Method and apparatus for penguin hatchling rotation monitoring

Abstract

The present application relates to the technical field of ecological monitoring, and provides a method and device for penguin hatching rotation monitoring, which comprises: for target penguins with subcutaneously implanted passive response chips storing unique identification ID, forming a near-field induction area covering only the nest center by an inductor arranged in the nest; recording the identification ID of the triggering event, the absolute triggering time and the received signal strength and converting them into instantaneous in-nest indication, and obtaining in-nest residence confidence through sliding window integration; generating in-nest interval sequence based on double-threshold hysteresis decision and missed reading bridging to filter out false triggers and compensate for occasional missed readings; for the same-nest parent individuals, taking the in-nest interval handover overlap length being not less than a preset threshold as a criterion to identify effective hatching rotation, and marking the no-individual-in-place interval as no-guarding event; and calculating the hatching rotation interval and off-nest foraging duration according to the standard format and outputting. The present application can accurately identify real hatching rotation and distinguish breeding interruption under dense colony interference.

Description

Technical Field

[0001] This invention relates to the field of ecological monitoring technology, and in particular to a method and apparatus for monitoring penguin hatching rotation. Background Technology

[0002] In monitoring polar marine ecosystems, penguin hatching rotation parameters are key indicators for assessing population reproductive effectiveness and marine food resource abundance. Current technologies for automated monitoring of penguin hatching rotation primarily rely on radio frequency identification (RFID) technology, broadly categorized into community-based and nest-based methods. However, community-based methods can only identify events where individuals cross community boundaries, failing to accurately attribute individual activity to specific nests and thus hindering the reconstruction of complete hatching rotation chains. While nest-based methods are deployed at nest locations, they typically equate single reads directly to visits or departures, neglecting to address the specific interference issues in densely populated breeding communities. In dense communities with nest spacing of only about 0.5 meters, weak coupling signals from adjacent nests or brief triggers from passing individuals are easily misinterpreted as hatching handover within the same nest, significantly increasing the error rate in identifying rotation events. Furthermore, passive transponder chips are prone to occasional missed reads when penguins are lying down, due to body obstruction, leading to the misinterpretation of continuous incubation periods. More importantly, existing methods do not use the overlap of parental interaction in the nest as a criterion, relying solely on the entry and exit records of individual individuals to determine the completion of the rotation. This makes it impossible to reliably distinguish between genuine hatching rotation and abandoned or empty nest events where one parent leaves the nest without a replacement. Consequently, interruptions in reproductive care are either missed or falsely reported as normal rotation. Furthermore, in extremely cold environments, it is difficult to balance low power consumption and high reliability. Therefore, there is an urgent need for a monitoring method that can accurately identify genuine hatching rotation and distinguish reproductive interruptions under conditions of dense colony disturbance. Summary of the Invention

[0003] Therefore, it is necessary to address the problem that existing technologies cannot accurately identify actual hatching rotations and distinguish reproductive interruptions under dense colony interference, and to provide a method for monitoring penguin hatching rotations, so as to achieve highly reliable automatic monitoring based on the parental overlap criterion.

[0004] This invention provides a method for monitoring penguin hatchery rotation, the method comprising: For target penguins with passive transponder chips implanted subcutaneously to store unique identifiers, sensors deployed in the nest periodically emit radio frequency excitation signals to form a near-field sensing zone that covers only the center of the current nest, and synchronize absolute timestamps; when the passive transponder chip enters the near-field sensing zone, is excited, and transmits the identifier ID back, the identifier ID, absolute trigger time, and received signal strength of each trigger event are recorded. The received signal strength of each triggered event is converted into the instantaneous in-nest indication of the individual to which the passive response chip belongs relative to the nest center, and the instantaneous in-nest indication is integrated over time within a sliding time window to obtain the in-nest residence confidence of the individual. Based on preset upper and lower thresholds, a hysteresis decision is made on the confidence of the in-nest dwell time to identify the arrival time and departure time of the individual. Adjacent triggering events with an interval less than the preset missed read gap tolerance within the in-nest interval are bridged to filter out false triggering events caused by passing through the nest and neighbor interference, and to compensate for the occasional missed reads of the passive response chip, thereby generating the in-nest interval sequence of the individual. Within a preset learning time window, based on the cumulative effective nesting time of each identifier ID and the overlap ratio of the nesting intervals of any two identifier IDs, the two identifier IDs with a mutual exclusion degree lower than a preset mutual exclusion threshold and the highest cumulative effective nesting time are identified as the parent individuals of the nest. The one currently maintaining nesting is called the in-situ individual, and the one entering the near-field sensing area is called the returning individual. When the overlap time of the nesting intervals of the returning individual and the in-situ individual is not less than a preset minimum handover overlap threshold, a valid hatching rotation is determined to have occurred. When the in-situ individual leaves the nest and there is no returning individual present, resulting in a gap in the nest where no individual is present, it is determined to be an unattended event and is marked separately, not included in the valid hatching rotation. The unattended event is classified as an empty nest event or an abandoned nest event based on its duration and whether it subsequently recovers its effective nesting status. Based on the identified valid hatching rotation events, the hatching rotation interval and the time spent foraging away from the nest are calculated and output according to the preset ecological monitoring standard data format.

[0005] In one embodiment, the sensor periodically transmits radio frequency excitation signals, including: transmitting radio frequency excitation signals with an adaptive duty cycle, specifically: The duration of time experienced by the individual after this departure from the nest. The independent variable is based on the individual's expected foraging time. and its dispersion Adaptively adjust the transmission period of the radio frequency excitation signal : , in, and These represent the preset maximum and minimum excitation periods, respectively. This represents the preset tracking time constant, and... Limiting The range; the expected foraging time. Updated using an exponentially weighted moving average: , in, This indicates the preset update weight. This indicates the duration of the individual's most recent foraging trip away from the nest; so that the sensor is excited at a high frequency within a preset short period after the individual leaves the nest and near its expected return time window, and at a low frequency during the middle of the foraging trip in the open ocean.

[0006] In one embodiment, converting the received signal strength of each triggered event into an instantaneous presence indication of the individual to which the passive acknowledgment chip belongs relative to the nest center includes: Based on the field-calibrated signal strength-distance empirical model, and compensated for the influence of ambient temperature on antenna readout characteristics, the estimated distance of the individual relative to the excitation antenna in the triggering event is calculated. : , in, and These represent the reference distance and its corresponding reference received signal strength, respectively. Indicates the current received signal strength. Indicates the empirical decay index, The amount of low-temperature detuning compensation is calculated using the following formula: , in, Indicates the temperature compensation coefficient. Indicates the calibration temperature. Indicates the current ambient temperature; based on the estimated distance and the preset near-field sensing zone radius. The instantaneous in-nest indication is calculated using a soft-boundary logistic function. : , in, This represents the preset boundary softening coefficient.

[0007] In one embodiment, the step of integrating the instantaneous in-nest indication over a sliding time window to obtain the in-nest residence confidence of the individual specifically involves: In length of Within the sliding time window, the instantaneous nest indication Perform time integration and normalization to obtain the individual. At any moment Nest dwell confidence : , in, This indicates the length of the sliding time window. Indicates the current moment. Represents the time variable of integration. Represents the individual At any moment The instantaneous indication in the nest; This ensures that the dwell confidence of individuals that are continuously lying down and incubating converges to one, while the dwell confidence of individuals passing through the nest is suppressed to a low level.

[0008] In one embodiment, the hysteresis decision on the nest dwell confidence based on preset upper and lower thresholds is used to identify the arrival and departure times of the individual, and to bridge adjacent triggered events within the nest dwell interval that are less than a preset missed read gap tolerance, including: Set upper threshold With lower threshold ,and Less than When the individual Nest dwell confidence The upper threshold is first crossed from below. When the individual arrives at the nest, the starting point of the current nesting interval is back-marked to the moment when the first instantaneous nesting indication is valid within that interval; When the confidence level of nest dwell is mentioned Crossing the lower threshold And within the preset tolerance for missed reading intervals If no valid trigger is obtained within the period, it is determined that the individual has left the nest, and the end point of the nesting interval is marked to the last moment when the nesting indication was valid within the interval; For intervals within the nested interval that are smaller than the missed read gap tolerance The adjacent triggering events are bridged to maintain the continuous nesting state.

