Bird repelling mode switching method, system, device and medium of a bird repelling device

Abstract

The application discloses a bird repelling mode switching method, system, device and medium of a bird repelling device. The method comprises the following steps: collecting bird activity monitoring data in the action area of the bird repelling device, determining a comprehensive activity intensity sequence based on the bird activity monitoring data; determining a pre-action window, a post-action window and a control window according to the comprehensive activity intensity sequence; calculating a net instant bird repelling effect value according to the comprehensive activity intensity corresponding to the three windows; calculating a continuous habituation index based on the net instant bird repelling effect value; determining whether the current bird repelling mode is invalid according to the continuous habituation index and a second comprehensive activity intensity; if it is determined that the current bird repelling mode is invalid, screening a candidate bird repelling mode set meeting a preset condition from a pre-established bird repelling mode pool; evaluating the control parameters of each bird repelling mode in the candidate bird repelling mode set to determine a target bird repelling mode based on the evaluation result. Therefore, the application can accurately switch the bird repelling mode according to the real reaction of the bird.

Description

Technical Field

[0001] This invention relates to the field of bird repellency mode switching, and more particularly to a bird repellency mode switching method, system, device, and medium for a bird repellency device. Background Technology

[0002] Birds frequently perch, nest, or congregate near transmission line towers, insulator strings, crossarms, and conductors, easily causing line faults such as bird droppings flashover and insulator breakdown, posing a serious threat to the safe operation of the power grid. Currently, bird-repelling devices such as laser beams, flashing lights, ultrasonic waves, voice warnings, and pan-tilt-zoom (PTZ) scanning have been widely deployed along transmission lines. These devices typically produce a significant deterrent effect in the initial deployment phase. However, as the same bird-repelling stimuli (such as fixed-frequency laser scanning, fixed voice content, or repetitive PTZ scanning paths) are repeatedly applied to the same target birds in the same area over a long period, the birds gradually realize that these stimuli do not pose a substantial threat, thus developing habituation. Once habituation occurs, even if the devices continue to perform bird-repelling actions, the frequency of bird appearance, the duration of their stay, or their proximity to towers and insulators no longer decreases significantly, and may even completely return to pre-determination levels. At this point, the devices are in a state of latent failure—"action performed but effect lost"—not only causing power consumption and equipment wear but also exposing the transmission lines to bird damage risks again. Therefore, bird deterrence strategies that rely solely on fixed patterns and operate over a long period of time are insufficient to address the adaptive behaviors of birds. It is necessary to switch bird deterrence patterns to maintain a long-term and effective deterrence effect.

[0003] Existing bird-repelling devices primarily employ two strategies to adjust bird-repelling modes: First, they operate continuously in a single fixed mode, changing operating parameters only when the device is restarted at set intervals or during manual intervention, without any spontaneous switching of bird-repelling modes. Second, they use a timed or random rotation strategy, switching between multiple bird-repelling modes at preset time intervals (e.g., daily, weekly) or randomly selecting the next mode from a mode pool. However, the fixed-mode approach cannot detect whether it has become ineffective; even if birds have fully adapted, the device continues to output ineffective stimuli. In summary, existing technologies lack accurate judgment of the birds' actual habitual state when switching bird-repelling modes and selecting the next mode, making it difficult to achieve accurate decision-making to "switch only when truly ineffective and switch to a truly effective mode." Summary of the Invention

[0004] This invention provides a method, system, device, and medium for switching bird-repelling modes in a bird-repelling device, which can accurately switch bird-repelling modes based on the actual reactions of birds.

[0005] The first aspect of the present invention provides a method for switching bird-repelling modes of a bird-repelling device, comprising: Collect bird activity monitoring data within the area of ​​effect of the bird deterrent device, and determine a comprehensive activity intensity sequence based on the bird activity monitoring data; Using action events as identifiers, the pre-action window, post-action window, and control window are determined based on the comprehensive activity intensity sequence. The net instant bird deterrence effect value is calculated based on the first comprehensive activity intensity corresponding to the pre-action window, the second comprehensive activity intensity corresponding to the post-action window, and the third comprehensive activity intensity corresponding to the control window. Based on the net immediate bird deterrence effect value and the baseline bird deterrence effect value of the current bird deterrence mode of the bird deterrence device, a continuous habituation index is obtained. The current bird deterrence mode is determined to be ineffective based on the continuous habituation index and the second comprehensive activity intensity. If the bird deterrence mode fails, a set of candidate bird deterrence modes that meet the preset conditions will be selected from the pre-established bird deterrence mode pool. The bird control parameters for each bird control mode in the candidate bird control mode set are evaluated to determine the target bird control mode based on the evaluation results.

[0006] This invention collects bird activity monitoring data and normalizes it into a comprehensive activity intensity sequence, transforming multi-source heterogeneous observation information into a unified quantitative measurement benchmark, providing a comparable and calculable data foundation for subsequent effect evaluation. Based on this, a time window mechanism using action events as identifiers is established, defining a pre-action window, a post-action window, and a control window. The control window is used to deduct activity intensity changes caused by non-bird-repelling factors such as weather, time of day, and natural bird migration, ensuring that the calculated net immediate bird-repelling effect accurately reflects the true deterrence contribution of the bird-repelling action itself, avoiding the bias of misjudging external factor fluctuations as bird-repelling effectiveness. Furthermore, by comparing this net immediate bird-repelling effect value with the current model's baseline bird-repelling effect value and generating a continuous habituation index, the trend of gradual decay of bird-repelling effect with repeated use can be captured from a time dimension. Combined with the second comprehensive activity intensity of the post-action window for dual-threshold joint judgment, it confirms the continuous decay of the effect while further verifying that birds maintain a high level of activity after the bird-repelling action. Failure judgment is triggered only when both conditions are met, effectively excluding the possibility that birds have already left naturally. The system mitigates the misleading influence of interference such as abnormal single observations on switching decisions, ensuring a precise correspondence between the switching trigger timing and the birds' actual habituation state. After determining failure, a candidate mode set is further screened using preset conditions, and the executable control parameters of the candidate modes are evaluated to determine the target mode. This ensures that the selected target mode not only has sufficient stimulus difference from the current failed mode to break the birds' adaptive expectations, but also meets safety boundaries and the device's executable constraints, avoiding the selection of unsafe or unsupported modes in pursuit of stimulus difference. Ultimately, each step is interconnected and progressive, jointly driving the bird deterrent device to switch only when the current mode has truly become habituated and failed. During the switch, an executable target mode is selected from the available differentiated modes, ensuring that the timing and direction of mode switching are driven by the birds' actual behavioral responses, rather than relying on fixed periods, random numbers, or scene rule matching. This achieves the technical effect of "precisely switching bird deterrent modes based on the birds' actual responses," effectively extending the long-term effective operating time of the bird deterrent device, reducing ineffective idling and equipment wear, and lowering manual inspection and maintenance costs.

[0007] Furthermore, determining the comprehensive activity intensity sequence based on the bird activity monitoring data includes: Determine the identification confidence level of each data point in the bird activity monitoring data; Data with an identification confidence level lower than a preset confidence threshold are filtered out from the bird activity monitoring data to obtain valid activity monitoring data; Normalization is performed on each available activity feature in the effective activity monitoring data to obtain the normalization result; Based on the normalization results described above, a comprehensive activity intensity sequence updated over time is obtained.

