A method for automated and intelligent management of laying hen farming

By establishing virtual queues and synchronization boundaries, and identifying delivery offsets and queue count mismatches, the problem of erroneous propagation of abnormal egg sources in egg-laying hen farming was solved, enabling accurate positioning and control, and improving detection accuracy.

CN123100560APending Publication Date: 2026-11-17CHAOYI EGG PRODUCTS (SHANGHAI) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611369235.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-04
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

In large-scale egg-laying hen farms, after multiple source branches converge into the main conveyor line, due to reasons such as conveyor deviation, partial slippage of eggs, missed detection or repeated detection, the correspondence between the actual eggs and the virtual source queue is mismatched, causing the problem of erroneous egg source propagation.

Method used

By acquiring egg input detection events and conveying status signals, a virtual queue carrying source attributes is established. Conveying segment offset and queue count mismatch are identified, the source branch is controlled to stop inputting eggs, a synchronization boundary is formed, and constraint matching is performed to determine the source information of abnormal eggs.

Benefits of technology

Without altering the mechanical structure of the egg collection and conveying system, accurate location of the source of abnormal eggs and management of breeding were achieved, reducing the risk of spreading abnormal information and improving detection accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123100560A_ABST
    Figure CN123100560A_ABST
Patent Text Reader

Abstract

This invention relates to the field of intelligent livestock and poultry farming and automatic control technology, and discloses an automated intelligent farming management method for laying hens. The method includes: acquiring egg input detection events and conveying status signals from multiple source branches; establishing a virtual queue carrying source attributes and generating predicted passing events; identifying queue mismatches based on the time, quantity, and sequence residuals of actual passing events and predicted passing events; determining the pre-synchronization boundary and source confusion section when a mismatch occurs, stopping source branch input while maintaining the main conveyor line operation to form a continuous egg-free state and determining the post-synchronization boundary; performing constraint matching on actual passing events and virtual queue items within the source confusion section; associating egg abnormality detection events with the matching results to determine the source or candidate source of abnormal eggs and generating farming management instructions. This invention eliminates the need for physical tags, enables source tracking and source reconstruction after mismatch, and improves the reliability of source traceability.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of intelligent livestock and poultry farming and automatic control technology, specifically to an automated intelligent farming and management method for laying hens. Background Technology

[0002] Large-scale egg-laying hen farms typically have multiple egg collection branches. Each branch receives eggs from different chicken houses, cage rows, cage layers, or other breeding areas, and they converge into the main conveyor line during transport. The main conveyor line then transports the eggs to grading, appearance inspection, crack inspection, soiling inspection, or other egg product inspection stations.

[0003] In the case of multi-source branch line confluence conveying, in order to reverse-engineer the source of abnormal eggs detected downstream, the conveying position and source relationship of eggs can be tracked based on the egg input detection signals from each source branch line, the operating status of the conveyor line, and the passing signals of eggs at the detection nodes. This method avoids the need to place physical tags on the surface of each egg.

[0004] However, in actual transportation, factors such as conveyor belt start / stop, conveyor speed fluctuations, egg rolling or sliding relative to the conveyor surface, adjacent eggs touching each other, and missed or duplicate detections at detection nodes can all cause time, quantity, or sequence discrepancies between the actual egg passage event and the predicted queue established based on the input detection events. When the correspondence between a certain egg and a virtual queue item becomes incorrect, if the source attribute continues to be propagated according to the original queue order, it is easy for the source information of multiple subsequent eggs to be continuously incorrect, thus causing abnormal eggs detected downstream to be incorrectly associated with other source branches or breeding areas. Summary of the Invention

[0005] The purpose of this invention is to provide an automated intelligent management and control method for laying hens, which solves the technical problem that after multiple source branches converge into the main conveyor line, the correspondence between the actual eggs and the virtual source queue is mismatched due to conveyor offset, partial slippage of eggs, missed detection, repeated detection or abnormal event sequence, thereby causing the erroneous propagation of abnormal egg sources.