[0009] In one embodiment, identifying the two identifiers with a mutual exclusion degree lower than a preset mutual exclusion threshold and the highest effective nesting time as the parent individuals of the nest includes: Within the preset learning time window, the peak retention confidence level should not be lower than the upper threshold. Furthermore, the peak value of the received signal strength is not lower than the preset threshold value for the main peak strength of the nested signal. The nesting interval is denoted as the valid nesting interval; The two identifiers with the highest cumulative effective nesting time and a mutual exclusivity relationship are identified as the parent individuals of the nest; the mutual exclusivity relationship is determined by the degree of mutual exclusivity. Characterization: , in, Indicates that the identifier ID is and The total effective overlap time of the two individuals in the nest interval. and These represent the effective cumulative time spent in the nest for both, and the mutual exclusion degree is... Less than the preset mutual exclusion threshold; The peak retention confidence level is lower than the upper threshold. Or the peak value of the received signal strength is lower than the nested main peak strength threshold. The identifier ID is used to identify individuals that have passed through the nest or are from neighboring nests, and these individuals are removed.

[0010] In one embodiment, determining that a valid hatching rotation has occurred when the overlap time between the returning individual and the in-situ individual in the nesting interval is not less than a preset minimum handover overlap threshold includes: When the parent birds are in power The nesting period has not yet ended, and the returning parent birds... The starting point of the nesting interval precedes the in situ parent bird. When the two appear at the end of the nesting interval, they overlap, and the overlap duration is... for: , in, Indicates the in-situ parent bird The end point of the nesting interval, This indicates the returning parent birds The starting point of the nesting interval; When the overlap duration Not less than the preset minimum overlap threshold And from the aforementioned parent birds in situ The returning parent birds after leaving the nest When an individual remains the sole in-situ individual, a valid hatching rotation is determined, and the rotation time is recorded. The time of departure of the parent birds in situ is specified as follows: .

[0011] In one embodiment, the calculation of the hatching rotation interval is: Parent birds The The duration of continuous incubation within the nest is used as the interval for a single incubation cycle, i.e., the incubation time within the nest. : , in, and Each represents the parent bird No. The start and end points of the nesting interval; the interval between two adjacent incubation cycles is equal to the duration of the corresponding nesting interval of the in-situ parent bird.

[0012] In one embodiment, the calculation of the hatching rotation interval and the time spent foraging away from the nest, and the output of the data according to a preset ecological monitoring standard data format, includes: Parent birds The time spent away from the nest between two adjacent nesting intervals is taken as the time spent away from the nest foraging. : , in, Indicates the first The starting point of the segment in the nest interval, Indicates the first The end point of the segment within the nest interval; Statistical period The percentage of time at least one parent bird is in the nest (i.e., the union length of the intervals between the two parents in the nest, obtained by subtracting the total overlapping time between the two parents from the sum of their total time in the nest according to the inclusion-exclusion principle) is used as the nest site protection rate. : , The hatching rotation intervals, foraging time away from the nest, and nest site protection rates, along with the corresponding individual identifiers (IDs) and absolute timestamps, are output according to the preset ecological monitoring standard data format. Specifically, the cumulative time both parents spend in the nest and the total overlapping time between parental interactions are counted based on the time falling within the statistical period. Nesting intervals that cross the boundaries of the statistical period are counted based on their intersection with the statistical period.

[0013] The present invention also provides an apparatus for monitoring penguin hatching rotation, for performing the method for monitoring penguin hatching rotation as described in any of the preceding claims, comprising: The passive response chip is configured to be implanted under the skin of the target penguin and stores a unique identifier ID inside. The sensor, installed in the nest, includes an excitation antenna, a radio frequency readout module connected to the excitation antenna, a microcontroller electrically connected to the radio frequency readout module, and an energy supply module, a real-time clock module, an ambient temperature module, a data storage module, and a data feedback module, all electrically connected to the microcontroller. The microcontroller is configured as follows: The excitation antenna transmits a radio frequency excitation signal to form a near-field sensing zone that covers only the center of the current nest and synchronizes the absolute timestamp. The received signal strength of each triggered event is combined with the output of the ambient temperature module to convert it into an instantaneous in-nest indication. The in-nest residence confidence is obtained by sliding window integration. Based on the dual threshold hysteresis decision and missed read bridging, an in-nest interval sequence is generated to filter out in-nest passage and adjacent false trigger events. Within a preset learning time window, based on the effective cumulative nesting time of each identifier ID and the overlap ratio of the nesting intervals of any two identifier IDs, the two identifier IDs with a mutual exclusion degree lower than a preset mutual exclusion threshold and the highest effective cumulative nesting time are identified as the parent individuals of the nest. The one currently maintaining its nesting status is called the in-situ individual, and the one entering the near-field sensing area is called the returning individual. When the overlap time of the nesting intervals of the returning individual and the in-situ individual is not less than a preset minimum handover overlap threshold, a valid hatching rotation is determined to have occurred. When the in-situ individual leaves the nest and there is no returning individual present, resulting in a gap in the nest where no individual is present, it is determined to be an unattended event and is marked separately. It is not included in the valid hatching rotation. The unattended event is classified into an empty nest event or an abandoned nest event according to its duration and whether it subsequently recovers its in-situ status. Based on this, the hatching rotation interval and the time spent foraging outside the nest are calculated, and the data is output by the data feedback module according to the preset ecological monitoring standard data format.

[0014] The aforementioned method and apparatus for monitoring penguin incubation rotation, by constructing a near-field sensing zone covering only the center of the nest, physically limits the sensing range. Combined with the instantaneous presence indication calculated from the received signal strength and the confidence level of presence obtained from integrating a sliding time window, it effectively distinguishes between continuous incubation while lying prone and brief passage through the nest or weakly coupled signals from adjacent locations, significantly reducing the false trigger rate in dense colony environments. Furthermore, a dual-threshold hysteresis decision mechanism is employed to identify the nest arrival and departure times, bridging adjacent trigger events smaller than the missed read gap tolerance. This avoids decision jitter caused by posture adjustments and tolerates occasional missed reads caused by the passive response chip's body obstruction, ensuring the continuity of the presence sequence within the nesting interval. Furthermore, this invention abandons the traditional approach of relying solely on a single reading event to determine the rotation, and instead uses the overlap duration of the two parent individuals in the nest area as the core criterion for identifying a valid hatching rotation. Only when the overlap duration between the returning individual and the in-place individual reaches a preset threshold is it confirmed as a genuine transition, thereby reliably distinguishing between normal hatching rotation and abandoned or empty nest events where one parent leaves the nest without replacement, filling the gap in the existing technology for identifying interruptions in breeding care.

[0015] The aforementioned technical effects are not simply the sum of the functions of each module, but rather an inseparable result of the synergistic effect of near-field physical isolation, confidence integration and hysteresis decision, missed read bridging, and handover overlap criteria. Specifically, without the physical limitation of the near-field sensing area, confidence integration and hysteresis decision at the signal processing level alone cannot effectively suppress strong interference from neighboring nests in dense colonies; without sliding window integration and hysteresis decision, near-field isolation alone cannot distinguish between the continuous lying down of parent birds in the same nest and their brief passage through the nest; without a missed read bridging mechanism, even if the pre-processing is correct, the continuous incubation process will still be incorrectly fragmented into multiple segments due to occasional missed reads from the passive chip, leading to statistical distortion of the rotation interval; without handover overlap criteria, even if the sequence within the nest interval is accurate, it is still impossible to distinguish between the actual handover of eggs between the parents and a brief brush past each other, let alone identify abandoned nests or empty nest events. The above-mentioned links constitute a complete anti-interference link from the signal layer to the decision layer. The absence of any link will lead to a significant degradation in the overall monitoring performance. The synergistic effect of each module enables this application to achieve performance that existing technologies cannot achieve in the three core technical effects of anti-interference in dense nesting areas, accurate identification of overlapping parent handover, and reliable differentiation between abandoned nests and rotation. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a schematic flowchart of a method for monitoring penguin hatching rotation according to an embodiment of the present invention; Figure 2 This is a timing diagram illustrating the confidence level of nest dwell and its delayed decision in an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the timing of parental incubation rotation and handover in an embodiment of the present invention; Figure 4 This is a schematic diagram of the device for monitoring penguin hatching rotation according to an embodiment of the present invention.