[0008] By filtering valid data using confidence levels and normalizing and weighting them, interference from low-quality observations is eliminated, providing accurate input for subsequent effect assessment and supporting precise switching based on the birds' actual responses.

[0009] Further, the calculation of the net immediate bird deterrence effect value based on the first comprehensive activity intensity corresponding to the pre-action window, the second comprehensive activity intensity corresponding to the post-action window, and the third comprehensive activity intensity corresponding to the control window includes: The ratio of the actual number of valid activity monitoring data to the theoretical sampling number is calculated separately in the pre-action window, the post-action window, and the control window, and is used as the observation completeness of each window; The arithmetic mean of the identification confidence scores corresponding to the valid activity monitoring data in the pre-action window, post-action window, and control window is calculated respectively, and used as the average identification confidence score for each window; If the observation completeness of all windows is not lower than the preset completeness threshold and the average identification confidence of all windows is not lower than the preset average confidence threshold, then the front-end comprehensive activity intensity and the back-end comprehensive activity intensity of the control window are determined according to the third comprehensive activity intensity. Calculate the first difference between the first comprehensive activity intensity and the second comprehensive activity intensity, and the second difference between the front-end comprehensive activity intensity and the back-end comprehensive activity intensity; The net decrease obtained by subtracting the second difference from the first difference is used as the numerator, and the sum of the first comprehensive activity intensity and the preset zero constant is used as the denominator to calculate the original bird deterrence effect value. The original bird deterrence effect value is limited to a preset range to obtain the net real-time bird deterrence effect value.

[0010] This triple gating system, which uses observation completeness, confidence level, and matching difference to determine measurability, blocks invalid data from participating in the calculation and ensures that failure determination is based solely on reliable observations.

[0011] Furthermore, the continuous habituation index is obtained based on the net immediate bird deterrence effect value and the baseline bird deterrence effect value of the current bird deterrence mode of the bird deterrence device, including: Determine the baseline bird-repelling effect value of the current bird-repelling mode of the bird-repelling device, wherein the baseline bird-repelling effect value is obtained by updating the net instantaneous bird-repelling effect value corresponding to each measurable action event of the current bird-repelling mode during the period when habituation failure is not determined through exponential moving average. The net instant bird deterrence effect value is compared with the baseline bird deterrence effect value to obtain the degree of single habituation decay corresponding to the action event. The continuous habituation index is obtained by weighting the degree of single habituation decay corresponding to each of the most recent consecutive measurable action events under the current bird deterrence mode with time decay.

[0012] By using multi-dimensional matching such as propensity score matching to obtain a control window, and deducting the influence of natural factors such as weather and time of day on the intensity of the activity, the net effect value truly reflects the bird-repelling action itself.

[0013] Furthermore, the step of selecting a set of candidate bird-repelling patterns that meet preset conditions from a pre-established bird-repelling pattern pool includes: Read the mode identifier, failure status flag, required execution unit list, cooldown timer parameters and execution control parameters corresponding to each bird-repelling mode in the bird-repelling mode pool; Bird-repelling modes whose mode identifier is the same as the current mode's mode identifier, and bird-repelling modes whose failure status is marked as failed, are excluded from the bird-repelling mode pool to obtain the first screening result. For each bird-repelling mode in the first screening result, the current working status of the device corresponding to the required execution unit list is read, the cooling timer parameter is read to determine the current remaining cooling time, and the current environmental parameters are obtained to verify the safety execution conditions corresponding to the execution control parameters item by item. If there is an offline or faulty unit in the current device working status, the remaining cooling time is greater than zero, or any of the safety execution conditions are not met, then the bird-repelling mode is excluded. The remaining bird-repelling patterns after elimination are combined into the candidate bird-repelling pattern set.

[0014] This standardized formula quantitatively calculates the net immediate bird deterrence effect after deducting natural variations, providing a comparable and cumulative quantitative basis for single-event attenuation indicators.

[0015] Further, the step of evaluating the bird control parameters for each bird control mode in the candidate bird control mode set to determine the target bird control mode based on the evaluation results includes: Determine the key stimulus type differences and control method differences between each candidate bird-repelling pattern in the candidate bird-repelling pattern set and the current bird-repelling pattern; Based on the differences in the key stimulus types and the differences in the control methods, the candidate bird deterrence pattern set is prioritized and graded to obtain the grading results. Based on the grading results and the preset mode priority of each candidate bird deterrence mode in the corresponding target bird species and the current scene, a target bird deterrence mode is determined from the set of candidate bird deterrence modes.

[0016] By updating the benchmark through exponential moving average and combining it with dual threshold determination, we can accurately identify true failures where birds are still lingering despite continuous decline, thus avoiding false switching caused by a single fluctuation.

[0017] Furthermore, the present invention also includes: After executing the target bird deterrence mode, the current action event is used as an identifier. Based on the newly acquired updated comprehensive activity intensity sequence, the window before the updated action, the window after the updated action, and the updated control window are re-determined, and the updated bird deterrence effect value is recalculated. When there is no set of candidate bird deterrence modes that meet the preset conditions, or when multiple different bird deterrence modes in the bird deterrence mode pool have been determined to have become habitually ineffective, the system enters a safety maintenance state, stops switching bird deterrence modes, and generates maintenance prompt information. When a preset number of consecutive action events are marked as events with unmeasurable effects, a sensor verification prompt message is generated.

[0018] This lays the resource foundation for differentiated choices and ensures that the switched model has the ability to effectively stimulate change.

[0019] Another embodiment of the present invention provides a bird-repelling mode switching system for a bird-repelling device, comprising: the system being used to perform the steps of the bird-repelling mode switching method of the bird-repelling device as described in the present invention.

[0020] Another embodiment of the present invention provides a terminal device, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the steps of the bird-repelling mode switching method of the bird-repelling device as described in the present invention.

[0021] Another embodiment of the present invention provides a computer-readable storage medium item, including: a stored computer program, which, when the computer program is running, controls the device where the computer-readable storage medium is located to perform the bird-repelling mode switching method of the bird-repelling device as described in the present invention. Attached Figure Description

[0022] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0023] Figure 1 This is a flowchart illustrating an embodiment of the bird-repelling mode switching method of the bird-repelling device provided in this application; Figure 2This is a flowchart illustrating one embodiment of steps S201 to S206 provided in this application; Figure 3 This is a flowchart illustrating one embodiment of steps S301 to S303 provided in this application. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0025] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0026] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0027] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0028] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0029] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0030] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0031] See Figure 1 To accurately switch bird-repelling modes based on the actual reactions of birds, an embodiment of the present invention provides a method for switching bird-repelling modes of a bird-repelling device, including steps S101 to S107: Step S101: Collect bird activity monitoring data within the area of ​​effect of the bird deterrent device, and determine a comprehensive activity intensity sequence based on the bird activity monitoring data; In some embodiments, bird activity monitoring data is collected within the area of ​​effect of the bird deterrent device. Specifically, the monitoring terminal collects at least one of image, video, audio, infrared or millimeter-wave radar data within the area covered by the bird deterrent device according to a unified time reference, and forms bird activity monitoring data arranged by time. Each record in the bird activity monitoring data includes at least the recording time, bird species or target identifier, and at least one of the following: frequency of bird appearance, number of targets, dwell time, closest distance to the tower, insulator or conductor at risk, and activity density. Each record is also accompanied by an identification confidence level.