[0006] The objective of this invention can be achieved through the following technical solutions: An automated and intelligent management method for laying hen farming, the method comprising: S1: Obtain egg input detection events from multiple source branches and the conveying status signals of the source branches and the main conveyor line, establish a virtual queue carrying source attributes, and generate predicted pass events for each virtual queue item at multiple detection nodes; S2: Based on the distribution of time residuals, quantity residuals, and sequence residuals between actual passing events and predicted passing events, identify overall shift, local slippage shift, or queue count mismatch of the transport section; S3: When queue counting mismatch is identified, determine the pre-synchronization boundary and source confusion section, control the source branch to stop inputting eggs into the main conveyor line, and keep the main conveyor line running until the downstream detection node detects a continuous state of no eggs; S4: Determine the blank conveying interval corresponding to the continuous egg-free state as the post-synchronization boundary, and restore the source branch line to input eggs into the main conveying line in a predetermined order; S5: Based on the number of source branch inputs between the pre-synchronization boundary and the post-synchronization boundary, the number of actual passing events, the input order, and the allowed transmission time interval, perform constraint matching on the actual passing events and virtual queue items in the source confusion section; S6: Associate the egg product anomaly detection event with the constraint matching result to determine the source information of the abnormal eggs and generate the corresponding breeding control instructions.

[0007] As a further aspect of the present invention: in step S1, each virtual queue item includes at least a source branch identifier, a corresponding cage or cage layer identifier, an input detection time, an input sequence within the branch, an initial encoder reading, a predicted position, and a predicted passage time interval; the transport displacement is calculated based on the start / stop status in the transport status signal, the encoder cumulative pulse, and the pulse displacement conversion coefficient, and the predicted position and predicted passage time interval are updated based on the transport displacement.

[0008] As a further aspect of the present invention: in step S2, identifying the overall offset or local slippage offset of the conveyor section includes: When the time residuals of at least a set number of consecutive actual passing events at the same detection node have the same sign, the absolute value of the difference between each time residual and the average value of the time residuals of the set number of consecutive actual passing events is less than the first threshold and the quantity residual is zero, it is determined that the overall displacement of the conveying section is corrected according to the average value of the time residuals. When the absolute value of the time residual of the actual passing event of the target is greater than the second threshold, and the absolute values ​​of the time residuals of the preceding and following actual passing events that are immediately adjacent to the actual passing event of the target are not greater than the second threshold and the quantity residual is zero, it is determined to be a local slip offset. The time offset and pulse displacement conversion coefficient of the corresponding transport segment are kept unchanged, the predicted passing time interval of the virtual queue item corresponding to the actual passing event of the target is expanded, and the virtual queue item and its adjacent virtual queue items in the same source branch are used as the candidate virtual queue item set of the actual passing event of the target.

[0009] As a further aspect of the present invention: In step S2, when the correspondence between the actual passing event and the predicted passing event satisfies at least one of the following conditions, it is determined that the queue counting is mismatched: within a set statistical interval, the number of actual passing events is not equal to the number of predicted passing events; the actual passing time of an actual passing event falls within the predicted passing time interval of two or more virtual queue items simultaneously; the same virtual queue item does not have a corresponding actual passing event at two or more consecutive detection nodes; the order of the actual passing events corresponding to each virtual queue item within the same source branch is inconsistent with the input order of each virtual queue item; After determining that the queue count mismatch is true, the assignment of source values ​​to each virtual queue item in the source confusion section is stopped, and all candidate correspondences are retained until the source attribute is re-determined through constraint matching in step S5.

[0010] As a further aspect of the present invention: in steps S3 and S4, the continuous egg-free state is determined by the continuous egg-free time of the downstream detection node or the continuous idle encoder displacement of the main conveyor line; when the continuous idle encoder displacement is not less than the sum of the maximum occupied length of the egg in the conveying direction and twice the maximum allowable position error, it is determined that the blank conveying interval is formed, so as to isolate the source confusion section and subsequent eggs through the blank conveying interval.

[0011] As a further aspect of the present invention: in step S4, the step of restoring the input of eggs from the source branch line to the main conveyor line in a predetermined order includes: After the post-synchronization boundary is formed, the target source branch to be restored is determined in a predetermined order, the target source branch is controlled to resume feeding eggs into the main conveyor line, and the remaining source branches are controlled to remain stopped feeding eggs into the main conveyor line. The first egg input detection event after the target source branch resumes input is obtained, and the first actual pass event obtained by the downstream detection node after the post-synchronization boundary is obtained. The first actual pass event is bound to the source attribute of the target source branch, and the bound virtual queue item is used as the starting item of the virtual queue after the post-synchronization boundary. After binding is completed, the next source branch is determined as the target source branch in the predetermined order and the above steps are repeated until all source branches resume feeding eggs into the main conveyor line.