[0018] Figure label: 410 Passive Response Chip; 420 Sensor; 421 Excitation Antenna; 422 RF Readout Module; 423 Microcontroller; 424 Power Supply Module; 425 Real-Time Clock Module; 426 Ambient Temperature Module; 427 Data Storage Module; 428 Data Feedback Module; 430 Near-Field Sensing Zone. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, 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.

[0020] Unless otherwise defined, all technical and scientific terms used in this specification have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used in this specification includes any and all combinations of one or more of the associated listed items.

[0021] The following is combined with Figures 1-4 The present invention describes a method and apparatus for monitoring penguin hatching rotation.

[0022] like Figure 1 As shown, in one embodiment, a method for monitoring penguin incubation rotation is provided. This method automates the monitoring of penguin incubation behavior and includes the following steps: Step S110: For the target penguin with a passive response chip that has been implanted under its skin to store a unique identifier ID, the sensors deployed in the nest periodically emit radio frequency excitation signals to form a near-field sensing zone that only covers the center of the current nest, and synchronize the absolute timestamp; when the passive response chip enters the near-field sensing zone, is excited and sends back the identifier ID, the identifier ID, absolute trigger time and received signal strength of each trigger event are recorded.

[0023] By strictly limiting the sensor's radio frequency excitation range to the center region of the current nest, physical anti-interference isolation is created, ensuring that only individuals truly located above the egg cups of the nest can be reliably read. Parent birds in adjacent nests or individuals merely passing through the colony are attenuated in coupling signal strength because they are at the edge or outside the sensing area. This physical limitation of the near-field sensing area, combined with synchronous recording of absolute timestamps, provides a high signal-to-noise ratio raw trigger event stream for subsequent data processing, avoiding the neighbor misjudgment problem caused by overlapping sensing areas in traditional wide-coverage schemes.

[0024] Step S120: Convert the received signal strength of each triggered event into an instantaneous presence indication of the individual to which the passive response chip belongs relative to the nest center, and perform time integration on the instantaneous presence indication within the sliding time window to obtain the individual's presence confidence.

[0025] The system does not directly equate a single read event with in-nest behavior, but instead introduces a signal strength to confidence level conversion mechanism. Since the received signal strength varies with distance, converting it into an instantaneous in-nest indication quantifies the relative position of the individual to the nest center. Furthermore, by performing time integration within a sliding time window, discrete instantaneous indications are transformed into continuous dwelling confidence levels. The principle is that the signal from a parent bird continuously incubating has temporal continuity, and after integration, the confidence level converges to a high level; while the weakly coupled signal from a passing individual or neighboring nest only manifests as a brief pulse, and after smoothing through the sliding window, its confidence level is suppressed to a low level. Thus, a clear distinction is formed in the confidence level values ​​between genuine incubation behavior and instantaneous interference behavior.

[0026] Step S130: Based on the preset upper and lower thresholds, perform a hysteresis decision on the confidence of the in-nest dwell time to identify the arrival time and departure time of the individual, and bridge adjacent triggering events with an interval less than the preset missed reading gap tolerance within the in-nest interval to filter out false triggering events caused by passing through the nest and neighbor interference, and compensate for the occasional missed reading of the passive response chip, and generate the in-nest interval sequence of the individual.

[0027] Through this step, continuous confidence levels are transformed into discrete state intervals. A dual-threshold hysteresis decision logic is employed: when the confidence level rises above the upper threshold, it is determined that the penguin has arrived in the nest; when it falls below the lower threshold, it is determined that the penguin has left the nest. The hysteresis band between the upper and lower thresholds effectively absorbs signal fluctuations caused by penguin posture adjustments or brief standing up within the nest, preventing frequent jitter in state decisions. Simultaneously, for occasional read failures of the passive response chip due to penguin body obstruction or positional changes, the system sets a read gap tolerance. Adjacent trigger events with intervals less than this tolerance are treated as continuous maintenance of the same nesting process and bridged. This processing method eliminates false triggers from individuals not in the nest and ensures that the actual incubation interval is not incorrectly segmented due to signal interruptions, thus generating an interval sequence that accurately reflects the actual nesting time of an individual.

[0028] Step S140: Within a preset learning time window, based on the effective cumulative nesting time of each identifier ID and the overlap ratio of the nesting intervals of any two identifier IDs, the two identifier IDs with a mutual exclusion degree lower than a preset mutual exclusion threshold and the highest effective cumulative nesting time are identified as the parent individuals of the nest. The one currently remaining in the nest is called the in-situ individual, and the one entering the near-field sensing area is called the returning individual. When the overlap time of the nesting intervals of the returning individual and the in-situ individual is not less than a preset minimum handover overlap threshold, a valid hatching rotation is determined to have occurred. When the in-situ individual leaves the nest and there is no returning individual present, resulting in a gap in the nest where no individual is present, it is determined to be an unattended event and is marked separately, not included in the valid hatching rotation. The unattended event is classified into an empty nest or abandoned nest event according to its duration and whether it subsequently recovers its in-situ status.

[0029] The system leverages the biological characteristic of parents taking turns incubating to identify true mating pairs from numerous trigger records through temporal mutual exclusion. Based on this, it uses the overlap of time both parents spend in the nest as a necessary condition for effective egg transfer, rather than relying solely on single entry and exit records. Substantial egg transfer is only confirmed when the time spent simultaneously in the nest by the returning individual and the in-place individual reaches a preset minimum overlap threshold. Conversely, if no returning individual is detected after the in-place individual leaves the nest, leaving the nest unattended, the system independently marks this as an abandoned or empty nest event. This decision-making logic allows the monitoring system to distinguish between normal incubation rotation and interrupted reproductive care.

[0030] Step S150: Based on the identified valid hatching rotation events, calculate the hatching rotation interval and the time spent foraging away from the nest, and output the data according to the preset ecological monitoring standard data format.

[0031] After completing behavioral recognition, the system automatically calculates the incubation time in the nest for each effective rotation and the foraging time between two rotations. These indicators directly reflect the energy budget of the parent birds and the abundance of marine food resources. The output data includes incubation rotation statistics fields compatible with CEMP penguin monitoring indicators, including individual identifiers, absolute timestamps, average rotation intervals, average foraging time in the nest, and nest site guarding rate, facilitating correlation analysis with data from ecological monitoring programs.

[0032] In one embodiment, the sensor periodically transmits radio frequency (RF) excitation signals, including transmitting RF excitation signals with an adaptive duty cycle. Specifically, this adaptive duty cycle strategy does not use a fixed time interval for wake-up, but rather uses the duration experienced by the individual since leaving the nest. The independent variable is based on the individual's expected foraging time. and its dispersion Adaptively adjust the transmission period of the radio frequency excitation signal The calculation formula is as follows: , in, and These represent the preset maximum and minimum excitation periods, respectively. This represents the preset tracking time constant, and... Limiting Interval. This formula constructs a dual-incentive mechanism based on biological behavior prediction. The first term within the square brackets on the right side of the formula is an exponential function. The tracking weight represents the initial tracking weight when an individual has just left the nest. When the value is small, this term approaches 1, causing the excitation cycle to shrink to a minimum. This allows for high-frequency sampling to capture handover overlap events that may not have fully left the nest, preventing the loss of crucial rotation information due to delays in nest departure determination. As the nest departure time increases, this term rapidly decays, and the system enters a low-power mode. The second term within the square brackets on the right side of the formula is a Gaussian function. This represents the waiting weight for the expected return window. When the time spent away from the nest... Approaching the expected foraging time At this point, the Gaussian term approaches 1 again, causing the excitation cycle to contract again. This allows for intensive detection during the window when individuals are most likely to return to the nest. During the middle of the migration, i.e., the foraging phase at sea, both exponential functions approach zero, and the excitation period automatically widens to its maximum value. This significantly reduces the number of invalid wake-ups.