[0032] In some embodiments, determining the comprehensive activity intensity sequence based on the bird activity monitoring data includes: determining the identification confidence level of each data point in the bird activity monitoring data; filtering out data points with identification confidence levels lower than a preset confidence threshold from the bird activity monitoring data to obtain valid activity monitoring data; performing normalization on each available activity feature in the valid activity monitoring data to obtain a normalization result; and obtaining a comprehensive activity intensity sequence updated over time based on each of the normalization results.

[0033] In some embodiments, after collecting various types of raw monitoring data, the system outputs an identification confidence score for each bird activity record. This confidence score is derived from the internal score of the target detection model or identification algorithm on the reliability of the current observation result. Simultaneously, the system sets a preset confidence threshold for each record. Then, the identification confidence score of each record is compared with the preset confidence threshold. Records with confidence scores below the threshold are considered unreliable observations and are directly filtered out from subsequent processing. Only records that reach or exceed the threshold are retained as valid activity monitoring data and enter the normalization stage. For valid records that pass the gating, the system reads all available activity features that exist simultaneously and performs normalization according to the physical meaning of each feature, obtaining a normalization result. When only one available feature exists in a valid record at a certain moment, the system directly uses the normalization result of that feature as the comprehensive activity intensity at that moment. When multiple available features exist at the same moment, the system performs a weighted sum of the normalization results according to a pre-fixed set of weight coefficients, with the sum of the feature weights being 1, and uses the weighted composite result as the comprehensive activity intensity at that moment. After processing each valid record, the overall activity intensity at all times is arranged in chronological order, thus forming a sequence of overall activity intensity that is updated over time.

[0034] It should be noted that the preset confidence threshold can be pre-written into the configuration parameters, or multiple sets of thresholds can be maintained according to different bird species, time periods or sensor types.

[0035] It should be noted that each available activity characteristic includes at least one of the following: frequency of bird appearance, number of targets, length of stay, closest distance to the risk location, and activity density.

[0036] It should be noted that normalization is performed according to the physical meaning of the features. Specifically, the frequency of occurrence, number of targets, dwell time and activity density are normalized by the maximum and minimum values ​​or logarithmic compression to convert them into unidirectional feature values ​​that increase with the increase of the value in the range of 0 to 1. The nearest distance is first converted into proximity before participating in the normalization, so that the closer the distance, the higher the value, which also falls in the range of 0 to 1.

[0037] By filtering valid data using confidence levels and normalizing and weighting them, interference from low-quality observations is eliminated, providing accurate input for subsequent effect assessment and supporting precise switching based on the birds' actual responses.

[0038] Step S102: Using action events as identifiers, determine the pre-action window, post-action window, and control window based on the comprehensive activity intensity sequence; In some embodiments, an action event is established using each bird-repelling action as the basic identifier unit. This action event records the mode identifier, start and end times of the bird-repelling action, and the corresponding control parameter distribution records. For this action event, the system extracts a pre-action window and a post-action window from the comprehensive activity intensity sequence. Each window contains comprehensive activity intensity data points from multiple sampling times, and the number of valid records in each window must reach a preset proportion of the number of records to be obtained according to the sampling period to ensure statistical representativeness. Additionally, the system searches for non-action periods without overlapping influence from other bird-repelling actions in the bird-repelling device's own historical records as historical control windows. After the three windows are determined, the arithmetic mean of the comprehensive activity intensities corresponding to each valid record in the pre-action window is recorded as the first comprehensive activity intensity, the arithmetic mean of the post-action window is recorded as the second comprehensive activity intensity, and the arithmetic mean of the first and second segments of the control window, obtained using the same method, are used as the first and second components of the third comprehensive activity intensity.

[0039] It should be noted that the pre-action window is a continuous time period before the bird deterrence action begins, and the post-action window is a continuous time period after the bird deterrence device's response delay ends.

[0040] It should be noted that bird activity records can also be collected from spatial areas outside the current bird deterrence device's effective range and where no bird deterrence actions are performed simultaneously. After generating a control sequence according to the same comprehensive activity intensity calculation rules as the action area, the preceding and following segments corresponding to the start and end times of the action window are extracted to determine the historical spatial control window.

[0041] It should be noted that, regardless of whether spatial or historical comparison is used, the system must match the current action event with four dimensions: target bird species, time period, weather, and initial activity intensity. Specifically, it uses propensity score matching, nearest neighbor matching, or hierarchical matching to ensure that the difference in each matching dimension does not exceed the corresponding preset threshold. If there are multiple candidate comparison windows, the one with the smallest weighted comprehensive difference in each dimension is selected as the final matching result.

[0042] Step S103: Calculate the net instant bird deterrence effect value based on the first comprehensive activity intensity corresponding to the pre-action window, the second comprehensive activity intensity corresponding to the post-action window, and the third comprehensive activity intensity corresponding to the control window. Please refer to Figure 2 In some embodiments, step S103 includes steps S201 to S206: Step S201: Calculate the ratio of the actual number of valid activity monitoring data to the theoretical sampling number in the pre-action window, post-action window, and control window respectively, and use this as the observation completeness of each window; In some embodiments, for the pre-action window, post-action window, and control window that have been determined, the system counts the number of valid activity monitoring data records actually retained in each window, and calculates the total number of records that should theoretically be obtained based on the time span of each window and the preset sampling period; the ratio of the actual number of records to the theoretical total number of records is the observation completeness of that window.

[0043] It should be noted that observation completeness is used to quantify whether the effective data within the window is sufficient to support subsequent statistical calculations. Different windows, including the pre-action, post-action, and the first and last parts of the control window, need to be statistically analyzed independently.

[0044] Step S202: Calculate the arithmetic mean of the identification confidence scores corresponding to the valid activity monitoring data in the pre-action window, post-action window and control window respectively, and use it as the average identification confidence score of each window; In some embodiments, the system reads the recognition confidence score value that is output along with each valid activity monitoring data in the pre-action window, post-action window, and control window. This confidence score is the internal score of the original target detection or recognition model on the reliability of the current observation. The system calculates the arithmetic mean of all confidence scores in each window, that is, the sum of the confidence scores of each valid record in the window is divided by the number of valid records in the window, and the result is used as the average recognition confidence score of the window.

[0045] It should be noted that the three windows are calculated separately to obtain the average confidence level of the window before the action, the average confidence level of the window after the action, and the average confidence level of the control window.

[0046] Step S203: If the observation completeness of all windows is not lower than the preset completeness threshold and the average identification confidence of all windows is not lower than the preset average confidence threshold, then the front-end comprehensive activity intensity and the back-end comprehensive activity intensity of the control window are determined according to the third comprehensive activity intensity. In some embodiments, the observation completeness of each window is compared with a preset completeness threshold, and the average recognition confidence of each window is compared with a preset average confidence threshold. Only when the observation completeness of the pre-action window, post-action window, and control window is all not lower than the preset completeness threshold, and the average recognition confidence of the three is all not lower than the preset average confidence threshold, is the action event marked as a measurable event, and the system allows the subsequent calculation of net real-time bird deterrence effect to continue. After the above measurable conditions are met, the system divides the control window into a front segment and a back segment according to time sequence, extracts the comprehensive activity intensity of each valid record in the corresponding time segment, and calculates the arithmetic mean, which is used as the comprehensive activity intensity of the front segment and the comprehensive activity intensity of the back segment of the control window.

[0047] It should be noted that the comprehensive activity intensity of the pre-window segment and the comprehensive activity intensity of the control window segment are collectively referred to as the two components of the third comprehensive activity intensity, which are used for comparison with the pre- and post-action responses, respectively.

