes a rigid bearing cover, a piezoelectric transducer array 102, and a shock-absorbing base. The rigid bearing cover is located on the top layer, and its upper surface is in direct contact with the sow's limbs. The rigid bearing cover is made of insulating material with high mechanical strength to ensure effective transmission of mechanical waves and isolate electromagnetic interference.

[0024] The piezoelectric transducer array 102, located between the rigid bearing cover and the vibration-damping base, consists of multiple piezoelectric ceramic transducer units arranged in a matrix. The piezoelectric transducer array 102 is configured to generate periodic mechanical displacement along the z-axis when receiving a drive signal. This periodic mechanical displacement is transmitted to the sow's hooves and skeletal system through the rigid bearing cover. The vertical displacement generated by the active piezoelectric platform module 100... Defined as:

[0025] in, This represents the vertical displacement of the platform surface at time t. The reference amplitude representing the mechanical excitation; Indicates the mechanical excitation frequency; The value range is set to be an interval that can cause micron-level vibrations in biological tissues without triggering stress responses in sows; It is set to a low-frequency band that can be transmitted through the bones to the pelvic region.

[0026] The shock-absorbing base is located at the bottom layer and is used to fix the piezoelectric transducer array 102 and block environmental vibration interference from the ground, ensuring that the mechanical waves generated by the active piezoelectric platform module 100 are transmitted unidirectionally upward.

[0027] The quasi-static edge electric field detection array 200 includes a first electrode plate assembly and a second electrode plate assembly. The first electrode plate assembly and the second electrode plate assembly are respectively embedded in the non-metallic sidewalls on the left and right sides of the restraint fence, and are arranged relatively parallel to each other in the y-axis direction. The height positions of the first electrode plate assembly and the second electrode plate assembly correspond to the torso and pelvic region of the sow when she is standing.

[0028] The first electrode plate assembly includes at least one set of transmitting electrodes, which are connected to the excitation output terminal of the multimodal signal processing unit 300. The second electrode plate assembly includes at least one set of receiving electrodes 202a, which are connected to the signal input terminal of the multimodal signal processing unit 300. The transmitting electrodes are configured to transmit a frequency of [frequency value missing] into the space inside the limiting barrier. The high-frequency electromagnetic carrier wave, the receiving electrode 202a is configured to sense the electric field signal transmitted through the space medium.

[0029] The multimodal signal processing unit 300 includes a dual-channel direct digital frequency synthesizer 320, a high-voltage power drive circuit 330, a weak signal analog front-end 340, and a quadrature phase-locked demodulation circuit 350. The dual-channel direct digital frequency synthesizer 320 is used to generate a mechanical excitation frequency of... The mechanical drive signal and the space transmission frequency are The electromagnetic carrier signal is used. A high-voltage power drive circuit 330 is connected between the dual-channel direct digital frequency synthesizer 320 and the active piezoelectric platform module 100 to drive the piezoelectric transducer array 102. A weak signal analog front-end 340 is connected to the receiving electrode 202a of the quasi-static edge electric field detection array 200 to amplify the weak electric field signal. A quadrature phase-locked demodulation circuit 350 simultaneously receives the output signal from the weak signal analog front-end 340 and the mechanical excitation from the dual-channel direct digital frequency synthesizer 320. The reference signal is used to extract the signal component that is synchronized with the mechanical excitation frequency.

[0030] In this system architecture, the active piezoelectric platform module 100 serves as the source of mechanical waves, and the quasi-static edge electric field detection array 200 serves as the transceiver of electromagnetic waves. The two are orthogonally distributed in space and work together through the timing control of the multi-modal signal processing unit 300 to jointly form a mechanical dielectric intermodulation detection environment for deep physiological tissues of sows.

[0031] See attached document Figure 1 The active piezoelectric platform module 100 provided by the present invention adopts a sandwich-type composite laminate structure in physical structure and a frequency domain multiplexing topology design in electrical connection, thereby realizing both static weight monitoring function and dynamic mechanical wave emission function.

[0032] The active piezoelectric platform module 100 is rigidly connected from top to bottom in the vertical direction to include: a rigid force transmission cover plate, a piezoelectric transducer array 102, and an acoustic damping base.

[0033] The rigid force transmission cover is a rectangular flat plate structure made of high-modulus insulating composite material. The upper surface of the rigid force transmission cover has anti-slip textures to provide friction when the sow is standing. The thickness of the rigid force transmission cover is determined based on the Young's modulus of the material to ensure that it does not deform when bearing the weight of the sow and to guarantee the uniform distribution of mechanical vibration energy on the surface of the rigid force transmission cover.

[0034] The piezoelectric transducer array 102 is located directly below the rigid force transmission cover plate. The piezoelectric transducer array 102 consists of M rows and N columns of piezoelectric stacked units 102a. Each piezoelectric stacked unit 102a is mechanically coupled to the rigid force transmission cover plate above and the acoustic damping base below by high-strength structural adhesive. The piezoelectric stacked unit 102a is formed by stacking multiple layers of PZT-5 (lead zirconate titanate) piezoelectric ceramic sheets in series along the polarization direction, which can obtain a large axial displacement under low voltage drive.

[0035] The acoustic damping base is made of high-damping rubber or polyurethane elastomer material. The acoustic damping base is configured to absorb the vibration energy radiated downward by the piezoelectric transducer array 102, prevent mechanical waves from leaking to the ground and causing energy loss, and at the same time isolate the background vibration noise transmitted from the external environment to the active piezoelectric platform module 100 through the ground.

[0036] To enable a single component to handle both static and dynamic forces, the active piezoelectric platform module 100 also includes a dual-mode interface circuit. The electrode leads of each piezoelectric stack unit 102a are connected to the dual-mode interface circuit. The dual-mode interface circuit includes a static weighing signal path and a dynamic excitation drive path.