[0033] Preferably, to adapt to the physiological rhythms of polar penguins and the energy constraints of the extreme cold environment, the above parameters can be set within the following range: minimum excitation cycle. The value range is from 2s to 10s to ensure that second-level overlap can be distinguished; maximum excitation period The value ranges from 5 minutes to 30 minutes, maximizing the sleep time while ensuring no return-to-home events are lost; tracking time constant. The value ranges from 60s to 300s, used to adjust the duration of high-frequency sampling during the initial stage of nest departure, covering the typical handover and farewell behavior duration. This parameter configuration allows the sensor to maintain high sensitivity during critical biological event windows while maintaining extremely low power consumption during long non-event periods, mitigating the shortcomings of fixed-frequency excitation schemes that consume too much power during long-term monitoring.

[0034] In order to enable the system to adapt to fluctuations in foraging time caused by different individuals and environmental changes, the expected foraging time is... Updated using an exponentially weighted moving average: , in, This indicates the preset update weight. This represents the individual's most recently measured foraging time away from the nest. Through this update mechanism, the monitoring system can adjust the expected foraging time online. Update Weight The preferred value range is 0.2 to 0.4, which ensures... Able to respond promptly to the most recent measured value Changes in the foraging time (e.g., prolonged or shortened foraging time due to changes in marine food abundance) are observed, and sufficient historical memory is preserved to smooth out disturbances caused by single anomalous observations. Dispersion The standard deviation of the measured foraging time away from the nest can be used to dynamically adjust the width of the return-to-nest waiting window: when the individual's foraging behavior is highly regular... Smaller size, narrow and precise monitoring window; when behavior is highly random. The sensor's detection window is relatively large, and it automatically widens to prevent missed detections. Through this predictive stimulus scheduling, the sensor achieves a dynamic balance between energy consumption and data acquisition efficiency, which helps the device maintain a longer operating cycle and ensure data integrity in unattended, cold environments.

[0035] In one embodiment, converting the received signal strength of each trigger event into an instantaneous presence indication of the individual belonging to the passive acknowledgment chip relative to the nest center includes: calculating the estimated distance of the individual relative to the excitation antenna during the trigger event based on a field-calibrated signal strength-distance empirical model, and compensating for the influence of ambient temperature on antenna readout characteristics. Considering that drastic temperature changes in polar environments can cause impedance detuning of the sensor antenna and drift in readout sensitivity, directly using a signal-distance model calibrated at room temperature would result in significant ranging errors. Therefore, a temperature compensation term is introduced to correct for the received signal strength. The formula for estimating the distance is as follows: , in, and These represent the reference distance and its corresponding reference received signal strength, which are typically measured at the calibration temperature during system deployment. Indicates the current received signal strength; The empirical attenuation index, determined by field calibration, characterizes the signal attenuation characteristics with distance under near-field inductive coupling at 134.2 kHz (obtained through on-site measurement and fitting, rather than from a free-space radiation propagation model). The preferred value range is 2.0 to 4.0; in this embodiment, it is taken as... ; The amount of low-temperature detuning compensation is calculated using the following formula: , in, This represents the temperature compensation coefficient, used to quantify the signal strength deviation caused by a unit temperature difference, with a preferred value range of 0.02 dB / ℃ to 0.08 dB / ℃. Indicates the calibration temperature; in this embodiment, it is taken as... ; This represents the current ambient temperature. Through this compensation mechanism, when the ambient temperature is lower than the calibration temperature, the low-temperature detuning compensation ΔST(t) is positive. It is subtracted from the exponent of the estimated distance calculation formula to reduce the overestimation of distance caused by the decrease in antenna efficiency and signal attenuation due to low temperature. This ensures that the estimated distance calculated under different thermodynamic environments can truly reflect the physical distance between the individual and the nest center, avoiding the drift of the nest discrimination benchmark caused by seasonal or diurnal temperature differences.

[0036] It should be noted that the distance estimation method of the above-mentioned field-calibrated signal strength-distance empirical model is a preferred implementation method of this embodiment. Its model parameters must be calibrated on-site according to near-field inductive coupling characteristics. In other embodiments, the conversion of instantaneous nest presence indication can also adopt a field-calibrated signal quality-nest presence confidence lookup table model. That is, during the system deployment phase, received signal strength samples are collected under different temperatures, implantation depths, and body positions to establish a mapping table between signal strength and nest presence probability. During operation, the instantaneous nest presence indication is directly obtained through table lookup and interpolation, without relying on analytical distance estimation. Such alternative solutions can also convert received signal strength into an instantaneous nest presence indication reflecting the relative positional relationship between the individual and the nest center, without departing from the technical concept of this application.

[0037] After obtaining the estimated distance, based on the estimated distance and the preset near-field sensing zone radius... Instantaneous in-nest indication is calculated using soft-boundary logistic stipulation functions. : , in, This represents the preset boundary softening coefficient, preferably ranging from 0.03m to 0.08m. Unlike traditional hard threshold decision (i.e., 1 within the radius and 0 outside the radius), the logistic function constructs a continuous and smooth transition band. When the estimated distance is significantly smaller than... hour, A value approaching 1 indicates that the individual is stably located in the nest center region; when the estimated distance is significantly greater than 1... hour, A value approaching 0 indicates that the individual is far from the nest or is only a weakly coupled signal from neighboring nests. Boundary softening coefficient. The steepness of the transition zone was controlled, and the smaller Make the boundary sharper, larger This makes the boundary smoother. This soft boundary design can effectively absorb distance estimation noise caused by multipath effects or small changes in penguin position, prevent the instantaneous nest indication from frequently jumping between 0 and 1, and provide a stable input signal for subsequent integration processing.

[0038] Furthermore, by performing a time integral on the instantaneous in-nest indication within a sliding time window, the confidence level of an individual's in-nest residence is obtained. Specifically, within a time window of length... The sliding time window indicates the instantaneous nest indication. Perform time integration and normalization to obtain the individual. At any moment Nest dwell confidence The calculation formula is as follows: , in, Indicates the length of the sliding time window. Indicates the current moment. Represents the time variable of integration. Represents an individual At any moment The instantaneous nest presence indication is used. This integration process is equivalent to low-pass filtering, used to distinguish between "continuous incubation while lying down" and "brief nest-crossing" from a temporal perspective. For parent birds that are incubating, their instantaneous nest presence indication remains high for a long time, and the residence confidence level is obtained after integration and normalization. The convergence is 1; however, for individuals that merely pass by the nest or visit neighboring nests, their instantaneous nest presence indication is only a narrow pulse, which is significantly diluted after averaging through a sliding window, and the dwell confidence is suppressed to a low level. Sliding window length The selection of the length needs to consider two aspects: if it is too short, it cannot effectively filter out interference from passing paths; if it is too long, it will increase the lag in state recognition. Preferably, The value range is from 30s to 120s. In this embodiment, the value is... This duration covers the typical time it takes for an individual to cross the sensing area (usually less than 30 seconds) and is far less than the minimum duration for parent birds to incubate and rotate, thus maintaining reasonable temporal resolution while ensuring anti-interference capability. It should be noted that the instantaneous nest indication... Generate discretely based on trigger events; during the time interval between two adjacent valid trigger events, zero-order hold is performed using the most recent valid instantaneous in-nest indicator, with the hold duration limited by the read gap tolerance. The upper limit is the time elapsed since the last valid trigger. If no new valid trigger is obtained, then Set it to zero, so that the above continuous-time integral has a definite and reproducible value on the discrete read sequence.