[0048] Step S204: Calculate the first difference between the first comprehensive activity intensity and the second comprehensive activity intensity, and the second difference between the front-end comprehensive activity intensity and the back-end comprehensive activity intensity; In some embodiments, the system records the arithmetic mean of the overall activity intensity corresponding to each valid record in the window before the action as the first overall activity intensity, and the arithmetic mean of the window after the action as the second overall activity intensity. Then, the system subtracts the second overall activity intensity from the first overall activity intensity to obtain a first difference reflecting the change in activity intensity within the action area before and after the bird-repelling action. At the same time, the system subtracts the overall activity intensity of the control window from the overall activity intensity of the control window at the beginning of the control window to obtain a second difference reflecting the change in natural activity intensity under the condition of no bird-repelling action during the same period.

[0049] Step S205: The net decrease obtained by subtracting the second difference from the first difference is used as the numerator, and the sum of the first comprehensive activity intensity and the preset zero constant is used as the denominator to calculate the original bird deterrence effect value. In some embodiments, the first difference is subtracted from the second difference, and the resulting net decrease is taken as the net decrease in activity intensity attributable to bird-repelling actions after deducting changes in natural activity; the net decrease in activity intensity is taken as the numerator, and the sum of the first comprehensive activity intensity and the preset zero constant is taken as the denominator, and the ratio of the two is calculated, and the result is the original bird-repelling effect value.

[0050] It should be noted that the zero-prevention constant is a small positive number greater than 0, and it only plays a stabilizing role when the average activity intensity before the action approaches zero.

[0051] It should be noted that the physical meaning of this calculation formula is: the ratio of the net decrease in bird descent to the initial activity intensity before the action, reflecting the additional bird-repelling effect brought about by the bird-repelling action relative to natural changes.

[0052] Step S206: Limit the original bird deterrence effect value within a preset range to obtain the net real-time bird deterrence effect value.

[0053] In some embodiments, the original bird deterrence effect value is truncated at both upper and lower limits: if the ratio is less than 0, it is set to 0; if it is greater than 1, it is set to 1; if it is between 0 and 1, the original value is retained. The value output after this truncation is the final net real-time bird deterrence effect value, which is restricted to a closed interval between 0 and 1, so as to enable unified comparison and subsequent benchmark updates and habituation decay calculations between different bird deterrence modes and different action events.

[0054] In some embodiments, the calculation formula is as follows: ; In the formula, This represents the net immediate bird deterrence effect value. The intensity of the first comprehensive activity, The second comprehensive activity intensity, and These represent the comprehensive activity intensity at the beginning and end of the control window, respectively. This is a preset zero-prevention constant.

[0055] This triple gating system, which uses observation completeness, confidence level, and matching difference to determine measurability, blocks invalid data from participating in the calculation and ensures that failure determination is based solely on reliable observations.

[0056] Step S104: Based on the net immediate bird deterrence effect value and the baseline bird deterrence effect value of the current bird deterrence mode of the bird deterrence device, a continuous habituation index is obtained. Please refer to Figure 3 In some embodiments, step S104 includes steps S301 to S303: Step S301: Determine the baseline bird-repelling effect value of the current bird-repelling mode of the bird-repelling device, wherein the baseline bird-repelling effect value is obtained by updating the net instantaneous bird-repelling effect value corresponding to each measurable action event of the current bird-repelling mode during the period when habituation failure is not determined through exponential moving average. In some embodiments, a baseline bird-repelling effect value is maintained independently for each bird-repelling mode. This baseline value is updated only while the current mode is in an effective state. Specifically, when a mode is first put into use, the system uses the net instantaneous bird-repelling effect values ​​calculated from multiple consecutive measurable action events in the initial stage of the mode as the basic sequence, and establishes an initial baseline using an exponential moving average method. In subsequent operations, only the net instantaneous bird-repelling effect values ​​that are marked as measurable and for which the mode has not yet been determined to be in a period of habitual failure are sequentially updated using the exponential moving average method. Each time the net instantaneous bird-repelling effect value is updated, the newly entered net instantaneous bird-repelling effect value is weighted and fused with the current baseline value according to a preset smoothing coefficient to obtain the updated baseline bird-repelling effect value.

[0057] It should be noted that if an action event is marked as having an unmeasurable effect, or if the pattern has been determined to be habitually ineffective at the corresponding time, the net real-time bird deterrence effect value at the corresponding time will not participate in the update of the pattern's baseline value. If a pattern lacks measurable events for a long period of time, its baseline value will remain at the value of the most recent effective update until a new measurable event arrives, at which point the moving average update will resume.

[0058] Step S302: Compare the net instant bird deterrence effect value with the baseline bird deterrence effect value to obtain the degree of single habituation decay corresponding to the action event; In some embodiments, it is determined whether the net immediate bird deterrence effect value reaches or exceeds the benchmark bird deterrence effect value. If so, the single-action habituation attenuation degree of the current action is directly set to the minimum value. If the net immediate bird deterrence effect value is lower than the benchmark bird deterrence effect value, the ratio of the difference between the two to the historical benchmark value is calculated, and this ratio is quantified as the single-action habituation attenuation degree of the current action. The greater the decrease in the net immediate bird deterrence effect value relative to the benchmark value, the higher the single-action habituation attenuation degree is. This single-action index only represents the attenuation status of the current bird deterrence action relative to the previous effective level of the pattern, and does not independently trigger any pattern failure determination. It is only used as the basic input for the subsequent calculation of the continuous habituation index.

[0059] Step S303: Perform time-decay weighted summation on the single habituation decay degree corresponding to each of the most recent consecutive measurable action events under the current bird deterrence mode to obtain the continuous habituation index.

[0060] In some embodiments, a sliding record window of the most recent consecutive measurable action events is maintained for the current bird deterrence mode. Whenever a new measurable action event is completed and its single habituation decay degree is calculated, the system pushes the value into the record window and removes the oldest data in the window. The window size R is a fixed positive integer. Subsequently, the system performs time decay weighting on the R single habituation decay degrees stored in the current window: the decay degree closer to the current time is assigned a higher weight, and the decay degree farther away is assigned a lower weight. Specifically, a preset time decay coefficient λ (0 < λ ≤ 1) is used to decrease step by step in the form of a power function. Each decay degree is multiplied by its corresponding weight, summed, and then divided by the sum of all weights to normalize. The result is the continuous habituation index at the current time.

[0061] In some embodiments, the formula for calculating the habituation metric is: ,in, This is a continuous habituation indicator; R is the number of most recently measurable action events included in the calculation. Let λ be the single decay index of the r-th measurable action event from the current time backward, where λ is the time decay coefficient and 0 < λ ≤ 1; the smaller λ is, the more emphasis is placed on the most recent action.

[0062] It should be noted that a higher continuous habituation index indicates a more severe decline in the effectiveness of the bird-repelling mode with repeated use. Ultimately, it is used to compare with a preset habituation threshold to trigger a failure determination.

[0063] By using multi-dimensional matching such as propensity score matching to obtain a control window, and deducting the influence of natural factors such as weather and time of day on the intensity of the activity, the net effect value truly reflects the bird-repelling action itself.