[0037] The static weighing signal path includes a charge amplifier and a low-pass filter. When the sow stands on the rigid force-transmitting cover, her weight causes static or quasi-static mechanical compressive deformation in the piezoelectric stack unit 102a. According to the positive piezoelectric effect, polarization charge is generated on the surface of the piezoelectric stack unit 102a. The charge amplifier converts this polarization charge into a voltage signal, and a low-pass filter with a cutoff frequency below 10Hz filters out high-frequency interference, outputting a static voltage signal representing the sow's weight. .

[0038] The dynamic excitation drive path includes a high-voltage drive amplifier and an impedance matching network. When the system enters the active monitoring mode, the dynamic excitation drive path receives the excitation control signal from the multi-mode signal processing unit 300 and amplifies it into a high-voltage AC drive signal, which is then applied to both ends of the piezoelectric stack unit 102a. According to the inverse piezoelectric effect, the piezoelectric stack unit 102a generates mechanical expansion and contraction deformation at the same frequency as the drive voltage.

[0039] The axial displacement response of the piezoelectric stack element 102a under the action of an electric field follows the piezoelectric constitutive equation. Considering only the axial direction (i.e., the 33 direction, parallel to the z-axis) and neglecting lateral coupling, the axial strain equation of the piezoelectric stack element 102a is described as follows: ; in, Represents mechanical strain (dimensionless) along the polarization axis direction (i.e., the perpendicular direction). Represents the elastic compliance coefficient under a constant electric field (unit: m) 2 / N); This represents the mechanical stress (unit: Pa) along the polarization axis, which is mainly composed of the pre-tightening pressure generated by the sow's body weight. This represents the longitudinal piezoelectric strain constant (unit: m / V). This represents the intensity of the driving electric field applied along the polarization axis (unit: V / m).

[0040] During the active emission phase, the driving voltage applied by the dynamic excitation drive path... The driving electric field intensity generated inside the piezoelectric stack unit 102a With piezoelectric ceramic monolayer thickness The relationship is: ; Combining the two formulas above, the total vertical displacement generated by the piezoelectric stack unit 102a under the action of the driving voltage is... Represented as: ; Where n represents the number of piezoelectric ceramic sheets stacked in piezoelectric stack unit 102a. This formula shows that by controlling the driving voltage output of the dual-mode interface circuit... The amplitude and frequency of the signal can be precisely controlled to control the mechanical wave parameters generated on the surface of the active piezoelectric platform module 100, thereby injecting a controllable mechanical excitation signal into the sow's limbs. The isolation capacitor in the dual-mode interface circuit is used to prevent the DC component in the static weighing signal path from entering the dynamic excitation drive path, and the wave-blocking inductor is used to prevent the high-frequency high-voltage signal of the dynamic excitation drive path from damaging the low-voltage devices in the static weighing signal path.

[0041] See attached document Figure 1 The quasi-static edge electric field detection array 200 provided by the present invention adopts a non-contact capacitive sensing structure based on active shielding technology and is configured to construct a three-dimensional quasi-static electric field with a specific penetration depth within the physical space of the limit bar.

[0042] The quasi-static edge electric field detection array 200 is physically divided into a first sidewall sensing component and a second sidewall sensing component. The first sidewall sensing component is installed on the left side of the gestation stall, and the second sidewall sensing component is installed on the right side of the gestation stall. The vertical center height of the first and second sidewall sensing components is set to be flush with the anatomical position of the reproductive tract of an adult sow.

[0043] Both the first and second sidewall sensing components adopt a multi-layer composite shielding structure design, which includes, from the inside out: a dielectric isolation window, a signal electrode layer, an insulating substrate layer, an active shielding layer, and an electromagnetic grounding shell.

[0044] The dielectric isolation window is made of high-density polyethylene or polytetrafluoroethylene sheet, which has low dielectric loss and is corrosion resistant. The dielectric isolation window is used to prevent feces and water vapor inside the limit bar from directly contacting the internal circuit, while maintaining the continuity of the detection electric field at the dielectric interface.

[0045] The signal electrode layer is attached to the inner surface of the insulating substrate layer. The signal electrode layer is etched from highly conductive copper foil, and its geometry is arranged in a macroscopic comb-like interdigitated or segmented rectangular array pattern. In this embodiment, to maximize the detection depth, the signal electrode layer is configured as a pair with a specific spacing. and specific width The two electrodes are coplanar. One electrode is defined as the transmitting electrode, and the other electrode is defined as the receiving electrode, or the entire first sidewall sensing assembly is used as the transmitting end, and the entire second sidewall sensing assembly is used as the receiving end.

[0046] The active shielding layer is located on the outer surface of the insulating substrate layer, and its area completely covers and extends beyond the projected area of ​​the signal electrode layer. In terms of circuit connection, the active shielding layer is connected to the equipotential drive terminal of the preamplifier 303. The voltage potential on the active shielding layer is driven to follow the signal potential on the signal electrode layer in real time, but the two remain electrically isolated. This equipotential drive technology eliminates the parasitic capacitance between the signal electrode layer and the grounded metal fence behind it, forcing the emitted electric field lines to bend and extend forward into the target area (within the sow's body), forming an edge electric field, preventing leakage to the rear.

[0047] The penetration depth of the edge electric field generated by the quasi-static edge electric field detection array 200 in the direction perpendicular to the electrode plane (i.e., into the sow's torso) is controlled by the electrode geometry. The potential distribution function Φ(x,z) of the detection electric field follows the Laplace equation in space. For a geometric period length... The periodic finger-shaped electrode structure generates an electric field strength modulus of... The decay characteristics with vertical depth z approximately follow the following exponential decay law: ; in, This represents the magnitude of the electric field intensity at a vertical depth z from the electrode surface; denoted by , where represents the initial electric field strength on the electrode surface; k represents the electric field space wavenumber.