[0039] Reference Figure 2 Based on preset upper and lower thresholds, a delayed decision is made on the confidence level of nest dwell time to identify the arrival and departure times of individuals, and adjacent triggered events within the nest dwell interval with an interval smaller than a preset missed read tolerance are bridged. In specific implementation, an upper threshold is set... With lower threshold ,and Less than When individuals Nest dwell confidence First time crossing the threshold from below When an individual arrives at the nest, the starting point of the current nesting interval is back-marked to the moment when the nesting indication is valid within that interval; when the confidence level of nesting stay is... Underpass lower threshold And within the preset tolerance for missed reading intervals If no valid trigger is obtained within the specified time, the individual is determined to have left the nest, and the end point of the nesting interval is marked to the last instantaneous moment within that interval when the nesting indication was valid. Upper threshold. The preferred value range is 0.6 to 0.8 (0.7 in this embodiment), and the lower threshold is... The preferred value range is 0.3 to 0.5 (0.4 in this embodiment). The hysteresis band between the upper and lower thresholds constitutes a hysteresis comparator, which can effectively shield the confidence fluctuations caused by penguins adjusting their posture or briefly getting up in the nest, preventing false state reversals during these normal incubation behaviors. At the same time, by backtracking the arrival time to the first valid indication time and setting the departure time to the last valid indication time, the phase lag caused by sliding window integration is eliminated, ensuring the accuracy of the nest time statistics.

[0040] In addition, to address occasional read failures of passive response chips caused by penguin body obstruction or changes in body position, a tolerance of read gaps smaller than the missed read gap is applied within the nesting area. Adjacent triggering events are bridged to maintain the continuity of the nested state. Missed read gap tolerance. The preferred value range is 5 min to 20 min. In this embodiment, the value is taken as... This mechanism acknowledges the physical limitations of passive RFID in biological applications, namely that missed reads during prone positions are probabilistic events rather than state changes. As long as the gap between two valid reads is shorter than... The system recognizes the gap as a false departure caused by signal obstruction and automatically fills it with a continuous incubation interval. This avoids incorrectly segmenting a complete incubation process into multiple fragmented segments, ensuring the integrity and reliability of the incubation rotation interval calculation. The combined use of these steps maintains a high level of confidence in discriminating penguin incubation behavior in complex polar environments.

[0041] In one embodiment, for parent individuals within the same nest determined based on time-exclusive relationships, the method includes: within a preset learning time window, ensuring that the peak retention confidence level is not lower than an upper threshold. Furthermore, the peak value of the received signal strength is not lower than the preset threshold value for the main peak strength of the nested signal. The nesting interval is defined as the effective nesting interval; the two identifiers with the highest cumulative effective nesting time and satisfying a time mutual exclusion relationship are identified as the parent individuals of that nest. Penguin breeding colonies are extremely densely distributed. Even if the sensor constructs a near-field sensing area that only covers the center of the nest, it may still weakly couple to parent birds in adjacent nests due to edge effects, or be triggered by non-parent individuals moving within the colony. To accurately locate the true pair of mates from complex triggering records, the system sets a preset learning time window, preferably ranging from 3 to 7 days, to accumulate sufficient behavioral samples to establish a baseline. Within this window, the system first filters the effective nesting interval through a dual threshold: the peak residence confidence level is not lower than the upper threshold. This ensured that the visit had a continuous incubation timeline, eliminating transient interference; the peak received signal strength was not lower than the in-nest main peak strength threshold. This ensures that the individual is indeed located in the core area of ​​the nest rather than on the edge of the sensing zone, thus eliminating weak signal interference from neighboring nests. Among all the identified IDs that passed the screening, the system automatically selects the two with the highest cumulative effective nest time as candidate parents, which is consistent with the biological principle that penguin parents share the incubation task.

[0042] Furthermore, the confirmation of candidate parents also requires a temporal mutual exclusion relationship, which is determined by the degree of mutual exclusion. Characterization: , in, Indicates that the identifier ID is and The total effective overlap time of the two individuals in the nest interval. and Represent the effective cumulative time spent in the nest and the mutual exclusivity of the two, respectively. The mutual exclusion threshold is lower than the preset threshold. Because penguins employ a strict rotational incubation strategy, except for brief handover moments, parents rarely appear in the nest simultaneously. Therefore, the mutual exclusion of true pairings should be maintained at an extremely low level. Conversely, if two individuals frequently appear in the nest simultaneously (e.g., due to misinterpretation of neighboring nest individuals due to location shifts, or social visits by non-pairing individuals), their mutual exclusion will significantly increase and they will be excluded. This behavioral pattern-based pairing verification mechanism, compared to identification methods that solely rely on signal strength, is better adapted to the complex electromagnetic environment and biological behavioral noise in dense colonies, ensuring the accuracy of subsequent rotation analysis.

[0043] After identifying the parent individuals, a valid hatching rotation is considered to have occurred when the overlap time between the returning individual and the in-situ individual in the nesting area is not less than a preset minimum handover overlap threshold. (Refer to...) Figure 3This criterion addresses the shortcomings of traditional nest location recording schemes, which directly equate a single read event with a visit or departure, and easily misjudge the brief passing of parents without actual egg exchange as a rotation. In this embodiment, when the in-situ parent birds... The nesting period has not yet ended, and the returning parent birds... The starting point of the nesting interval precedes that of the in-situ parent birds. When the two appear at the end of the nesting interval, they overlap, and the overlap duration is... for: , in, Indicating the presence of parent birds The end point of the nesting interval, Indicating the return of parent birds to their nest The starting point of the nesting interval. Only when the overlap duration is... Not less than the preset minimum overlap threshold And freely in the position of parent birds Parent birds returning to the nest after leaving A valid hatching rotation is determined only when the individual remains the sole in-situ entity, and the rotation time is recorded. The time of departure of the in-situ parent birds is marked as follows: Minimum overlap threshold The preferred physical time required for penguins to complete egg transfer is between 30 seconds and 300 seconds. This embodiment uses [the specified value]. Minimum overlap threshold The system mandates that both parents must coexist in the nest for a sufficient period of time to be considered a substantial transfer of hatching responsibility. If the overlap time is less than... Even if there are records of both parents appearing at the same time, they are considered invalid handovers or abnormal social behaviors and are not included in the valid rotation statistics, thus avoiding false rotation events caused by signal jitter or brief appearances together.

[0044] Specifically, when a parent bird leaves the nest without a returning individual, resulting in a gap in the nest's presence, the system classifies this as an unattended event and marks it separately, excluding it from the effective hatching rotation. This unattended event is further categorized as an empty nest event or an abandoned nest event based on its duration and whether effective nesting resumes subsequently. Specifically, when a parent bird is detected leaving the nest, if within a preset missed interval tolerance... No valid nesting indications for any individual were detected, and no subsequent events occurred that satisfied the nesting requirement. If conditions overlap, that unattended gap is marked independently. To avoid misjudging nest abandonment due to prolonged lying down obstructing readings, an unattended gap tolerance is used in the nest abandonment determination. The preferred setting is significantly greater than the missed read bridging tolerance. (For example Pick Several times or even more than the nest temperature drop, the confidence level of the last valid read showing a smooth decline rather than an abrupt break (and other supporting evidence), that is, only when the blank duration exceeds A nest abandonment event is defined as an event where there is no overlapping or overlapping of nesting areas that meet the conditions, and no individual re-establishes a valid nesting area; otherwise, the absence of a caregiver gap does not exceed [a certain threshold]. If an individual subsequently re-establishes a valid nesting area, it is considered an empty nest event (short-term lack of care) to distinguish it from voluntary nest abandonment. This logic strictly separates normal hatching rotation from interrupted reproductive care, preserving crucial temporal evidence for researchers to analyze the causes of reproductive failure (such as predation by natural enemies, voluntary nest abandonment due to environmental stress, etc.), and avoiding data contamination caused by the inability to distinguish between rotation and nest abandonment in previous monitoring systems.