[0012] As a further aspect of the present invention: in step S5, the constraint matching includes: Feasible matching results between actual passing events and virtual queue items within the source confusion segment are generated according to uniqueness constraints, intra-source order constraints, and delivery time constraints. Specifically, the uniqueness constraint ensures that each actual passing event matches at most one virtual queue item, and each virtual queue item matches at most one actual passing event; the intra-source order constraint ensures that the order of actual passing events matched by virtual queue items belonging to the same source branch is the same as the input order of the virtual queue items; and the delivery time constraint ensures that the actual passing time of an actual passing event falls within the allowed delivery time interval of its matched virtual queue item. When there is only one feasible matching result, the source attribute corresponding to the virtual queue item in the feasible matching result is output as the determined source; when there are two or more feasible matching results, the union of the source attributes corresponding to the same actual passing event in each feasible matching result is taken and output as the candidate source set of the actual passing event.

[0013] As a further aspect of the present invention: in step S6, generating the corresponding aquaculture control instructions includes: Count the number or proportion of abnormal eggs associated with the same source branch, cage column or cage layer within a preset time window; When the number of abnormal eggs or the proportion of abnormal eggs reaches a preset threshold, a separate operation instruction is generated. The separate operation instruction is used to control the source branch to input eggs into the main conveyor line separately in the next egg collection cycle, and to control the other source branches to stop inputting eggs into the main conveyor line in the egg collection cycle. Acquire egg product anomaly detection events during the individual operation period, and associate the egg product anomaly detection events with the source attribute corresponding to the source branch; When the number or proportion of abnormal eggs during the individual operation still reaches the preset threshold, an inspection instruction is generated for the breeding area corresponding to that source branch.

[0014] The beneficial effects of this invention are: This invention establishes a virtual queue carrying source attributes by using egg input detection events, transport status signals, and actual pass events of detection nodes, enabling source tracking without setting physical tags on eggs. When a mismatch is detected in the queue, new source inputs are stopped, and a blank transport interval is formed for resynchronization, isolating the transported objects before and after the mismatch. Furthermore, the source confusion section is constrained and reconstructed using transport time, internal source order, and quantity relationships. When the source cannot be uniquely determined, candidate sources are retained instead of being forcibly assigned incorrect values, thereby reducing the risk of a single detection anomaly spreading to subsequent eggs. Egg anomaly results can also be used for verification of the corresponding source area or for breeding inspections. Attached Figure Description

[0015] The invention will now be further described with reference to the accompanying drawings.

[0016] Figure 1 This is a flowchart of a method according to an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figure 1 As shown, in one embodiment, an automated intelligent farming and management method for laying hens is provided, the method comprising: S1: Obtain egg input detection events from multiple source branches and the conveying status signals of the source branches and the main conveyor line, establish a virtual queue carrying source attributes, and generate predicted pass events for each virtual queue item at multiple detection nodes; S2: Based on the distribution of time residuals, quantity residuals, and sequence residuals between actual passing events and predicted passing events, identify overall shift, local slippage shift, or queue count mismatch of the transport section; S3: When queue counting mismatch is identified, determine the pre-synchronization boundary and source confusion section, control the source branch to stop inputting eggs into the main conveyor line, and keep the main conveyor line running until the downstream detection node detects a continuous state of no eggs; S4: Determine the blank conveying interval corresponding to the continuous egg-free state as the post-synchronization boundary, and restore the source branch line to input eggs into the main conveying line in a predetermined order; S5: Based on the number of source branch inputs between the pre-synchronization boundary and the post-synchronization boundary, the number of actual passing events, the input order, and the allowed transmission time interval, perform constraint matching on the actual passing events and virtual queue items in the source confusion section; S6: Associate the egg product anomaly detection event with the constraint matching result to determine the source information of the abnormal eggs and generate the corresponding breeding control instructions.

[0019] Specifically, this embodiment is applied to an existing egg collection and conveying system for laying hens. The egg collection and conveying system includes multiple source branches, a main conveying line connected to each source branch, multiple detection nodes set along the conveying path, egg quality anomaly detection nodes, and a controller. Each source branch corresponds to a different hen house, cage row, cage layer, or egg collection area. The method utilizes signals output from existing detection equipment, conveying control equipment, and egg quality detection equipment, without altering the mechanical structure of the egg collection and conveying system.

[0020] In step S1, the controller acquires the egg input detection events from each source branch, as well as the conveying status signals from each source branch and the main conveyor line.