[0048] The electric field spatial wavenumber k is determined by the geometric period length of the signal electrode layer. The decision, its defined relationship is: ; Geometric period length Numerically, it is equal to the distance between the center lines of two adjacent electrodes of the same polarity. From the two formulas above, it can be seen that the effective penetration depth of the edge electric field (defined as the electric field intensity decaying to the surface intensity) is... (Depth of time) With geometric period length Proportional: ; To ensure that the detection electric field can effectively cover the reproductive organs deep in the sow's pelvic cavity (approximately 15 to 25 centimeters from the body surface), the geometric period length of the quasi-static edge electric field detection array 200 is [not specified]. It is designed to be no less than 50 cm. Specifically, in this embodiment, the signal electrode layer within the first sidewall sensing component is configured as a single large-area electrode plate, forming a quasi-parallel mutual capacitance detection structure with the opposite second sidewall sensing component. At this time, the geometric period length... The electric field approaches infinity mathematically, transforming the electric field mode from a surface edge field to a volume-penetrating field. This overcomes the limitation of traditional capacitive sensors that can only detect the epidermis, enabling direct sensing of the dielectric properties of deep tissues in sows.

[0049] See attached document Figure 1 The multimodal signal processing unit 300 provided by the present invention serves as the control core of the entire monitoring system. It is configured to accurately synchronize mechanical excitation signals and electromagnetic detection signals in the time and frequency domains and perform hardware-level signal demodulation operations.

[0050] The multimodal signal processing unit 300 mainly includes the following components in its electrical architecture: a central main control microprocessor, a dual-channel direct digital frequency synthesizer 320, a high-voltage power drive circuit 330, a weak signal analog front-end 340, and an orthogonal phase-locked demodulation circuit 350.

[0051] The central control microprocessor employs a digital signal processor or field-programmable gate array (FPGA) with floating-point arithmetic capabilities. It establishes data communication connections with the dual-channel direct digital frequency synthesizer 320 and the quadrature phase-locked demodulation circuit 350 via an internal high-speed bus. The central control microprocessor is configured to issue frequency control words, phase control words, and amplitude control words to set the mechanical excitation frequency. and electromagnetic carrier frequency .

[0052] The dual-channel direct digital frequency synthesizer 320 includes a first signal output channel and a second signal output channel. The first signal output channel is connected to the input terminal of the high-voltage power drive circuit 330. The output frequency of the first signal output channel is the mechanical wave frequency. The second signal output channel is connected to the transmitting electrode of the quasi-static edge electric field detection array 200, and the output frequency of the second signal output channel is the electromagnetic carrier frequency. The radio frequency carrier signal. The dual-channel direct digital frequency synthesizer 320 uses the same reference clock source to ensure that the signal output from the first signal output channel and the signal output from the second signal output channel maintain a strict phase correlation in the time domain. This is a necessary prerequisite for subsequent phase demodulation.

[0053] A high-voltage power drive circuit 330 is connected between the dual-channel direct digital frequency synthesizer 320 and the active piezoelectric platform module 100. The high-voltage power drive circuit 330 includes a linear power amplifier and a step-up transformer. Because the piezoelectric transducer array 102 exhibits capacitive load characteristics, the high-voltage power drive circuit 330 is designed to have a high voltage swing and low output impedance, driving the piezoelectric transducer array 102 to generate sufficient mechanical displacement while avoiding waveform distortion.

[0054] The input of the weak signal analog front-end 340 is directly connected to the receiving electrode 202a of the quasi-static edge electric field detection array 200. The weak signal analog front-end 340 sequentially includes a transimpedance amplifier, a bandpass filter, and a variable gain amplifier. The transimpedance amplifier converts the high-impedance displacement current signal (in the microampere or even nanoampere range) induced by the receiving electrode 202a into a low-impedance voltage signal. The center frequency of the bandpass filter is set to the electromagnetic carrier frequency. The passband width is set to at least twice the mechanical excitation frequency (i.e., ), ensuring that the sideband signals contain physiological information It can pass through without loss while filtering out power frequency interference (50 Hz / 60 Hz) and low-frequency baseline drift.

[0055] The quadrature phase-locked loop (QLL) demodulation circuit 350 is connected between the output of the weak signal analog front-end 340 and the data input of the central main control microprocessor. The specific circuit implementation of the QLL demodulation circuit 350 includes a first-stage mixer, a second-stage mixer, and a low-pass filter bank. The QLL demodulation circuit 350 receives a carrier synchronization signal from the dual-channel direct digital frequency synthesizer 320 as a first reference signal, and a mechanical synchronization signal from the dual-channel direct digital frequency synthesizer 320 as a second reference signal.

[0056] The quadrature phase-locked demodulation circuit 350 performs a two-stage demodulation process: the first stage demodulation utilizes the first reference signal. The weak signal is down-converted to baseband from the high-frequency modulated signal output by the analog front-end 340, and the envelope signal containing mechanical modulation information is extracted; the second-stage demodulation utilizes the second reference signal. The envelope signal is synchronously detected. Based on the lock-in amplification principle, the quadrature lock-in demodulation circuit 350 outputs an in-phase component. and orthogonal components . In-phase component and orthogonal components The values ​​satisfy the following orthogonal projection relationship: ; ; in, This means that the frequency of the measured signal is exactly equal to... The amplitude of the component, Indicates the phase of the measured signal. This indicates the phase of the second reference signal. The central microprocessor reads the in-phase component. and orthogonal components Based on this, the amplitude characterizing the mechanoelectric response properties of sow reproductive tract tissue was calculated. and phase : ; Through the aforementioned hardware connection architecture, the multimodal signal processing unit 300 constructs a closed-loop physical field excitation and detection circuit, enabling the system to extract the weak dielectric disturbance characteristics induced by active mechanical waves from the high-noise aquaculture environment.

[0057] See attached document Figure 2 This section details how the low-frequency mechanical waves generated by the active piezoelectric platform module 100 are transmitted within the sow's body, and establishes a mathematical mapping relationship between muscle physiological state and mechanical conduction characteristics.

[0058] In the z-axis direction perpendicular to the ground, the hindquarters structure of the sow (including hind limb bones, joint soft tissues and pelvic organs) is equivalent to a single-degree-of-freedom or multi-degree-of-freedom coupled mass, spring, and damping mechanical vibration system. The input end of this mechanical vibration system is the hoof part in contact with the active piezoelectric platform module 100, and the output end is the deep pelvic reproductive tract tissue covered by the quasi-static edge electric field detection array 200.