[0045] Based on the identified valid hatching rotation events, the hatching rotation interval and the time spent foraging away from the nest are calculated, and the data is output according to the preset ecological monitoring standard data format. The specific calculation process is as follows: [The text abruptly shifts to a different topic] The The duration of continuous incubation within the nest is used as the interval for a single incubation cycle, i.e., the incubation time within the nest. : , in, and They represent the parent birds. No. The starting and ending points of the nesting interval; the interval between two adjacent incubation cycles equals the duration of the corresponding nesting interval for the in-situ parent bird. This indicator directly reflects the endurance and energy consumption level of the parent bird during a single incubation cycle. Simultaneously, the parent bird... The time spent away from the nest between two adjacent nesting intervals is taken as the time spent away from the nest foraging. : , in, Indicates the first The starting point of the segment in the nest interval, Indicates the first The segment marks the end point of the nesting interval. The duration of foraging away from the nest is a core parameter for assessing the abundance of marine food resources and the foraging efficiency of parent birds. Additionally, the statistical period... The percentage of time at least one parent bird is in the nest (i.e., the union length of the intervals between the two parents in the nest, obtained by subtracting the total overlapping time between the two parents from the sum of their total time in the nest according to the inclusion-exclusion principle) is used as the nest site protection rate. : .

[0046] Nest site protection rate quantifies the degree to which eggs are continuously protected. Theoretically, it should approach 1 during the incubation period, and the degree of deviation from 1 represents the proportion of time the egg is exposed to risk. Finally, the system outputs the incubation rotation intervals, foraging time away from the nest, and nest site protection rate, along with the corresponding individual identifier ID and absolute timestamp, according to a preset ecological monitoring standard data format (such as the CCAMLR Ecosystem Monitoring Program, CEMP standard). This standardized data output method allows the monitoring data obtained by this invention to be output in a format compatible with CEMP penguin monitoring indicators, facilitating correlation analysis and cross-project comparison with data from ecological monitoring programs. Specifically, the cumulative time parents spend in the nest and the total overlapping time are both counted based on the time falling within the statistical period; nesting intervals crossing the boundaries of the statistical period are counted based on their intersection with the statistical period. Specifically, each output record in the standard data format includes at least: nest number, individual identifier ID, event type, arrival time, departure time, rotation time, hatching rotation interval, foraging time away from the nest, nest site guarding rate, and the type and duration of unattended events.

[0047] like Figure 4 As shown, in one embodiment, a device for monitoring penguin hatching rotation is provided. This device serves as the physical carrier of the monitoring methods described in the preceding embodiments. Through a specific hardware architecture and module collaboration, it ensures long-term, reliable, automated data acquisition and processing in harsh polar environments. The device includes a passive transponder chip 410 and a sensor 420.

[0048] The passive transponder chip 410 is configured for subcutaneous implantation within the target penguin and stores a unique identifier ID. Specifically, the passive transponder chip 410 preferably uses a passive implantable tag conforming to the ISO 11784 / 11785 animal identification standard, operating at a frequency of 134.2 kHz. Since it contains no internal power source and relies entirely on the radio frequency excitation signal emitted by the sensor 420 for near-field coupling power and data transmission, it features small size, good biocompatibility, long-term maintenance-free operation, and no risk of battery failure. This makes it suitable for one-time subcutaneous implantation in the back of the neck during penguin marking operations. This passive design eliminates additional physiological burden on the penguin and avoids the risk of battery failure in extremely cold environments, providing a stable physical identification basis for long-term individual identification.

[0049] Sensor 420 is installed in the nest and includes an excitation antenna 421, a radio frequency readout module 422 connected to the excitation antenna 421, a microcontroller 423 electrically connected to the radio frequency readout module 422, and a power supply module 424, a real-time clock module 425, an ambient temperature module 426, a data storage module 427, and a data feedback module 428, all of which are electrically connected to the microcontroller 423. All modules work collaboratively under the scheduling of the microcontroller 423, forming an edge sensing and computing node.

[0050] The excitation antenna 421 is used to transmit radio frequency excitation signals to form a near-field induction zone 430 that covers only the center of the current nest. In a specific implementation, the excitation antenna 421 adopts a ring or multi-turn coil structure, arranged around the edge of the egg cup in the nest. By precisely designing the number of coil turns and geometric dimensions, the magnetic field strength it generates decays rapidly in space, thereby constructing an equivalent radius at the physical level. The near-field sensing zone 430 is approximately 0.15m to 0.35m. This spatial selectivity ensures that only penguins actually lying on top of their own nest's egg cup can be stably stimulated, while parent birds in adjacent nests or individuals merely passing through the colony passageway are unable to trigger effective readouts due to being in a weak magnetic field coupling zone, thus suppressing common neighbor interference and false triggers in dense colonies.

[0051] The radio frequency (RF) readout module 422 is responsible for driving the excitation antenna 421 and demodulating the feedback signal from the passive acknowledgment chip 410 to obtain the identifier ID and received signal strength. This module communicates with the microcontroller 423 via a serial interface, transmitting the raw RF data to the microcontroller 423 in real time for subsequent processing. The microcontroller 423, as the core of the device's computation and control, is configured to execute the monitoring method described in the aforementioned embodiment, including: transmitting an RF excitation signal through the excitation antenna 421 and synchronizing an absolute timestamp; converting the received signal strength of each triggered event into an instantaneous nest presence indication by combining it with the output of the ambient temperature module 426, obtaining the nest presence confidence level through sliding window integration, and generating a nest presence interval sequence based on dual-threshold hysteresis decision and missed read bridging; for parent individuals in the same nest determined by time mutual exclusion relationship, identifying effective hatching rotation based on the criterion that the overlap time between the two individuals in the nest interval is not less than a preset minimum overlap threshold, and determining the gap without an individual in place as a no-caretaker event (abandoned nest or empty nest event) and marking it separately; and calculating the hatching rotation interval and the time spent foraging away from the nest accordingly. The microcontroller 423 transforms abstract algorithm logic into precise timing control of hardware resources by running embedded firmware, and completes signal acquisition and behavior discrimination at the edge.

[0052] The energy supply module 424 provides continuous and stable power to the entire device, and is specifically optimized for ultra-low temperature polar environments. Specifically, the energy supply module 424 employs a hybrid power supply architecture combining a low-temperature lithium thionyl chloride primary battery and a supercapacitor in parallel. Lithium thionyl chloride batteries possess extremely high energy density and excellent low-temperature discharge performance, enabling stable operation within a temperature range of -55℃ to +85℃, making them suitable as a basic energy source; however, these batteries exhibit voltage hysteresis during high-current pulse discharge. To address this, the supercapacitor connected in parallel acts as an instantaneous power buffer, providing peak current during high-power moments such as when the microcontroller 423 initiates RF excitation or data transmission, preventing battery voltage drops that could cause system resets. This combination of a high-energy-density battery and a high-power capacitor balances the power requirements for long-term operation and high reliability under extremely cold conditions, ensuring the device can operate continuously for months during unattended incubation periods.

[0053] The real-time clock module 425 is connected to the microcontroller 423, providing a high-precision absolute time reference. Considering that the low polar temperatures may cause crystal oscillator frequency deviation, the real-time clock module 425 preferably incorporates a temperature-compensated crystal oscillator (TCXO) to ensure the accuracy of timestamp recording across the entire temperature range, providing a reliable time scale for subsequent calculations of hatching rotation intervals and foraging time away from the nest. The ambient temperature module 426 collects temperature data near the nest in real time and feeds it back to the microcontroller 423. This temperature data is not only used for environmental monitoring but also serves as a key parameter input into the low-temperature detuning compensation algorithm, correcting antenna impedance drift and reading distance deviation caused by temperature changes, ensuring the robustness of instantaneous nest indication calculations.

[0054] The data storage module 427 is used for local caching of monitoring data. In the event of a communication link interruption or data transmission failure, the microcontroller 423 writes the generated nesting interval sequence, rotation events, and statistical indicators into the data storage module 427. Once the link is restored, transmission resumes, preventing the loss of valuable ecological data. The data transmission module 428 outputs data according to a preset ecological monitoring standard data format. Specifically, the data transmission module 428 supports dual-mode redundant transmission of low-power long-range wireless communication (such as LoRa) and satellite short message service. In areas with LoRa gateway coverage, the LoRa protocol is used preferentially for extremely low-power data transmission; when in remote breeding areas without gateway coverage or when the LoRa link fails, it automatically switches to satellite short message mode. This dual-mode design ensures that monitoring data can still be reliably transmitted to the host computer or cloud server even under the complex geographical and meteorological conditions of Antarctica.