[0021] The egg input detection event indicates that an egg has entered the corresponding source branch, including the source branch identifier, the input detection time, and the input sequence within the branch. The conveying status signal indicates the operating status, conveying speed, conveying displacement, or conveying section switching status of the source branch and the main conveyor line.

[0022] When an egg is detected entering a source branch, the controller establishes a virtual queue entry based on the source branch identifier and input order corresponding to that egg. The virtual queue entry represents the virtual transport object of the corresponding egg during the egg collection and transport process, and carries the source attribute corresponding to the source branch, chicken house, cage row, cage layer, or egg collection area.

[0023] Based on the egg's input detection time, the transport path information from the source branch to each detection node, and the transport status signal, the controller predicts the arrival time and order of each virtual queue item at multiple detection nodes, and generates a predicted pass event for each virtual queue item at the detection node.

[0024] In step S2, the controller acquires the actual passing events of the eggs detected by each detection node and compares the actual passing events with the corresponding predicted passing events.

[0025] The actual passage event includes at least the detection node identifier, the actual passage time, the actual passage order, and the actual number of passages within the set detection interval. The controller determines the time residual, quantity residual, and order residual based on the actual passage event and the predicted passage event.

[0026] Among them, the time residual is used to represent the difference between the actual time when the egg passes through the detection node and the predicted time; the quantity residual is used to represent the difference between the actual number of passing events and the predicted number of passing events within the set detection interval; and the order residual is used to represent the difference between the actual passing order and the predicted passing order.

[0027] The controller identifies the conveying status based on the changes in the time residuals in consecutive actual passing events, the changes in the quantity residuals, and whether the sequence residuals satisfy the input order within the source branch. When multiple consecutive actual passing events show time deviations in the same direction and with similar magnitudes, it is identified as an overall shift in the conveying section; when the time deviations are concentrated in a single or a small number of actual passing events, and adjacent actual passing events do not show the same deviation, it is identified as a local slippage shift; when a stable correspondence cannot be established between actual passing events and predicted passing events, it is identified as a queue count mismatch.

[0028] In step S3, when a queue count mismatch is identified, the controller determines the position of the last virtual queue that can establish a reliable correspondence with the actual event before the queue mismatch occurs, and determines this position as the pre-synchronization boundary.

[0029] The controller identifies the transport segment following the pre-synchronization boundary, where a reliable source correspondence has not yet been restored, as a source confusion segment. Within this source confusion segment, there are instances where the actual passing events and virtual queue items have uncertain sources, inconsistent quantities, or inconsistent sequences.

[0030] Subsequently, the controller stops each source branch from feeding new eggs into the main conveyor line, while maintaining the main conveyor line's operation to allow eggs already on the main conveyor line to continue being transported downstream. The controller continuously acquires the detection results from downstream detection nodes.

[0031] A continuous egg-free state is determined when no egg passage event is detected by the downstream detection node within a continuous detection time or within a continuous transport displacement range. This continuous egg-free state indicates that the eggs within the source confusion section have passed the corresponding downstream detection position, providing an empty interval for subsequent queue resynchronization.

[0032] In step S4, the controller determines the idle section of the main conveyor line corresponding to the continuous egg-free state as the post-synchronization boundary. The pre-synchronization boundary and the post-synchronization boundary together define the range of the source confusion section.

[0033] After determining the post-synchronization boundary, the controller restores the input of eggs to the main conveyor line from each source branch according to the predetermined source branch recovery sequence. The controller uses the post-synchronization boundary as the re-establishment position for subsequent virtual queues and re-establishes subsequent virtual queue items based on the restored egg input detection events and conveyor status signals.

[0034] For a newly established virtual queue item, the controller regenerates its predicted pass event at the downstream detection node and corrects the predicted pass event using the actual pass event obtained from the downstream detection node. When the actual pass event and the regenerated predicted pass event maintain a correspondence, the virtual queue resumes normal tracking after the post-synchronization boundary is determined.

[0035] In step S5, the controller acquires the transport data between the front synchronization boundary and the back synchronization boundary, including the number of eggs input by each source branch in the section, the actual number of passing events, the order of egg input, and the allowable transport time interval corresponding to each source branch.

[0036] Based on the aforementioned transmission data, the controller establishes candidate correspondences between actual passing events and virtual queue items within the source confusion section. Candidate correspondences that do not satisfy the relationships of the number of inputs from the source branch, the number of actual passing events, the input order within the same source branch, or the allowed transmission time interval are deleted by the controller; candidate correspondences that satisfy the above relationships are retained by the controller.