[0059] When the active piezoelectric platform module 100 applies a frequency of During forced displacement excitation, mechanical waves propagate upwards along the skeletal conduction chain. Definition This represents the vertical displacement of the platform. This represents the vertical response displacement of the pelvic tissues. According to mechanical vibration theory, the mechanical transfer function from the platform to the pelvis... Defined as the ratio of output amplitude to input amplitude.

[0060] The mechanical impedance characteristics of biological tissues are primarily determined by the recovery stiffness of muscles. In sow reproductive cycle management, the standing reflex, unique to estrus, leads to tonic contraction of the longissimus dorsi and biceps femoris muscles. This physiological change is represented in physical models by the equivalent stiffness coefficient of biological tissues. The increase and equivalent damping coefficient The relative change.

[0061] Steady-state response amplitude of pelvic tissues With the amplitude of the platform excitation The relationship between them is described by the following displacement transmissibility formula: ; in, Indicates the steady-state response amplitude of pelvic tissues (unit: meters); (This refers to the amplitude of the excitation applied by the active piezoelectric platform module 100, in meters). The equivalent stiffness coefficient of the biological tissues of the sow's hindquarter support system (unit: Newton / meter) under non-estrus relaxation state. At a low value; during estrus tonic state, Significantly increased; The equivalent damping coefficient of the sow's hindquarter support system (unit: Newton-second / meter) reflects the ability of soft tissue to dissipate mechanical energy. Indicates the equivalent mass of the rear body involved in the vibration (unit: kilogram); Indicates the frequency of mechanical excitation (unit: radians / second).

[0062] To maximize the signal difference between estrus and non-estrus states, the system is set to apply mechanical excitation at a frequency of [frequency value missing]. Selected in the low-frequency region far from the natural resonant frequency of the organism (usually much smaller than) Under these conditions, the above formula can be simplified by introducing a mechanical coupling factor that describes the degree of muscle tension. : ; in, Represents the physiological tension of muscles; when sows are not in estrus, their muscles are relaxed, and the equivalent stiffness coefficient of biological tissue is [missing information]. The smaller size and higher soft tissue damping result in a lower mechanical coupling factor. The value is low, and the mechanical wave is significantly attenuated during upward propagation, with almost no follow-up displacement of the pelvic tissue; when the sow enters the peak of estrus, the standing reflex caused by the influence of estrogen levels puts the muscle fibers in a state of high tension, and the equivalent stiffness coefficient of biological tissue is low. A sharp increase occurs, and the body structure behaves as a near-rigid body, leading to a mechanical coupling factor. The value increases significantly, approaching 1. At this point, the micron-level vibration of the active piezoelectric platform module 100 can be efficiently and non-destructively transmitted to the pelvic reproductive tract, driving the reproductive tract tissue to perform synchronous micro-displacement in the electric field.

[0063] This model reveals the physical essence of how the present invention can distinguish between lying down and standing estrus: the former blocks the mechanical wave transmission channel, while the latter establishes a mechanical wave transmission channel. This difference determines the presence or absence of the subsequent electric field modulation signal.

[0064] See attached document Figure 2This section elaborates on how mechanical vibrations are converted into characteristic sidebands in electrical signals, revealing the microscopic physical mechanism of the mechanical and dielectric intermodulation effect.

[0065] Within the detection space constructed by the quasi-static edge electric field detection array 200, the pelvic region of the sow is regarded as a non-uniform dielectric distribution field, which is mainly composed of two parts: background tissue (including fat, bone and intestinal gas) and target tissue (i.e. reproductive tract and uterine horn).

[0066] The background structure has a relatively low and stable dielectric constant. During estrus, the target tissue experiences a significant increase in water content due to physiological congestion and tissue fluid infiltration (i.e., edema), resulting in a dielectric constant that is significantly higher than that of the background tissue. The dielectric contrast between the two is defined as .

[0067] When the mechanical waves excited by the active piezoelectric platform module 100 are conducted to the pelvic cavity via the bones, the target tissue, surrounded by background tissue, executes at a frequency of [frequency missing]. The periodic forced displacement, although its amplitude is on the order of micrometers, results in a non-uniform edge electric field generated by the quasi-static edge electric field detection array 200, leading to varying electric field strengths. As the spatial location changes drastically, the positional fluctuations of the target tissue in a non-uniform electric field cause periodic changes in the equivalent capacitance between the detection electrodes.

[0068] According to the theory of electrostatic field perturbation, the instantaneous change in equivalent capacitance caused by the micro-displacement of the target tissue is... It can be described as: ; in, The volume integral domain representing the target tissue (i.e., the displaced edematous reproductive tract); This represents the displacement vector of a mass point within the target tissue. This displacement vector is controlled by the mechanical transmission characteristics; The sensitivity field gradient represents the detection electric field. It describes the weighted distribution of the contribution of the change in dielectric constant at a point in space to the total capacitance. For edge electric field sensors, this gradient has a non-zero value in the region in front of the electrode.

[0069] The above formula shows that the instantaneous change in equivalent capacitance The dielectric contrast is With mechanical displacement amplitude The product of . This means that only when there is edema in the reproductive tract (high) ) and muscle rigidity transmission ( When they occur simultaneously, Only then can it have a clearly detectable amplitude.

[0070] The receiving electrode 202a of the quasi-static edge electric field detection array 200 is connected to a preamplifier with high input impedance, and the system converts the change in equivalent capacitance into a change in phase. Let the transmitted carrier be... Then receive the signal The phase term contains the components of The resulting modulation component.

[0071] Phase modulation function of the received signal Expressed as: ; in, This indicates the reference phase delay determined by the stationary structure; This indicates phase lag caused by mechanical wave propagation; Defined as the intermodulation depth coefficient.