[0055] It should be understood that, although Figure 4This embodiment demonstrates a specific layout of the modules of sensor 420. However, in other embodiments, the physical arrangement of the modules can be flexibly adjusted according to the nest shape and packaging requirements. For example, the excitation antenna 421 can be separated from the circuit board to accommodate egg cups of different sizes, as long as the aforementioned electrical connections and functional configurations are maintained. The hardware architecture provided in this embodiment can support the long-term operation of the aforementioned monitoring method in extreme polar environments.

[0056] In one specific embodiment, a field monitoring application scenario for polar penguin breeding season is provided to verify the effectiveness and reliability of the monitoring methods and devices described in the aforementioned embodiments under harsh real-world conditions. This application scenario is set in an Antarctic penguin breeding colony, where nests are densely distributed with an average spacing of approximately 0.5 meters. During the monitoring period, the ambient temperature ranges from -20°C to -5°C, accompanied by persistent low temperatures and intermittent blizzards. In this scenario, the system must overcome multiple challenges, including crosstalk from neighboring nests, equipment drift due to low temperatures, and energy constraints, to achieve continuous and accurate recording of incubation rotation behavior.

[0057] During the deployment phase, researchers implanted a passive response chip storing a unique identifier (ID) subcutaneously on the back of the neck of the target penguins and installed sensors around the egg cups in each nest. The sensor's excitation antennas were arranged around the egg cups, forming an equivalent radius. The near-field sensing zone. Since the nest spacing is 0.5m, the center-to-center distance between adjacent nests is much greater than the diameter of the sensing zone. This physical spatial isolation, combined with the soft-boundary logistic function in the aforementioned embodiment, ensures that even if a parent bird in a neighboring nest is at the edge of its own nest's sensing zone, its instantaneous nest indication will not be detected. After distance decay and boundary smoothing, it is suppressed to an extremely low level, avoiding the common neighbor mis-triggering in high-density communities.

[0058] During monitoring operations, the sensors emit radio frequency excitation signals based on an adaptive duty cycle strategy. When parent birds leave the nest to forage, the system adjusts the signal based on the duration of their absence. Dynamically adjust incentive cycle For example, the excitation cycle automatically contracts to its minimum value in the early stages after parent birds have left the nest and near the expected return window. (e.g., 5s) to capture overlapping events using high-frequency sampling; while in the middle of the long ocean foraging period, the excitation period is relaxed to the maximum value. (e.g., 20 minutes), reducing ineffective power consumption. Simultaneously, for wide temperature fluctuations from -20℃ to -5℃, the microcontroller calls the ambient temperature module data in real time and compensates for low-temperature detuning. The received signal strength was corrected, eliminating ranging errors caused by antenna impedance temperature drift and ensuring the consistency of instantaneous in-nest indication calculation under different thermodynamic environments.

[0059] To verify the system's ability to ensure data continuity under extreme weather conditions, this embodiment simulates a brief power outage and restart event caused by a blizzard. When the sensor resets and restarts due to a momentary drop in power supply voltage, if the parent birds in the nest are still incubating, the system will read the parent bird's identification ID for the first time after the restart. Although there is a data gap of several minutes due to the power outage, this gap duration is less than the preset missed reading interval tolerance. The microcontroller, based on the missed read bridging logic, automatically merges the first trigger event after restarting with the last nesting interval before the power outage, treating it as the maintenance of the same continuous nesting process. This mechanism effectively prevents temporary equipment outages caused by environmental stress from being mistakenly interpreted as false behavior of parent birds leaving and returning to the nest, thus ensuring the hatching rotation interval. Completeness of statistics.

[0060] In distinguishing between incubation rotation and nest abandonment events, the overlapping handover criterion proved effective. In a normal incubation rotation, returning parent birds... Enter the sensing area and interact with the parent birds in situ. They stayed in the nest together for 150 seconds; this overlap duration... Greater than the preset minimum overlap threshold Based on this, the system determines that a valid hatching rotation has occurred and records the rotation time. Accurately identified as parent birds The timing of nest departure. In contrast, in another observation, the time from when the parent birds left the nest until the missed reading interval tolerance was exceeded. No individuals have been detected returning to the nest, resulting in a caretaker gap lasting several hours. The system did not misreport this gap as a rotation event, but instead marked it separately as a caretaker event; given that this caretaker gap far exceeds the nest abandonment tolerance threshold. Furthermore, if no individual subsequently returns to the nest to form a valid nesting interval, it is further determined to be a nest abandonment event, and a specific status code is assigned to the output data stream. This differentiated processing allows researchers to directly screen breeding failure samples from massive amounts of automated monitoring data without the need for manual backtracking of the original waveforms for secondary discrimination.

[0061] Finally, the microcontroller encapsulates the identified indicators, such as effective hatching rotation intervals, foraging time away from the nest, and nest site guarding rate, along with individual identifiers and absolute timestamps, in a data format compatible with CEMP penguin monitoring indicators. This data is then transmitted back to the host computer via LoRa or satellite short message links. This standardized output not only contains complete behavioral time-series information but also retains metadata about anomalous events such as nest abandonment, facilitating correlation analysis with polar ecological monitoring data and enabling quantitative analysis in polar ecological research.

[0062] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0063] The above-described embodiments are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. A method for monitoring penguin hatching rotation, characterized in that, The method includes: For target penguins with passive transponder chips implanted subcutaneously to store unique identifiers, sensors deployed in the nest periodically emit radio frequency excitation signals to form a near-field sensing zone that covers only the center of the current nest, and synchronize absolute timestamps; when the passive transponder chip enters the near-field sensing zone, is excited, and transmits the identifier ID back, the identifier ID, absolute trigger time, and received signal strength of each trigger event are recorded. The received signal strength of each triggered event is converted into the instantaneous in-nest indication of the individual to which the passive response chip belongs relative to the nest center, and the instantaneous in-nest indication is integrated over time within a sliding time window to obtain the in-nest residence confidence of the individual. Based on preset upper and lower thresholds, a hysteresis decision is made on the confidence of the in-nest dwell time to identify the arrival time and departure time of the individual. Adjacent triggering events with an interval less than the preset missed read gap tolerance within the in-nest interval are bridged to filter out false triggering events caused by passing through the nest and neighbor interference, and to compensate for the occasional missed reads of the passive response chip, thereby generating the in-nest interval sequence of the individual. Within a preset learning time window, based on the cumulative effective nesting time of each identifier ID and the overlap ratio of the nesting intervals of any two identifier IDs, the two identifier IDs with a mutual exclusion degree lower than a preset mutual exclusion threshold and the highest cumulative effective nesting time are identified as the parent individuals of the nest. The one currently maintaining nesting is called the in-situ individual, and the one entering the near-field sensing area is called the returning individual. When the overlap time of the nesting intervals of the returning individual and the in-situ individual is not less than a preset minimum handover overlap threshold, a valid hatching rotation is determined to have occurred. When the in-situ individual leaves the nest and there is no returning individual present, resulting in a gap in the nest where no individual is present, it is determined to be an unattended event and is marked separately, not included in the valid hatching rotation. The unattended event is classified as an empty nest event or an abandoned nest event based on its duration and whether it subsequently recovers its effective nesting status. Based on the identified valid hatching rotation events, the hatching rotation interval and the time spent foraging away from the nest are calculated and output according to the preset ecological monitoring standard data format.

2. The method according to claim 1, characterized in that, The sensor periodically transmits radio frequency excitation signals, including: transmitting radio frequency excitation signals with an adaptive duty cycle, specifically: The duration of time experienced by the individual after this departure from the nest. The independent variable is based on the individual's expected foraging time. and its dispersion Adaptively adjust the transmission period of the radio frequency excitation signal : ， in, and These represent the preset maximum and minimum excitation periods, respectively. This represents the preset tracking time constant, and... Limited to The range; the expected foraging time. Updated using an exponentially weighted moving average: ， in, This indicates the preset update weight. This indicates the duration of the individual's most recent foraging trip away from the nest; so that the sensor is excited at a high frequency within a preset short period after the individual leaves the nest and near its expected return time window, and at a low frequency during the middle of the foraging trip in the open ocean.