[0037] When a given pass event corresponds to only one source attribute after candidate relationship filtering, the controller determines that source attribute as the source information for the corresponding egg. When a given pass event still corresponds to multiple source attributes, the controller retains the multiple source attributes as candidate source information and does not force the given pass event to be assigned to a single source.

[0038] In step S6, the egg product anomaly detection node performs anomaly detection on the passing eggs and outputs an egg product anomaly detection event. The egg product anomaly detection event includes at least the anomaly detection time, the detection node identifier, the actual passing order, and the anomaly detection result.

[0039] When an abnormal egg is detected by the egg product anomaly detection node, the controller searches for the corresponding virtual queue item in the virtual queue re-established after the constraint matching result of the source confusion section or the post-synchronization boundary, based on the detection time of the anomaly detection event, the detection node identifier, and the actual passage order.

[0040] When a unique source can be identified, the controller determines the source branch, chicken house, cage row, cage layer, or egg collection area corresponding to the abnormal egg; when a unique source cannot be identified, the controller outputs candidate source information corresponding to the abnormal egg.

[0041] The controller generates aquaculture control instructions based on the determined source information or candidate source information, and sends the aquaculture control instructions to the corresponding conveying control equipment or aquaculture management terminal to perform abnormal source recording, source area verification, conveying process inspection or aquaculture area inspection.

[0042] Through the above technical solution, this embodiment establishes and maintains a virtual queue by utilizing egg input detection events, conveying status signals, and multiple detection nodes without changing the existing mechanical structure of the egg collection and conveying system or placing labels on the surface of the eggs. When a count mismatch occurs in the virtual queue, a continuous egg-free state is formed by stopping the input of the source branch line and keeping the main conveying line running, thereby establishing a new synchronization boundary. Then, constraint matching is performed on the source confusion sections between the synchronization boundaries to determine or narrow down the source range of abnormal eggs, and the source information is used for subsequent breeding management.

[0043] In one embodiment of the present invention, in step S1, each virtual queue item includes at least a source branch identifier, a corresponding cage or cage layer identifier, an input detection time, an input sequence within the branch, an initial encoder reading, a predicted position, and a predicted passage time interval; the transport displacement is calculated based on the start / stop status in the transport status signal, the encoder cumulative pulse, and the pulse displacement conversion coefficient, and the predicted position and predicted passage time interval are updated based on the transport displacement.

[0044] Specifically, in this embodiment, the controller establishes a corresponding virtual queue entry for each egg input detection event on the source branch. The virtual queue entry is stored in the controller's storage unit and is used to record the source information and predicted delivery status of the corresponding egg.

[0045] Each virtual queue item includes at least the source branch identifier, the corresponding cage or cage layer identifier, the input detection time, the input order within the branch, the initial encoder reading, the prediction position, and the prediction pass-through time interval.

[0046] The source branch identifier is determined by the source branch that detected the egg input event. The controller pre-stores the mapping relationship between source branches and cage rows or cage layers; after obtaining the source branch identifier, it determines the corresponding cage row or cage layer identifier according to the mapping relationship and writes it into the virtual queue item. The input order within a branch is determined according to the output order of each egg input detection event in the same source branch.

[0047] The controller acquires the start / stop status, encoder cumulative pulses, and pulse displacement conversion factor from the conveying status signal. The encoder cumulative pulses represent the cumulative pulse change of the corresponding conveying section since the reference time, and the pulse displacement conversion factor represents the conveying distance corresponding to a single encoder pulse.

[0048] The controller determines the conveying displacement of the corresponding egg in the current conveying section based on the pulse difference between the current cumulative encoder pulse and the initial encoder reading, as well as the pulse displacement conversion coefficient, and updates the predicted position of the virtual queue item based on the conveying displacement.

[0049] When the conveyor section is stopped, the controller maintains the predicted position of the corresponding virtual queue item unchanged. When the conveyor section resumes operation, the controller continues to update the predicted position based on the cumulative pulse changes of the encoder. When the predicted position of the virtual queue item reaches the path length corresponding to the end of the source branch, the controller switches the conveyor section of the virtual queue item to the main conveyor line according to the preset connection relationship between the source branch and the main conveyor line, and updates the predicted position to the starting position or corresponding confluence position of the main conveyor line.