[0072] Intermodulation depth coefficient This is the core physical quantity for state determination in this invention, and its analytical expression is: ; in, These are constants related to the electrode geometry; The average equivalent dielectric constant of the probe region; It is the effective characteristic scale of the electric field.

[0073] This formula illustrates the dynamic dielectric perturbation mechanism: mechanical vibration. As a physical carrier, the dielectric contrast of the reproductive tract is It is modulated into the phase spectrum of electromagnetic waves. When the sow is in estrus, (Due to edema) and (Due to the simultaneous increase in rigidity) the intermodulation depth coefficient is increased. It exhibits nonlinear enhancement, thereby generating high-amplitude sideband signals in the spectrum. This enables specific identification of estrus status.

[0074] See attached document Figure 2 Based on the dynamic dielectric perturbation mechanism, this section establishes an analytical mathematical model of the electrical signal at the receiving end of the quasi-static edge electric field detection array 200, and demonstrates the manifestation of the mechanical dielectric intermodulation effect in the frequency domain.

[0075] The raw analog voltage signal acquired by the receiving circuit in the multimodal signal processing unit 300 It is an undemodulated radio frequency signal, the original analog voltage signal. It includes not only the fundamental component of the transmitted carrier wave, but also the phase modulation component caused by the mechanical vibration of biological tissue.

[0076] The unmodulated carrier electric field signal emitted by the quasi-static edge electric field detection array 200 is set. for: ; in, This represents the peak intensity of the emitted electric field. The angular frequency of an electromagnetic carrier wave ( ).

[0077] When the carrier wave passes through the pelvic tissue of a sow undergoing forced vibration at an angular frequency, the receiver obtains the original analog voltage signal according to the dynamic dielectric perturbation mechanism. This manifests as a phase-modulated waveform. The instantaneous phase of this waveform no longer increases linearly. Instead, it is superimposed with a phase offset term that varies sinusoidally with time, the original analog voltage signal The time-domain expression is: ; in, The system link transmission gain coefficient represents the total conversion gain of the receiving electrode 202a and the preamplifier 303. The static phase delay (unit: radians) is determined by the reference distance between the transmitting electrode and the receiving electrode 202a and the average dielectric constant within the limit bar in the static state. The intermodulation depth coefficient (unit: radians) is the modulation index that reflects the dual characteristics of reproductive tract edema and muscle rigidity. Forced displacement excitation applied to the active piezoelectric platform module 100; The phase lag angle (unit: radians) caused by the propagation of mechanical waves within biological tissue.

[0078] To reveal the frequency domain characteristics of this signal, the original analog voltage signal was analyzed using a Bessel function of the first kind. Perform a Jacobi-Anger expansion, taking into account the intermodulation depth coefficient. It is usually present in small amounts in living organisms. The expansion retains only the zeroth-order and first-order terms, ignoring higher-order terms. The expanded received signal spectrum model. Represented as:

[0079] in, The value of the zeroth-order Bessel function of the first kind corresponds to the frequency of The amplitude weights of the carrier components, due to Very small This indicates that most of the energy is still concentrated at the carrier frequency, reflecting the static dielectric properties of the background structure. The value of a first-order Bessel function of the first kind corresponds to a frequency of The amplitude weights of the intermodulation sideband components, when When smaller, ; and These represent the combined phase angles of the upper and lower sidebands, respectively, and their values ​​include the static phase delay. and mechanical hysteresis angle Linear combination information.

[0080] This modulation model indicates that the sow's physiological state information (included in...) (The term) does not exist in the angular frequency representing an electromagnetic carrier wave. The amplitude was not entirely transferred to a frequency of . and In a pair of symmetrical sideband signals.

[0081] Therefore, the core task of signal processing in this invention is to accurately separate, in the received spectrum, the frequencies located on both sides of the carrier frequency with a frequency offset exactly equal to the mechanical excitation frequency. The weak sideband signal is detected, and the sideband amplitude is calculated. With carrier amplitude The ratio. This ratio directly corresponds to the intermodulation depth coefficient. Half of: ; This ratio eliminates system gain. and launch intensity The effects of fluctuations provide an absolute measurement that is only related to the biophysical characteristics of the sow. This forms the theoretical basis for the phase-locked demodulation algorithm in the subsequent signal processing chapter.

[0082] See attached document Figure 2 In this embodiment, the main task of the signal acquisition and preprocessing stage is to convert the weak high-impedance displacement current signal sensed by the receiving electrode 202a into a voltage signal with high signal-to-noise ratio and low impedance, and to complete the preliminary spectrum purification in the analog domain, so as to provide a high-quality input source for subsequent phase-locked demodulation.

[0083] The signal acquisition and preprocessing process is executed by the weak signal analog front-end 340, which contains, in sequence, a low-noise transimpedance amplifier, an active bandpass filter, and a programmable gain amplifier.

[0084] Current-to-voltage conversion and impedance matching: The receiving electrode 202a is directly connected to the inverting input of the low-noise transimpedance amplifier. Since the quasi-static edge electric field detection array 200 operates in non-contact mode, the equivalent coupling capacitance between the receiving electrode 202a and the transmitting electrode is extremely small. Therefore, the received signal is a displacement current source with high internal resistance.

[0085] The low-noise transimpedance amplifier uses a JFET input operational amplifier with a terahertz input impedance and a femtoampere input bias current to avoid loading the signal source. The low-noise transimpedance amplifier utilizes a feedback resistor. Input displacement current Converted to voltage signal .

[0086] Input displacement current The physical expression is: ; in, This represents the time-varying mutual capacitance modulated by the physiological activities of the sow; This indicates the transmitted carrier voltage.

[0087] Output voltage of low noise transimpedance amplifier for: ; To eliminate the shunting effect of parasitic capacitance in the connecting cable on the measurement, the low-noise transimpedance amplifier is also equipped with an active shielding drive circuit. The active shielding drive circuit buffers the output signal of the low-noise transimpedance amplifier with unity gain and then feeds it back to drive the shielding layer of the connecting cable, so that the potential of the shielding layer is synchronized with the potential of the core wire, thereby electrically eliminating the equivalent parasitic capacitance of the cable.