3. The method according to claim 1, characterized in that, The step of converting the received signal strength of each triggered event into an instantaneous presence indication of the individual to which the passive acknowledgment chip belongs relative to the nest center includes: Based on the field-calibrated signal strength-distance empirical model, and compensated for the influence of ambient temperature on antenna readout characteristics, the estimated distance of the individual relative to the excitation antenna in the triggering event is calculated. : ， in, and These represent the reference distance and its corresponding reference received signal strength, respectively. Indicates the current received signal strength. Indicates the empirical decay index, The amount of low-temperature detuning compensation is calculated using the following formula: ， in, Indicates the temperature compensation coefficient. Indicates the calibration temperature. Indicates the current ambient temperature; based on the estimated distance and the preset near-field sensing zone radius. The instantaneous in-nest indication is calculated using a soft-boundary logistic function. : ， in, This represents the preset boundary softening coefficient.

4. The method according to claim 3, characterized in that, The step of integrating the instantaneous in-nest indication within the sliding time window to obtain the in-nest residence confidence of the individual is specifically as follows: In length of Within the sliding time window, the instantaneous nest indication Perform time integration and normalization to obtain the individual. At any moment Nest dwell confidence : ， in, This indicates the length of the sliding time window. Indicates the current moment. Represents the time variable of integration. Represents the individual At any moment The instantaneous indication in the nest; This ensures that the residence confidence of individuals continuously lying down and incubating converges to one, while the residence confidence of individuals passing through the nest is suppressed to a low level.

5. The method according to claim 4, characterized in that, The method of performing a delayed decision on the confidence level of nest dwelling based on preset upper and lower thresholds to identify the arrival and departure times of the individual, and bridging adjacent triggered events within the nest dwelling interval that are less than a preset missed read gap tolerance, includes: Set upper threshold With lower threshold ,and Less than When the individual Nest dwell confidence The upper threshold is first crossed from below. When the individual arrives at the nest, the starting point of the current nesting interval is back-marked to the moment when the first instantaneous nesting indication is valid within that interval; When the confidence level of nest dwell is mentioned Crossing the lower threshold And within the preset tolerance for missed reading intervals If no valid trigger is obtained within the period, it is determined that the individual has left the nest, and the end point of the nesting interval is marked to the last moment when the nesting indication was valid within the interval; For intervals within the nested interval that are smaller than the missed read gap tolerance The adjacent triggering events are bridged to maintain the continuous nested state.

6. The method according to claim 5, characterized in that, The step of identifying the two identifiers with the highest mutual exclusion degree (below a preset mutual exclusion threshold) and the highest effective nesting time as the parent individuals of the nest includes: Within the preset learning time window, the peak retention confidence level should not be lower than the upper threshold. Furthermore, the peak value of the received signal strength is not lower than the preset threshold value for the main peak strength of the nested signal. The nesting interval is denoted as the valid nesting interval; The two identifiers with the highest cumulative effective nesting time and a mutual exclusivity relationship are identified as the parent individuals of the nest; the mutual exclusivity relationship is determined by the degree of mutual exclusivity. Characterization: ， in, Indicates that the identifier ID is and The total effective overlap time of the two individuals in the nest interval. and These represent the effective cumulative time spent in the nest for both, and the mutual exclusion degree is... Less than the preset mutual exclusion threshold; The peak retention confidence level is lower than the upper threshold. Or the peak value of the received signal strength is lower than the nested main peak strength threshold. The identifier ID is used to identify individuals that have passed through the nest or are from neighboring nests, and these individuals are removed.

7. The method according to claim 6, characterized in that, When the overlap time between the returning individual and the in-situ individual in the nesting interval is not less than a preset minimum handover overlap threshold, a valid hatching rotation is determined to have occurred, including: When the parent birds are in power The nesting period has not yet ended, and the returning parent birds... The starting point of the nesting interval precedes the in situ parent bird. When the two appear at the end of the nesting interval, they overlap, and the overlap duration is... for: ， in, Indicates the in-situ parent bird The end point of the nesting interval, This indicates the returning parent birds The starting point of the nesting interval; When the overlap duration Not less than the preset minimum overlap threshold And from the aforementioned parent birds in situ The returning parent birds after leaving the nest When an individual remains the sole in-situ individual, a valid hatching rotation is determined, and the rotation time is recorded. The time of departure of the parent birds in situ is specified as follows: .

8. The method according to claim 7, characterized in that, The calculation of the incubation rotation interval is as follows: Parent birds The The duration of continuous incubation within the nest is used as the interval for a single incubation cycle, i.e., the incubation time within the nest. : ， in, and Each represents the parent bird No. The start and end points of the nesting interval; the interval between two adjacent incubation cycles is equal to the duration of the corresponding nesting interval of the in-situ parent bird.

9. The method according to claim 8, characterized in that, The calculation of the hatching rotation interval and the time spent foraging away from the nest, and the output of the data according to the preset ecological monitoring standard data format, includes: Parent birds The time spent away from the nest between two adjacent nesting intervals is taken as the time spent away from the nest foraging. : ， in, Indicates the first The starting point of the segment in the nest interval, Indicates the first The end point of the segment within the nest interval; Statistical period The percentage of time that at least one parent bird remains in the nest is used as the nest site protection rate. : ， The hatching rotation intervals, foraging time away from the nest, and nest site protection rate, along with the corresponding individual identifiers (IDs) and absolute timestamps, are output according to the preset ecological monitoring standard data format. The cumulative time of both parents in the nest and the total overlapping time of parental interactions are both counted as the time falling within the statistical period. Nesting intervals that cross the boundary of the statistical period are counted as the intersection of the interval with the statistical period.

10. An apparatus for monitoring penguin hatching rotation, used to perform the method for monitoring penguin hatching rotation as described in any one of claims 1 to 9, characterized in that, include: The passive response chip is configured to be implanted under the skin of the target penguin and stores a unique identifier ID inside. The sensor, installed in the nest, includes an excitation antenna, a radio frequency readout module connected to the excitation antenna, a microcontroller electrically connected to the radio frequency readout module, and an energy supply module, a real-time clock module, an ambient temperature module, a data storage module, and a data feedback module, all electrically connected to the microcontroller. The microcontroller is configured as follows: The excitation antenna transmits a radio frequency excitation signal to form a near-field sensing zone that covers only the center of the current nest and synchronizes the absolute timestamp. The received signal strength of each triggered event is combined with the output of the ambient temperature module to convert it into an instantaneous in-nest indication. The in-nest residence confidence is obtained by sliding window integration. Based on the dual threshold hysteresis decision and missed read bridging, an in-nest interval sequence is generated to filter out in-nest passage and adjacent false trigger events. Within a preset learning time window, based on the cumulative effective nesting time of each identifier ID and the overlap ratio of the nesting intervals of any two identifier IDs, the two identifier IDs with a mutual exclusion degree lower than a preset mutual exclusion threshold and the highest cumulative effective nesting time are identified as the parent individuals of the nest. The one currently maintaining nesting is called the in-situ individual, and the one entering the near-field sensing area is called the returning individual. When the overlap time of the nesting intervals of the returning individual and the in-situ individual is not less than a preset minimum handover overlap threshold, a valid hatching rotation is determined to have occurred. When the in-situ individual leaves the nest and there is no returning individual present, resulting in a gap in the nest where no individual is present, it is determined to be an unattended event and is marked separately, not included in the valid hatching rotation. The unattended event is classified as an empty nest event or an abandoned nest event based on its duration and whether it subsequently recovers its effective nesting status. Based on this, the hatching rotation interval and the time spent foraging outside the nest are calculated, and the data is output by the data feedback module according to the preset ecological monitoring standard data format.

Images