[0050] The controller pre-stores the position parameters of each detection node on the corresponding conveyor section. When the predicted position of the virtual queue item is close to the position parameters of the target detection node, the controller generates a predicted passage time based on the current predicted position, the position parameters of the target detection node, the conveyor status, and the current conveyor speed. It also determines the prediction error range based on the encoder resolution, conveyor speed fluctuations, and the response time of the detection node to form a predicted passage time interval.

[0051] When the conveying status, conveying speed, or encoder cumulative pulses change, the controller recalculates the predicted position and predicted passage time interval for virtual queue items that have not yet reached the target detection node. For virtual queue items that have already generated actual passage events at the detection node, the controller retains the corresponding predicted passage record and actual passage record.

[0052] Through the above technical solution, this embodiment establishes a virtual queue item for each egg entering the source branch line, which includes source attributes, input order, initial encoder reading, predicted position, and predicted passage time interval. By dynamically updating the predicted position of the egg using the conveyor belt start / stop status, encoder cumulative pulses, and pulse displacement conversion coefficient, it achieves continuous tracking of the egg's transport position and expected arrival time in the source branch line and main conveyor line without setting physical tags on the eggs. This provides a data foundation for subsequent matching between actual and predicted passage events, identification of transport deviations, reconstruction of source confusion sections, and location of abnormal egg sources.

[0053] In one embodiment of the present invention, step S2, identifying the overall offset or local slip offset of the conveyor section, includes: When the time residuals of at least a set number of consecutive actual passing events at the same detection node have the same sign, the absolute value of the difference between each time residual and the average value of the time residuals of the set number of consecutive actual passing events is less than the first threshold and the quantity residual is zero, it is determined that the overall displacement of the conveying section is corrected according to the average value of the time residuals. When the absolute value of the time residual of the actual passing event of the target is greater than the second threshold, and the absolute values ​​of the time residuals of the preceding and following actual passing events that are immediately adjacent to the actual passing event of the target are not greater than the second threshold and the quantity residual is zero, it is determined to be a local slip offset. The time offset and pulse displacement conversion coefficient of the corresponding transport segment are kept unchanged, the predicted passing time interval of the virtual queue item corresponding to the actual passing event of the target is expanded, and the virtual queue item and its adjacent virtual queue items in the same source branch are used as the candidate virtual queue item set of the actual passing event of the target.

[0054] Specifically, this embodiment is used to distinguish between the overall offset of the conveyor section and the local slippage offset of a single egg based on the time residuals of multiple actual passing events at the same detection node. The time residual is determined by the difference between the actual passing time and the corresponding predicted passing time; when the actual passing time is later than the predicted passing time, the time residual is positive, and when the actual passing time is earlier than the predicted passing time, the time residual is negative.

[0055] The controller constructs an analysis sequence based on consecutively acquired actual passing events at the same detection node. This analysis sequence contains at least a predetermined number of actual passing events. The predetermined number is determined based on the detection frequency of the detection node, the conveyor line speed, and the normal egg input interval.

[0056] The controller calculates the average time residual of each actual passing event in the analysis sequence, and determines whether each time residual has the same sign, whether the absolute value of the difference between each time residual and the average value is less than a first threshold, and whether the quantity residual in the corresponding detection interval of the analysis sequence is zero.

[0057] When the above conditions are met simultaneously, it is determined that the corresponding conveyor segment has experienced an overall offset. The controller corrects the time offset of the conveyor segment based on the average time residual. If the time residual continuously changes with the increase of the conveying displacement, and the conveying status signal does not indicate that the corresponding conveyor segment has a start / stop switch or speed adjustment, the controller corrects the pulse displacement conversion coefficient based on the relationship between the time residual and the conveying displacement, and uses the corrected parameters to update the predicted position and predicted passage time interval of subsequent virtual queue items.

[0058] For a target's actual passage event, the controller obtains the time residual between its preceding and following actual passage events. When the target's actual passage event has both a preceding and following actual passage event, and simultaneously meets the following conditions, it is determined that the egg corresponding to the target's actual passage event has experienced local slippage: The absolute value of the time residual of the actual event passing through the target is greater than the second threshold. The absolute value of the time residual between the previous actual passing event and the next actual passing event is not greater than the second threshold. The residual number of targets actually passing through the detection interval where the event occurs is zero.

[0059] When the target's actual passage event lacks a preceding or subsequent actual passage event, the controller will not determine the local slip offset based on that event until the adjacent actual passage event is obtained.

[0060] After determining that a local slippage offset has occurred, the controller keeps the time offset and pulse displacement conversion factor of the corresponding conveying segment unchanged, and only expands the predicted passage time interval of the virtual queue item corresponding to the actual passing event of the target. The expanded range is determined based on the target time residual, the response error of the detection node, and the fluctuation range of the conveying speed.