[0088] Frequency domain window selection and interference filtering: The output of the low-noise transimpedance amplifier is connected to an active bandpass filter. The active bandpass filter is configured as a fourth-order Butterworth structure to achieve a flat passband response.

[0089] Center frequency of active bandpass filter Strictly set to match the electromagnetic carrier frequency Consistent, the passband width of the active bandpass filter Set to cover at least 4 times the mechanical excitation frequency (Right now ( ), usually set to 1kHz to 5kHz.

[0090] Frequency response function of active bandpass filter The following conditions must be met: ; in, This indicates the power frequency interference frequency (50Hz or 60Hz). Due to the carrier frequency... (MHz level) far higher than power frequency interference frequency Active bandpass filters can provide a power frequency rejection ratio of over 80dB, filtering out common power line noise, motor electromagnetic interference, and low-frequency baseline drift in aquaculture environments, while retaining only carrier frequency band signals containing modulation information.

[0091] Dynamic range matching: The output of the active bandpass filter is connected to a programmable gain amplifier. Due to variations in sow size and random positioning within their pens, the amplitude of the received signal fluctuates over a wide range. The programmable gain amplifier receives automatic gain control commands from the central microprocessor.

[0092] The programmable gain amplifier adjusts its gain coefficient in real time according to the peak level of the input signal. The final output signal The amplitude is calibrated within the optimal linear input region of the quadrature phase-locked loop demodulation circuit 350 (e.g., 80% of full scale) to prevent clipping distortion or excessive quantization noise in subsequent circuits. The final output signal after preprocessing... for: ; in, This represents a temporal convolution operation, where... That is, the original analog voltage signal The physical implementation of it is a set of pure analog voltage signals containing only the carrier wave and its intermodulation sidebands.

[0093] See attached document Figure 3 In this embodiment, the multimodal signal processing unit 300 adopts a cascaded dual synchronous demodulation architecture, which aims to accurately extract the signal from a strong noise background caused by forced displacement excitation. The modulated weak sideband components enable control over the intermodulation depth coefficient. The quantitative measurement. The demodulation process mainly includes the first-stage carrier quadrature down-conversion stage and the second-stage modulation envelope synchronous detection stage.

[0094] First-stage carrier orthogonal downconversion stage: The quadrature phase-locked demodulation circuit 350 first receives the pre-processed input signal. The input signal Includes the angular frequency of the electromagnetic carrier wave. The carrier components and frequencies are The sideband components. To remove the high-frequency carrier, the quadrature phase-locked demodulation circuit 350 uses the first reference signal to modulate the input signal. Perform quadrature mixing.

[0095] They are simultaneously fed into two parallel multiplier branches; in the in-phase branch, With in-phase carrier reference signal Multiplication; in orthogonal branches, Orthogonal carrier reference signal Multiply.

[0096] The multiplier's output then passes through the cutoff frequency. Slightly higher than the mechanical excitation frequency (e.g.) The first-stage low-pass filter removes all harmonic components, retaining only the baseband signal. The resulting in-phase baseband signal is... and orthogonal baseband signal They are represented as follows: Central microprocessor and Perform arctangent operation to calculate the instantaneous phase function. : ; This formula shows that after the first stage of demodulation, the high-frequency electromagnetic carrier has been removed, and the system obtains a signal containing a DC component. and communication weight The phase signal, in which the AC component is the target physiological signal that we need to extract.

[0097] Second-stage modulation envelope synchronous detection phase: In order to Extracting the weak intermodulation depth coefficient After eliminating other frequencies of biological motion noise (such as breathing, heartbeat, or unforced limb movements), the system performs a second-stage lock-in amplification process.

[0098] The second-stage phase-locked demodulation utilizes the drive signal of the active piezoelectric platform module 100 as the second reference signal; this process is completed in the digital domain. The central control microprocessor processes the instantaneous phase function. Two parallel digital multipliers are fed in again, and channel A will... With digital reference waveform Multiply; Channel B will With digital reference waveform Multiply.

[0099] The result of the multiplication operation has an integration time constant of: The digital integrator performs cumulative averaging. According to the orthogonality principle, only when the frequency is strictly equal to... The signal component, after integration, produces a non-zero DC value, while the DC bias... All other frequencies of noise (including random white noise) approach zero over a long time integration.

[0100] Output value of channel A and the output value of channel B It is calculated using the following definite integral formula:

[0101] The final feature calculation is performed by the central microprocessor based on... and Two core physical quantities were calculated: Intermodulation depth modulus Intermodulation depth coefficient :

[0102] This value directly reflects the product effect of genital tract edema and muscle rigidity transmission.

[0103] Mechanical conduction phase lag :

[0104] This value reflects the viscoelastic characteristics of biological tissues.

[0105] Through the aforementioned dual phase-locked loop mechanism, tracking and detection with an extremely narrow bandwidth (typically less than 0.1Hz) is achieved, ensuring that only dielectric changes caused by mechanical waves of a specific frequency excited by the active piezoelectric platform module 100 are recorded by the system, thereby shielding the complex environmental interference at the aquaculture site.

[0106] See attached document Figure 3 This section details how to convert the physical quantity output by the previous phase-locked demodulation circuit into a dimensionless characteristic parameter, namely the reproductive response index, which is independent of the sow's individual body size and standing position.

[0107] The feature extraction algorithm module running inside the central microprocessor receives two sets of input data streams from the quadrature phase-locked loop demodulation circuit 350: the first set of data streams is the static carrier vector, which is the in-phase baseband DC component output from the first-stage demodulation. and orthogonal baseband DC component The first data stream consists of a static carrier vector, which characterizes the static dielectric coupling strength of the background structure within the current detection area. The second data stream is the dynamic intermodulation vector, derived from the final integration result of the second-stage demodulation output. and Composition. This vector characterizes the intensity of the dynamic dielectric perturbation induced by mechanical excitation.