[0064] Step S105: Determine whether the current bird deterrence mode has failed based on the continuous habituation index and the second comprehensive activity intensity; In some embodiments, the calculated persistent habituation index is compared with a preset habituation threshold, and the second comprehensive activity intensity within the window after the action is compared with a preset residual intensity threshold. The system determines that the current bird-repelling mode has failed in habituation only when the persistent habituation index reaches or exceeds the habituation threshold and the second comprehensive activity intensity simultaneously reaches or exceeds the residual intensity threshold; otherwise, it is determined that the mode has not failed. Only when both conditions are met simultaneously does the system confirm that the current bird-repelling mode has lost its intended deterrent effect due to long-term repeated adaptation by birds.

[0065] In some embodiments, the formula for determining whether the current bird deterrence mode has become habitually ineffective is: ; In the above formula, The result of habituation failure determination for pattern m; To maintain habitual indicators, To preset the habituation threshold, The second comprehensive activity intensity, This is the preset residual strength threshold.

[0066] It should be noted that if the sustained habituation index reaches the threshold but the second comprehensive activity intensity is lower than the residual intensity threshold, it indicates that although the effect of the current pattern is continuously declining, the bird activity has dropped to a low level after the bird-driving. In this case, it may be a situation where the bird flock has left naturally, and it is not judged as habituation failure. If the second comprehensive activity intensity reaches the threshold but the sustained habituation index does not reach the threshold, it indicates that although the residual activity is high after this action, the effect has not yet shown a continuous declining trend. It may be a single action abnormality or occasional fluctuation, and it is also not judged as habituation failure.

[0067] Step S106: If the failure is determined, a set of candidate bird-repelling modes that meet the preset conditions is selected from the pre-established bird-repelling mode pool. In some embodiments, step S106 includes: reading the mode identifier, failure status mark, required execution unit list, cooling timer parameter, and execution control parameter corresponding to each bird-repelling mode in the bird-repelling mode pool; excluding bird-repelling modes whose mode identifier is the same as the mode identifier of the current mode, and bird-repelling modes whose failure status is marked as failed, from the bird-repelling mode pool to obtain a first screening result; for each bird-repelling mode in the first screening result, reading the current device working status corresponding to the required execution unit list, reading the cooling timer parameter to determine the current remaining cooling time, and obtaining the current environmental parameters to verify the safety execution conditions corresponding to the execution control parameters item by item; if there is an offline or faulty unit in the current device working status, the remaining cooling time is greater than zero, or any of the safety execution conditions are not met, then the bird-repelling mode is excluded; and the remaining bird-repelling modes after exclusion form the candidate bird-repelling mode set.

[0068] In some embodiments, after determining that the current bird-repelling mode has become habitually ineffective, the system immediately reads the pre-established bird-repelling mode pool and excludes modes with the same mode identifier as the current bird-repelling mode identifier, as well as modes marked as ineffective, from the mode pool to form a first screening result. For each remaining candidate mode in the first screening result, the system reads the current device operating status of each hardware unit corresponding to the required execution unit list for that mode. If any of the required execution units is offline or faulty, the mode is directly excluded. At the same time, the system reads the cooling timer parameter of the mode to determine the current remaining cooling time. If the remaining cooling time is greater than zero, the mode is still in the cooling period and is also excluded. Subsequently, each item in the execution control parameters is checked for its safe execution conditions. The laser irradiation boundary is used to confirm that the laser beam scanning range does not exceed the preset safe closed area. The noise limit is used to confirm that the sound pressure level does not exceed the local environmental protection or operation specifications. The gimbal mechanical stroke is used to confirm that each target point on the patrol path is within the physical limits of the gimbal's horizontal and vertical rotation angles. If any of the above verification conditions are not met, the mode is excluded from the candidates. After the aforementioned elimination process, the system will form a candidate bird-repelling mode set of all remaining bird-repelling modes. Each mode in this set must meet all the prerequisites: it is different from the current mode, it has not yet failed, all required equipment is online and in good working order, cooling has been completed, and it can be safely executed under the current environmental conditions.

[0069] It should be noted that the mode pool was established before the device was put into operation. Each bird deterrence mode in the pool has a corresponding mode identifier, failure status mark, list of required execution units, cooling timer parameters, and complete execution control parameters. The execution control parameters include specific control fields such as laser power, pulse frequency, scanning pattern, voice content, sound pressure level, playback rhythm, strong light flashing parameters, gimbal scanning path, scanning speed, and stopping point.

[0070] This standardized formula quantitatively calculates the net immediate bird deterrence effect after deducting natural variations, providing a comparable and cumulative quantitative basis for single-event attenuation indicators.

[0071] Step S107: Evaluate the bird control parameters for each bird control mode in the candidate bird control mode set to determine the target bird control mode based on the evaluation results.

[0072] In some embodiments, step S107 includes: determining the differences in key stimulus types and control methods between each candidate bird-repelling mode in the candidate bird-repelling mode set and the current bird-repelling mode; prioritizing the candidate bird-repelling mode set according to the differences in key stimulus types and control methods to obtain a prioritization result; and determining a target bird-repelling mode from the candidate bird-repelling mode set according to the prioritization result and the preset mode priority of each candidate bird-repelling mode in the corresponding target bird species and the current scene.

[0073] In some embodiments, after obtaining a set of candidate bird-repelling patterns, the key stimulus type field corresponding to each candidate bird-repelling pattern in the set is first read one by one, and the field is compared with the key stimulus type of the current failed bird-repelling pattern to determine whether the two belong to the same stimulus type; at the same time, the control method parameters such as the scanning pattern, playback content or rhythm, flashing rhythm, and patrol path of each candidate bird-repelling pattern are read and compared with the corresponding control method of the current mode item by item to determine whether there are differences in the specific control methods within the same stimulus type. Based on this, the system performs a two-level priority classification on the candidate bird deterrence mode set: The first level prioritizes candidate modes with key stimulus types different from the current mode. That is, if the current mode is laser stimulation, then modes with different stimulus types such as voice, strong light, or PTZ scanning are preferred. If there are no modes with different stimulus types from the current mode in the current candidate set, then the second level is entered, which selects candidate modes with at least one control method that is significantly different from the current mode in terms of scanning pattern, playback content or rhythm, flashing rhythm, and scanning path under the same stimulus type. For example, changing the pulse rhythm or scanning pattern in laser mode, changing the playback content or adjusting the playback rhythm in voice mode, and changing the scanning path or the distribution of stopping points in PTZ scanning mode, rather than selecting approximate modes that only change a very small angle or a small volume. After the above priority classification, if multiple candidate bird deterrence modes exist at the same priority level, the system reads the historical measurable net real-time bird deterrence effect value of each candidate mode in the corresponding target bird species and current scene, and selects the one with the highest historical effect value as the final target bird deterrence mode; if reliable historical data of the same bird species and the same scene is lacking, the system determines the unique target bird deterrence mode from that level according to the preset mode priority in the mode pool.

[0074] It should be noted that the key stimulus type field is predefined when the pattern pool is constructed, and is divided into categories such as laser, voice, strong light, gimbal scanning and their combinations according to the main deterrent means that birds can perceive.

[0075] It should be noted that the mode priority is pre-set based on the prior effectiveness ranking of various bird deterrence methods for the bird species before the device is put into operation.

[0076] By updating the benchmark through exponential moving average and combining it with dual threshold determination, we can accurately identify true failures where birds are still lingering despite continuous decline, thus avoiding false switching caused by a single fluctuation.