[0061] The controller determines the initial matching virtual queue item of the target's actual passage event, as well as the virtual queue item adjacent to the initial matching virtual queue item in the main conveyor line virtual queue, as a candidate virtual queue item set; then, based on the source branch, input order within the branch, predicted position, and predicted passage time interval of the candidate item, the controller performs subsequent matching on the target's actual passage event.

[0062] Through the above technical solution, in this embodiment, the local transport deviation of a single egg will not directly change the overall time offset or pulse displacement conversion coefficient of the corresponding transport segment, thereby avoiding the propagation of local deviation to other virtual queue items in the same transport segment.

[0063] It should be noted that the set number is determined based on the detection frequency of the detection node and the normal input interval of the egg, preferably 3 to 5 consecutive actual passing events; the first threshold is 1.2 times the sum of the encoder single pulse corresponding time and the conveying speed fluctuation time, or it can be determined by those skilled in the art through on-site calibration based on the actual operating parameters of the conveyor line; the second threshold is 2 to 3 times the first threshold, used to distinguish between the overall conveying deviation and the local slippage deviation of a single egg, and the specific value is determined by on-site calibration based on the speed fluctuation range of the conveyor line and the encoder resolution.

[0064] In one embodiment of the present invention, in step S2, when the correspondence between the actual passing event and the predicted passing event satisfies at least one of the following conditions, it is determined that the queue counting is mismatched: within a set statistical interval, the number of actual passing events is not equal to the number of predicted passing events; the actual passing time of an actual passing event falls within the predicted passing time interval of two or more virtual queue items simultaneously; the same virtual queue item does not have a corresponding actual passing event at two or more consecutive detection nodes; the order of actual passing events corresponding to each virtual queue item within the same source branch is inconsistent with the input order of each virtual queue item; After determining that the queue count mismatch is true, the assignment of source values ​​to each virtual queue item in the source confusion section is stopped, and all candidate correspondences are retained until the source attribute is re-determined through constraint matching in step S5.

[0065] Specifically, this implementation is used to determine whether a mismatch or a miscorrespondence has occurred between the actual passed events and the virtual queue items.

[0066] The controller uses detection nodes, transport path segments, or preset time intervals as analysis objects to obtain actual passing events and predicted passing events within the analysis objects. The actual passing events include the detection node identifier, actual passing time, and actual passing order; the predicted passing events include the corresponding virtual queue item, predicted passing time interval, and predicted passing order. The statistical interval is determined based on the transport path between adjacent detection nodes, the preset transport time, or the preset detection cycle.

[0067] Within the same statistical interval, the controller counts the number of actual passing events and the number of predicted passing events. When the two counts are not equal, it is determined that there is a queue count mismatch in the statistical interval, and the actual passing events or virtual queue items that cannot be matched are marked as undetermined corresponding items.

[0068] The controller compares the actual passage time of an actual passing event with the predicted passage time interval of a virtual queue item. When the actual passage time of an actual passing event falls within the predicted passage time interval of two or more virtual queue items, and the actual passing order, source branch input order, and transport path information still cannot rule out any of the virtual queue items, it is determined that there is a non-unique correspondence, and the actual passing event is associated with the multiple virtual queue items.

[0069] The controller checks the actual passing events of the same virtual queue item at consecutive detection nodes along the egg conveying direction. When the virtual queue item has reached the predicted passing time interval of the corresponding detection node, but no corresponding actual passing event is obtained at two or more consecutive detection nodes, it is determined that the virtual queue item has a cross-node missing correspondence.

[0070] If eggs are transported within the source branch in the order of input and there is no reverse movement or change in order during transport, the controller compares the input order of each virtual queue item within the same source branch with the order of its corresponding actual passing events. If the two are inconsistent, it is determined that there is a sequence mismatch in the queue segment corresponding to the source branch.

[0071] The controller determines that a queue count mismatch has occurred when at least one of the following conditions is met: quantity mismatch, non-unique correspondence, missing cross-node correspondence, or sequence mismatch within the source branch. The controller identifies the last virtual queue item that can establish a unique correspondence with the actual event and whose preceding consecutive queue items have all been uniquely corresponded as a reliable queue item; the transport segment corresponding to the virtual queue items that have not yet established a unique correspondence after the reliable queue item is identified as the source confusion segment.