[0108] To eliminate signal strength differences caused by the different standing positions of sows in the gestation crate (i.e., to eliminate path loss uncertainty), the central microprocessor first performs distance normalization calculation.

[0109] Reproductive Response Index It is defined as the ratio of the dynamic intermodulation vector magnitude to the static carrier vector magnitude. Its calculation formula is as follows:

[0110] in, Indicates calibration coefficients, used to map calculation results to a numerical range that is easy to display and process (e.g., setting...). ).

[0111] The reproductive response index was calculated. Subsequently, the central microprocessor further combines phase information to perform phase validity weighting, eliminating artifact signals generated by non-forced movements (such as the sow's active friction or impact with the fence).

[0112] Biological tissues exhibit specific viscoelasticity under forced vibration, with mechanical transmission phase hysteresis. It must fall within a specific physiological phase window Within (e.g., 30 to 60 degrees). The feature extraction algorithm module is based on the currently calculated mechanical conduction phase hysteresis. Calculate weighting factors :

[0113] in, This represents the phase weighting factor, with a value ranging from 0 to 1; This represents the center value of the physiological phase window; This represents the attenuation slope coefficient.

[0114] The weighted reproductive response index ultimately used for state determination The calculation is as follows:

[0115] Through the above calculation process, the weighted reproductive response index output by the multimodal signal processing unit 300 is... It is a highly robust eigenvalue that integrates amplitude intensity and phase consistency, and is valid only if the detected signal simultaneously satisfies two conditions: strong intermodulation amplitude (corresponding to genital tract edema and muscle rigidity) and specific phase hysteresis (corresponding to biological soft tissue damping characteristics). Only then will it show a high value.

[0116] The system establishes a classifier based on a double-threshold logic tree to process the calculated values. To make the judgment, the logic tree contains two key judgment thresholds: Threshold for determining estrus This value is set based on historical big data statistics and corresponds to the response level of sows in typical estrus (ankylosing spondylitis and edema).

[0117] Noise detection threshold This value is slightly higher than the ambient noise level and is used to distinguish biological signals from background noise.

[0118] The system executes the following classification logic: Scenario 1: Determined to be in peak estrus if the following conditions are met: The physical interpretation is as follows: A high-intensity intermodulation signal was detected, and the phase lag is consistent with biological tissue characteristics. This indicates that muscle rigidity (leading to efficient mechanical wave conduction) and reproductive tract edema (leading to dramatic fluctuations in dielectric constant) occurred simultaneously in the sow, which is a typical dual physiological characteristic of peak estrus.

[0119] System operation: Outputs estrus status flag, records the current timestamp, and prompts the optimal mating window.

[0120] Scenario 2: Determined to be in a non-estrus quiescent period, if the following conditions are met: Physical interpretation: A weak intermodulation signal was detected, indicating that although the sow was in situ and at rest, mechanical waves could be transmitted through her bones, but due to the lack of estrogen, there was edema in her reproductive tract. Muscle rigidity (large mechanical attenuation) caused by low or absent standing reflexes leads to insufficient intermodulation depth.

[0121] System operation: Output a non-estrus status flag.

[0122] Scenario 3: Determined as an abnormal or invalid measurement if the following conditions are met: The physical meaning is: no effective intermodulation signal was detected, which usually occurs when the sow is in an improper position (such as being completely on her side, causing the skeletal conduction pathway to be broken) or when the equipment is malfunctioning.

[0123] System operation: Discard this data and return to wait for a trigger again.

[0124] Step S403: To prevent false alarms caused by occasional interference, the system introduces a time smoothing mechanism. Only when the determination results are all in the peak estrus period for M consecutive times (e.g., 3 consecutive scan cycles) can the system officially send an estrus alarm confirming the status to the external management terminal through the communication interface.

[0125] Through the above-mentioned closed-loop logic, this system uses static triggering to select the optimal measurement time, uses active multiphysics to obtain deep physiological information, and uses dual threshold judgment to eliminate false positives, thereby achieving high-precision and automated monitoring of the sow reproductive cycle.

Claims

1. A breeding system for precise management of the sow reproductive cycle, characterized in that, It includes an active piezoelectric platform module (100), a quasi-static edge electric field detection array (200), and a multi-mode signal processing unit (300). The active piezoelectric platform module (100) is located at the bottom of the restraint bar and is used to support the sow and provide mechanical stimulation in the vertical direction to the sow's hooves and skeletal system. The quasi-static edge electric field detection array (200) is disposed in the two side wall panels of the limiting rail, and is used to construct a non-contact detection electric field covering the internal space of the limiting rail; The multimodal signal processing unit (300) establishes electrical connections with the active piezoelectric platform module (100) and the quasi-static edge electric field detection array (200), respectively; The multimodal signal processing unit (300) is configured to control the active piezoelectric platform module (100) to transmit mechanical excitation signals, and simultaneously control the quasi-static edge electric field detection array (200) to transmit electromagnetic carrier signals; The multimodal signal processing unit (300) receives the feedback signal sensed by the quasi-static edge electric field detection array (200), extracts the signal component that is synchronized with the frequency of the mechanical excitation signal from the feedback signal, obtains the characteristic parameters characterizing the mechanical dielectric response characteristics of the sow's reproductive tract tissue based on the signal component, and then determines the sow's reproductive cycle status.

2. The breeding system for precise management of the sow reproductive cycle according to claim 1, characterized in that, The active piezoelectric platform module (100) is provided with a rigid bearing cover, a piezoelectric transducer array (102) and a shock-absorbing base in the vertical direction from top to bottom; The rigid load-bearing cover is made of insulating material, and its upper surface is in direct contact with the sow. The piezoelectric transducer array (102) is composed of multiple piezoelectric ceramic transducer units arranged in a matrix, and the piezoelectric transducer array (102) is configured to generate periodic mechanical displacement in the vertical direction when a drive signal is received. The damping base is configured to fix the piezoelectric transducer array (102) and block environmental vibration interference from the ground, so that the periodic mechanical displacement is transmitted unidirectionally upward.