[0077] In some embodiments, the present invention further includes: after executing the target bird deterrence mode, using the current action event as an identifier, re-determining the update action pre-window, update action post-window, and update control window based on the re-acquired update comprehensive activity intensity sequence, and recalculating the update bird deterrence effect value; when there is no candidate bird deterrence mode set that meets the preset conditions, or when multiple different bird deterrence modes in the bird deterrence mode pool have been determined to have become habitually ineffective, entering a safety maintenance state, stopping the switching of bird deterrence modes, and generating maintenance prompt information; when a preset number of consecutive action events are marked as events with unmeasurable effects, generating sensor verification prompt information.

[0078] In some embodiments, after issuing a target bird-repelling mode to the bird-repelling execution unit and completing the mode switch, the bird-repelling action after the switch is used as a new action event identifier. An updated comprehensive activity intensity sequence is generated from the newly collected bird activity monitoring data in the same manner as described above. The updated sequence is then re-divided into a pre-update action window, a post-update action window, and an updated control window. Subsequently, the updated net real-time bird-repelling effect value of the target mode is recalculated according to the same observation completeness and identification confidence gating conditions. The verification result is written into the historical measurable effect record of the mode for priority updates during subsequent selections. When the system finds, after the aforementioned screening process, that there are no candidate bird-repelling modes that meet the safe execution conditions, are in good online condition, and are not in failure, or when multiple different bird-repelling modes in the bird-repelling mode pool have been successively determined to be habitually ineffective, the system no longer performs any mode switching operation, but enters a safe maintenance state—that is, maintaining the current bird-repelling mode at the lowest safety level. The system can either run or pause bird deterrence actions, simultaneously halting subsequent habituation detection and mode switching processes, and generating maintenance prompts. These prompts should include at least the triggering reason, the current mode identifier, a list of modes deemed invalid, the most recently measurable net real-time bird deterrence effect value, and the current operating status of each device. Based on the specific cause of failure, differentiated suggestions should be provided, such as device cleaning, adjustment of installation location or angle, expansion of the mode pool, or on-site manual verification. During normal monitoring, if the system continuously detects a preset number of action events (this number is preset as a system configuration parameter) and marks them as unmeasurable events due to insufficient recognition confidence, a low effective recording ratio, or excessive differences in the control window matching, the system will not trigger mode switching. Instead, it will generate sensor verification or communication status check prompts to guide maintenance personnel to prioritize troubleshooting external factors affecting data quality, such as sensor dirt, obstruction, device offline status, or abnormal communication links, ensuring that the bird activity monitoring data relied upon for subsequent failure determination has sufficient reliability.

[0079] This lays the resource foundation for differentiated choices and ensures that the switched model has the ability to effectively stimulate change.

[0080] To illustrate the feasibility of the present invention, the following embodiments are provided: Example 1: In an exemplary evaluation period, the system collected a comprehensive activity intensity of 0.800 before the action and 0.300 after the action, matching a decrease in activity intensity of 0.050 in the control window. Substituting these three intensity values ​​into the net immediate bird deterrence effect calculation formula, the net immediate bird deterrence effect of the current mode is approximately 0.562. Comparing this value with the historical baseline bird deterrence effect value of 0.700 for the current mode, the single-event habituation attenuation degree of this action is calculated to be approximately 0.197. If the single-event attenuation degree of the most recent five consecutive measurable action events is at a low level—for example, 0.12, 0.10, 0.09, 0.08, and 0.07 respectively—weighting these five attenuations with a time attenuation coefficient λ=0.8, the calculated continuous habituation index is approximately 0.168. This value does not reach the preset habituation threshold, therefore the system does not determine that the current mode is invalid.

[0081] In subsequent continuous monitoring, if the single-event decay rate of the five most recent measurable action events increases to 0.66, 0.64, 0.62, 0.60, and 0.58, respectively, the time decay coefficient λ=0.8 is still used for weighting, and the calculated continuous habituation index is approximately 0.629. When the overall activity intensity of the post-action window is simultaneously higher than the preset residual intensity threshold, both conditions are met, and the system determines that the current bird deterrence mode has failed to develop habituation. If an action event is marked as unmeasurable due to occlusion, device offline, or insufficient recognition confidence, then this event will not participate in the failure determination, nor will it participate in the update of the mode's baseline bird deterrence effect value.

[0082] After failure determination, the system excludes the current mode, failed modes, modes in the cooling-off period, and modes that do not meet the safe execution conditions from the preset bird deterrence mode pool. Assuming the current failed mode is a fixed-frequency laser fan-shaped scanning mode, the system prioritizes candidate modes with different key stimulus types, such as speech, strong light, or combined stimuli. If no cross-stimuli type mode exists in the candidate set, a laser mode with a pulse rhythm or scanning pattern that is significantly different from the current mode is selected. When multiple candidate modes exist at the same level of difference, the system selects the mode with the highest historical measurable net immediate bird deterrence effect for the corresponding target bird species and the current scene as the target bird deterrence mode. After executing the target mode, the system recalculates the net immediate bird deterrence effect using newly established action events and writes the verification result into the mode's historical record. The numerical values ​​in this embodiment are merely illustrative examples for explaining the calculation process and do not constitute any limitation on the scope of the invention.

[0083] Example 2: The bird deterrence mode pool can be constructed according to different execution units and output stimulus types. Based on the laser bird deterrence unit, it can form continuous light mode, pulse light mode, different pulse period mode, different scanning trajectory mode, different start and end angle mode, and different dwell point mode; based on the voice bird deterrence unit, it can form raptor call mode, similar warning sound mode, and environmental interference sound mode, and the playback frequency band, playback rhythm and volume level can be adjusted to obtain more differentiated voice modes; based on the pan-tilt-zoom (PTZ) scanning unit, it can form horizontal scanning mode, vertical scanning mode, segmented scanning mode, random dwell mode, key area repeated sweeping mode, and slow approach mode; in addition, the outputs of the above different units can be combined into laser plus voice combination mode, laser plus PTZ combination mode, voice plus PTZ combination mode, and other composite modes.

[0084] Once the current mode is deemed invalid, the system first excludes modes that are in a cooling-off period, do not meet safety limits, or have abnormal equipment status. Then, it categorizes modes based on differences in key stimulus types and control methods. For example, if the current invalid mode is a combination of "fixed-frequency laser fan-shaped scanning superimposed with fixed voice content playback," the system prioritizes modes with strong light or gimbal scanning as the primary stimulus type. If no available mode across stimulus types exists in the candidate set, the system selects modes with "different laser pulse rhythms combined with irregular gimbal paths" or "changing voice content and playback rhythm while altering the scanning pattern," rather than approximate modes that only slightly change the scanning angle or finely adjust the volume, to ensure that birds can perceive significant stimulus changes.

[0085] Example 3: In a specific application scenario, after the current bird deterrence mode is determined to have become habitually ineffective, the system reads a pre-established bird deterrence mode pool. For example, the current ineffective mode is "fixed-frequency laser fan-shaped scanning mode," and the available candidate modes in the mode pool include voice warning mode, pulsed laser irregular scanning mode, and PTZ patrol scanning mode. The system determines the target mode in the following order: first, modes that are in a cooling-off period, have abnormal equipment status, or do not meet safety boundary conditions are excluded; then, the remaining candidate modes are sorted according to the rule of "prioritizing different key stimulus types, followed by different control methods under the same stimulus type."