3. The breeding system for precise management of the sow reproductive cycle according to claim 2, characterized in that, The active piezoelectric platform module (100) also includes a dual-mode interface circuit, which includes a static weighing signal path and a dynamic excitation drive path. The static weighing signal path is configured to detect the polarization charge generated by the piezoelectric transducer array (102) due to the weight of the sow, convert the polarization charge into a voltage signal and filter out high-frequency interference, and output a static voltage signal representing the weight of the sow. The dynamic excitation drive path is configured to receive an excitation control signal from the multimodal signal processing unit (300), amplify the excitation control signal into an AC drive signal and apply it to the piezoelectric transducer array (102) to drive the piezoelectric transducer array (102) to generate the periodic mechanical displacement.

4. The breeding system for precise management of the sow reproductive cycle according to claim 1, characterized in that, The quasi-static edge electric field detection array (200) includes a first electrode plate assembly and a second electrode plate assembly; Both the first electrode plate assembly and the second electrode plate assembly adopt a multi-layer composite shielding structure, which includes, from the inside out: a dielectric isolation window, a signal electrode layer, an insulating substrate layer, an active shielding layer, and an electromagnetic grounding shell. The active shielding layer is connected to the equipotential driving terminal of the multimodal signal processing unit (300). The active shielding layer is driven to follow the signal potential of the signal electrode layer in real time, eliminating the parasitic capacitance between the signal electrode layer and the electromagnetic grounding shell, and causing the electric field lines to extend into the target area inside the limit bar.

5. The breeding system for precise management of the sow reproductive cycle according to claim 1, characterized in that, The multimodal signal processing unit (300) includes a dual-channel direct digital frequency synthesizer (320), a high-voltage power drive circuit (330), a weak signal analog front-end (340), and an orthogonal phase-locked demodulation circuit (350). The dual-channel direct digital frequency synthesizer (320) is configured to generate a mechanical drive signal with a frequency of mechanical wave frequency and an electromagnetic carrier signal with a frequency of electromagnetic carrier frequency, wherein the mechanical drive signal and the electromagnetic carrier signal share the same reference clock source. The high-voltage power drive circuit (330) is connected between the dual-channel direct digital frequency synthesizer (320) and the active piezoelectric platform module (100) to drive the active piezoelectric platform module (100). The weak signal analog front end (340) is connected to the quasi-static edge electric field detection array (200) and is used to convert the sensed weak displacement current signal into a voltage signal and perform filtering and gain amplification. The quadrature phase-locked demodulation circuit (350) is connected to the output of the weak signal analog front end (340) and is used to demodulate the pre-processed signal.

6. The breeding system for precise management of the sow reproductive cycle according to claim 5, characterized in that, The quadrature phase-locked demodulation circuit (350) is configured to perform a two-stage demodulation process: The first-stage demodulation uses the electromagnetic carrier frequency as the first reference signal to downconvert the signal output by the weak signal analog front-end (340) to the baseband and extract the envelope signal containing mechanical modulation information. The second-stage demodulation uses the mechanical wave frequency as the second reference signal to perform synchronous detection on the envelope signal, and outputs in-phase and quadrature components. The multimodal signal processing unit (300) calculates the intermodulation depth coefficient and mechanical conduction phase lag based on the in-phase component and the quadrature component.

7. A breeding system for precise management of the sow reproductive cycle according to claim 6, characterized in that, The weak signal analog front end (340) includes a low-noise transimpedance amplifier, an active bandpass filter and a programmable gain amplifier cascaded in sequence. The low-noise transimpedance amplifier is used to convert the high-impedance weak displacement current signal into a low-impedance voltage signal, and is equipped with an active shielding drive circuit to eliminate cable parasitic capacitance. The center frequency of the active bandpass filter is set to be consistent with the electromagnetic carrier frequency, and the passband width is set to cover at least twice the mechanical wave frequency, in order to retain sideband signals containing physiological information and filter out power frequency interference.

8. A breeding system for precise management of the sow reproductive cycle according to claim 6, characterized in that, The central main control microprocessor (310) within the multimodal signal processing unit (300) is configured to calculate the reproductive response index; The reproductive response index is defined as the ratio of the dynamic intermodulation vector magnitude to the static carrier vector magnitude; wherein, the dynamic intermodulation vector magnitude is obtained by calculating the square root of the sum of the squares of the in-phase component and the quadrature component, and the static carrier vector magnitude is obtained by calculating the magnitude of the DC component of the first-stage demodulation output; The central main control microprocessor (310) is also configured to calculate a phase weighting factor based on the mechanical conduction phase lag, and to use the phase weighting factor to numerically correct the reproductive response index to obtain a weighted reproductive response index.

9. A breeding system for precise management of the sow reproductive cycle according to claim 8, characterized in that, The multimodal signal processing unit (300) is configured to interpret the weighted reproductive response index based on a dual-threshold logic tree: If the weighted reproductive response index is greater than or equal to the preset estrus determination threshold and is satisfied for a preset number of consecutive times, then the sow is determined to be in the peak of estrus. If the weighted reproductive response index is less than the estrus determination threshold and greater than the preset noise determination threshold, then the sow is determined to be in a non-estrus quiescent period. If the weighted reproductive response index is less than or equal to the noise threshold, it is determined to be an abnormal or invalid measurement.

10. A breeding system for precise management of the sow reproductive cycle according to claim 1, characterized in that, The geometric period length of the quasi-static edge electric field detection array (200) is designed such that the effective penetration depth of the generated edge electric field in the vertical direction covers the reproductive organs deep in the sow's pelvic cavity. The frequency of the mechanical excitation signal generated by the active piezoelectric platform module (100) is set to a low frequency band that can be transmitted to the pelvic region through the sow's bones, and the frequency is far from the natural resonance frequency of the organism. The estrus state is identified by utilizing the change in mechanical coupling factors caused by muscle rigidity during estrus.

Images