[0086] If both the voice alert mode and the PTZ scanning mode belong to different stimulus types than the current laser mode, the system further reads the net immediate bird deterrence effect values ​​of the most recently measurable action events recorded for these two modes in the same target bird species and similar scenarios, and selects the one with the higher historical effect value as the sole target bird deterrence mode for this time. If reliable historical data for the same bird species and scenario is lacking, the target mode is determined according to the mode priority preset by the device. After the mode is executed, the system creates a new action event, verifies the mode with the actual measured net immediate bird deterrence effect after switching, and updates the effective status and subsequent selection priority of the mode accordingly, forming a closed loop.

[0087] Example 4: During system operation, if a certain action event fails to meet the measurable conditions due to the bird activity record recognition confidence level being lower than the preset threshold, insufficient effective record ratio, or control window matching difference exceeding the threshold, the system will mark the event as having unmeasurable effect. This event will not participate in the update of the model baseline bird deterrence effect value, nor in the calculation of the continuous habituation index. The system may extend the observation window as appropriate to obtain more effective records, or send a sensor check prompt to the operation and maintenance platform.

[0088] When the system finds that there are no candidate bird deterrence modes that meet the safety execution conditions, are in good condition, and have not yet failed after screening, it will no longer perform mode switching operations, but will enter a safety hold state—that is, maintain the current mode to run at the lowest safety level or suspend the output of bird deterrence actions—and send a review suggestion to the operation and maintenance platform, prompting the device to be cleaned, the installation angle adjusted, the mode pool expanded, or the on-site bird nest hazard investigation to be carried out.

[0089] Example 5: In an optional comparative verification method, three operating modes can be set respectively under the same target bird species, weather conditions, initial activity intensity, equipment status and observation duration: fixed single mode configuration, fixed cycle rotation configuration and effect-driven rotation configuration of the present invention. All three configurations record bird activity records before and after the action, mode identifiers, net immediate bird deterrence effect of each action, idling time after failure, number of mode switching times and number of times safety conditions are rejected.

[0090] The effect-driven configuration of this invention triggers switching only when the sustained habituation index of a measurable event and the residual activity intensity after the action simultaneously reach a preset threshold, and selects a target mode with a significant stimulus difference from modes that meet safety, equipment, and cooling conditions. By comparing the idling time after failure, the number of invalid switching attempts, and the measured net immediate bird deterrence effect after switching under the three configurations, the improvement of the effect-driven triggering mechanism and the differentiated target mode selection method compared to fixed single mode and fixed cycle rotation can be verified. The specific threshold settings and statistical results are based on field tests or actual project operation records, and the above verification methods do not constitute a necessary limitation of this invention.

[0091] Compared with the prior art, the present invention has at least the following beneficial effects: (1) By correcting the changes in activity intensity before and after bird deterrence actions through a matching control window, the changes caused by weather, time of day or natural departure of bird flocks can be distinguished from the net effect caused by bird deterrence actions. On this basis, habituation failure is determined by the dual threshold of attenuation index and residual intensity, reducing the risk of misjudging natural fluctuations as bird deterrence effects or habituation. (2) After failure, adopt the direct and interpretable next mode selection rule, first exclude unsafe, unexecutable, cooling or failed modes, and then prioritize the selection of modes with significantly different key stimulus types or control methods, and avoid switching to modes with different numbers but similar stimuli. (3) When there are multiple candidate modes at the same difference level, a next target mode is determined based on the target bird species and the historical measurable net real-time bird deterrence effect in the current scene or a preset priority. There is no need to solve the multi-mode sequence, which makes it easy to execute stably on edge devices. (4) After switching, re-establish the action event, match the control window and calculate the net real-time bird deterrence effect. Update the effective status of the selected mode and the subsequent selection priority with the actual test results to form a closed loop of "judgment failure - select a next mode - perform verification - judgment again". (5) The data dependencies of the whole scheme are clear: bird activity records form a comprehensive activity intensity time series; the same series forms the average values ​​of the pre-action, post-action and control windows; measurable events are used to calculate net immediate effect and continuous decay; dual threshold results trigger the selection of the next mode; the mode pool, safety conditions, stimulus differences and historical measurable effects jointly determine a next target mode, and the output of the previous step constitutes the input of the next step.

[0092] (6) Unlike existing technologies: Existing bird control products mainly rely on timed rotation, random rotation, or strategy library matching to trigger mode switching. The timing of rotation is unrelated to the actual bird response and cannot determine whether the current mode has truly become habituated and lost its effectiveness. This invention uses the net effect difference between the action area and the control area as the basis for habituation judgment. It uses a combination of continuous decay index and residual intensity thresholds to determine failure, and only triggers mode switching when the failure judgment is established, rather than relying on preset time intervals or random decisions. This makes the rotation timing driven by effect rather than by rules, which can avoid premature switching when the bird control mode is still effective and maintaining the original mode for a long time after habituation has lost its effectiveness.

[0093] This invention collects bird activity monitoring data and normalizes it into a comprehensive activity intensity sequence, transforming multi-source heterogeneous observation information into a unified quantitative measurement benchmark, providing a comparable and calculable data foundation for subsequent effect evaluation. Based on this, a time window mechanism using action events as identifiers is established, defining a pre-action window, a post-action window, and a control window. The control window is used to deduct activity intensity changes caused by non-bird-repelling factors such as weather, time of day, and natural bird migration, ensuring that the calculated net immediate bird-repelling effect accurately reflects the true deterrence contribution of the bird-repelling action itself, avoiding the bias of misjudging external factor fluctuations as bird-repelling effectiveness. Furthermore, by comparing this net immediate bird-repelling effect value with the current model's baseline bird-repelling effect value and generating a continuous habituation index, the trend of gradual decay of bird-repelling effect with repeated use can be captured from a time dimension. Combined with the second comprehensive activity intensity of the post-action window for dual-threshold joint judgment, it confirms the continuous decay of the effect while further verifying that birds maintain a high level of activity after the bird-repelling action. Failure judgment is triggered only when both conditions are met, effectively excluding the possibility that birds have already left naturally. The system mitigates the misleading influence of interference such as abnormal single observations on switching decisions, ensuring a precise correspondence between the switching trigger timing and the birds' actual habituation state. After determining failure, a candidate mode set is further screened using preset conditions, and the executable control parameters of the candidate modes are evaluated to determine the target mode. This ensures that the selected target mode not only has sufficient stimulus difference from the current failed mode to break the birds' adaptive expectations, but also meets safety boundaries and the device's executable constraints, avoiding the selection of unsafe or unsupported modes in pursuit of stimulus difference. Ultimately, each step is interconnected and progressive, jointly driving the bird deterrent device to switch only when the current mode has truly become habituated and failed. During the switch, an executable target mode is selected from the available differentiated modes, ensuring that the timing and direction of mode switching are driven by the birds' actual behavioral responses, rather than relying on fixed periods, random numbers, or scene rule matching. This achieves the technical effect of "precisely switching bird deterrent modes based on the birds' actual responses," effectively extending the long-term effective operating time of the bird deterrent device, reducing ineffective idling and equipment wear, and lowering manual inspection and maintenance costs.

[0094] Based on the above method embodiments, corresponding apparatus embodiments are provided; An embodiment of the present invention provides a bird-repelling mode switching system for a bird-repelling device, comprising: the system being used to perform the steps of the bird-repelling mode switching method of the bird-repelling device as described in the present invention.

[0095] It is understood that the above-described device embodiments correspond to the method embodiments of the present invention, and can implement the bird-repelling mode switching method of the bird-repelling device provided by any of the above-described method embodiments of the present invention.

[0096] It should be noted that the device embodiments described above are merely illustrative, and some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can specifically be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0097] Based on the above-described embodiments of the bird-repelling device bird-repelling mode switching method, another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the bird-repelling mode switching method of any embodiment of the present invention.

[0098] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the terminal device.

[0099] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0100] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.

[0101] Based on the above-described method embodiments, another embodiment of the present invention provides a computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the bird-repelling mode switching method of the bird-repelling device described in any of the above-described method embodiments of the present invention.

[0102] The modules / units integrated in the device / terminal equipment, if implemented as software functional units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0103] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for switching bird-repelling modes of a bird-repelling device, characterized in that, include: Collect bird activity monitoring data within the area of ​​effect of the bird deterrent device, and determine a comprehensive activity intensity sequence based on the bird activity monitoring data; Using action events as identifiers, the pre-action window, post-action window, and control window are determined based on the comprehensive activity intensity sequence. The net instant bird deterrence effect value is calculated based on the first comprehensive activity intensity corresponding to the pre-action window, the second comprehensive activity intensity corresponding to the post-action window, and the third comprehensive activity intensity corresponding to the control window. Based on the net immediate bird deterrence effect value and the baseline bird deterrence effect value of the current bird deterrence mode of the bird deterrence device, a continuous habituation index is obtained. The current bird deterrence mode is determined to be ineffective based on the continuous habituation index and the second comprehensive activity intensity. If the bird deterrence mode fails, a set of candidate bird deterrence modes that meet the preset conditions will be selected from the pre-established bird deterrence mode pool. The bird control parameters for each bird control mode in the candidate bird control mode set are evaluated to determine the target bird control mode based on the evaluation results.

2. The bird-repelling mode switching method of the bird-repelling device according to claim 1, characterized in that, The determination of the comprehensive activity intensity sequence based on the bird activity monitoring data includes: Determine the identification confidence level of each data point in the bird activity monitoring data; Data with an identification confidence level lower than a preset confidence threshold are filtered out from the bird activity monitoring data to obtain valid activity monitoring data; Normalization is performed on each available activity feature in the effective activity monitoring data to obtain the normalization result; Based on the normalization results described above, a comprehensive activity intensity sequence updated over time is obtained.

3. The bird-repelling mode switching method of the bird-repelling device according to claim 1, characterized in that, The calculation of the net immediate bird deterrence effect value based on the first comprehensive activity intensity corresponding to the pre-action window, the second comprehensive activity intensity corresponding to the post-action window, and the third comprehensive activity intensity corresponding to the control window includes: The ratio of the actual number of valid activity monitoring data to the theoretical sampling number is calculated separately in the pre-action window, the post-action window, and the control window, and is used as the observation completeness of each window; The arithmetic mean of the identification confidence scores corresponding to the valid activity monitoring data in the pre-action window, post-action window, and control window is calculated respectively, and used as the average identification confidence score for each window; If the observation completeness of all windows is not lower than the preset completeness threshold and the average identification confidence of all windows is not lower than the preset average confidence threshold, then the front-end comprehensive activity intensity and the back-end comprehensive activity intensity of the control window are determined according to the third comprehensive activity intensity. Calculate the first difference between the first comprehensive activity intensity and the second comprehensive activity intensity, and the second difference between the front-end comprehensive activity intensity and the back-end comprehensive activity intensity; The net decrease obtained by subtracting the second difference from the first difference is used as the numerator, and the sum of the first comprehensive activity intensity and the preset zero constant is used as the denominator to calculate the original bird deterrence effect value. The original bird deterrence effect value is limited to a preset range to obtain the net real-time bird deterrence effect value.

4. The bird-repelling mode switching method of the bird-repelling device according to claim 1, characterized in that, The continuous habituation index is obtained based on the net immediate bird deterrence effect value and the baseline bird deterrence effect value of the current bird deterrence mode of the bird deterrence device, including: Determine the baseline bird-repelling effect value of the current bird-repelling mode of the bird-repelling device, wherein the baseline bird-repelling effect value is obtained by updating the net instantaneous bird-repelling effect value corresponding to each measurable action event of the current bird-repelling mode during the period when habituation failure is not determined through exponential moving average. The net instant bird deterrence effect value is compared with the baseline bird deterrence effect value to obtain the degree of single habituation decay corresponding to the action event. The continuous habituation index is obtained by weighting the degree of single habituation decay corresponding to each of the most recent consecutive measurable action events under the current bird deterrence mode with time decay.

5. The bird-repelling mode switching method of the bird-repelling device according to claim 1, characterized in that, The step of selecting a set of candidate bird-repelling patterns that meet preset conditions from a pre-established pool of bird-repelling patterns includes: Read the mode identifier, failure status flag, required execution unit list, cooldown timer parameters and execution control parameters corresponding to each bird-repelling mode in the bird-repelling mode pool; Bird-repelling modes whose mode identifier is the same as the current mode's mode identifier, and bird-repelling modes whose failure status is marked as failed, are excluded from the bird-repelling mode pool to obtain the first screening result. For each bird-repelling mode in the first screening result, the current working status of the device corresponding to the required execution unit list is read, the cooling timer parameter is read to determine the current remaining cooling time, and the current environmental parameters are obtained to verify the safety execution conditions corresponding to the execution control parameters item by item. If there is an offline or faulty unit in the current device working status, the remaining cooling time is greater than zero, or any of the safety execution conditions are not met, then the bird-repelling mode is excluded. The remaining bird-repelling patterns after elimination are combined into the candidate bird-repelling pattern set.

6. The bird-repelling mode switching method of the bird-repelling device according to claim 1, characterized in that, The step of evaluating the bird control parameters for each bird control mode in the candidate bird control mode set, and determining the target bird control mode based on the evaluation results, includes: Determine the key stimulus type differences and control method differences between each candidate bird-repelling pattern in the candidate bird-repelling pattern set and the current bird-repelling pattern; Based on the differences in the key stimulus types and the differences in the control methods, the candidate bird deterrence pattern set is prioritized and graded to obtain the grading results. Based on the grading results and the preset mode priority of each candidate bird deterrence mode in the corresponding target bird species and the current scene, a target bird deterrence mode is determined from the set of candidate bird deterrence modes.

7. The bird-repelling mode switching method of the bird-repelling device according to claim 1, characterized in that, Also includes: After executing the target bird deterrence mode, the current action event is used as an identifier. Based on the newly acquired updated comprehensive activity intensity sequence, the window before the updated action, the window after the updated action, and the updated control window are re-determined, and the updated bird deterrence effect value is recalculated. When there is no set of candidate bird deterrence modes that meet the preset conditions, or when multiple different bird deterrence modes in the bird deterrence mode pool have been determined to have become habitually ineffective, the system enters a safety maintenance state, stops switching bird deterrence modes, and generates maintenance prompt information. When a preset number of consecutive action events are marked as events with unmeasurable effects, a sensor verification prompt message is generated.

8. A bird-repelling mode switching system for a bird-repelling device, characterized in that, include: The system is used to execute the bird-repelling mode switching method of the bird-repelling device as described in any one of claims 1-7.

9. A terminal device, characterized in that, include: One or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the steps of the bird-repelling mode switching method of the bird-repelling device as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, include: A stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform the steps of the bird-repelling mode switching method of the bird-repelling device as described in any one of claims 1-7.

Images