[0072] For virtual queue items within the aforementioned source confusion segment, the controller retains their initial source attributes but does not yet establish a unique correspondence between them and actual passing events; simultaneously, it retains candidate correspondences between actual passing events and virtual queue items and marks the relevant actual passing events and virtual queue items as undetermined correspondences for subsequent source relationship processing. Through the above technical solution, this embodiment identifies virtual queue mismatch from multiple aspects such as quantity differences, non-unique correspondence, missing correspondence across detection nodes, and abnormal input order within the source branch. After determining the mismatch, it does not forcibly determine the unique source of the eggs, but retains the candidate correspondence between the actual passing events and virtual queue items. This avoids the source information being mispropagated due to missed detection, repeated detection, egg sticking, or abnormal delivery order, and provides a data foundation for subsequent rematching of source confusion sections and locating the source of abnormal eggs.

[0073] In one embodiment of the present invention, in steps S3 and S4, the continuous egg-free state is determined by the continuous egg-free time of the downstream detection node or the continuous idle encoder displacement of the main conveyor line; when the continuous idle encoder displacement is not less than the sum of the maximum occupied length of the egg in the conveying direction and twice the maximum allowable position error, it is determined that the blank conveying interval is formed, so as to isolate the source confusion section and subsequent eggs through the blank conveying interval.

[0074] Specifically, after the source branch line stops feeding eggs into the main conveyor line, the controller acquires the egg passage detection signal from the downstream detection node and the cumulative pulse signal from the main conveyor line encoder. The moment when the last egg passes through the downstream detection node is taken as the start time of the continuous egg-free state; after the start time, if the downstream detection node fails to detect any egg passage events for a continuous period, it is determined that a continuous egg-free state exists.

[0075] The controller determines the continuous no-load conveying displacement of the main conveyor line based on the cumulative encoder pulse change and pulse displacement conversion coefficient from the start time of the continuous egg-free state to the current time. The maximum length occupied by the egg in the conveying direction refers to the maximum conveying length that a single egg may occupy along the conveying direction; the maximum permissible position error refers to the maximum predicted position deviation determined by encoder resolution, conveying speed fluctuation, conveyor belt start / stop response error, and detection node response error.

[0076] When the continuous empty conveying displacement is not less than the sum of the maximum occupied length and twice the maximum permissible position error, a blank conveying interval is determined to be formed. The twice maximum permissible position error is used to cover the possible opposite positional deviations of eggs on both sides before and after the blank conveying interval. The controller uses the blank conveying interval as the synchronization area between the source confusion section and subsequent eggs, and determines the downstream boundary of the blank conveying interval as the post-synchronization boundary.

[0077] When the continuous unloaded conveying displacement is less than the sum of the maximum occupied length and twice the maximum allowable position error, the controller continues to acquire the egg passing detection signal and the encoder accumulated pulse signal until the determination condition is met.

[0078] Through the above technical solution, this embodiment uses the continuous no-egg signal from the downstream detection node and the continuous no-load encoder displacement of the main conveyor line to jointly determine the blank conveying interval, and requires that the no-load conveying displacement is not less than the sum of the maximum occupied length of the egg and twice the maximum allowable position error, which can avoid misjudging the ordinary interval between adjacent eggs as the synchronization boundary.

[0079] It should be noted that the maximum permissible position error is determined by the superposition of encoder resolution error, conveyor belt start / stop response error, detection node response time error, and the rolling or slippage error that the egg may generate on the conveying surface. Its typical value range is 5% to 10% of the maximum length occupied by a single egg, or it can be obtained through on-site calibration. The double maximum permissible position error is used to cover the opposite positional deviation that the eggs on the front and back sides of the blank conveying section may generate. The set time threshold is preferably 1.5 times the normal egg conveying interval. When the continuous egg-free time has met condition ① but the continuous empty encoder displacement has not yet reached the standard, the encoder displacement condition is used as the standard to continue waiting, so as to avoid misjudgment caused by the temporary loss of detection node signal. Conversely, if the encoder displacement has reached the standard but the continuous egg-free time has not been met due to the response delay of the detection node, the encoder displacement condition is used as the standard to avoid delaying synchronous recovery.

[0080] In one embodiment of the present invention, step S4, which involves restoring the input of eggs from the source branch line to the main conveyor line in a predetermined order, includes: After the post-synchronization boundary is formed, the target source branch to be restored is determined in a predetermined order, the target source branch is controlled to resume feeding eggs into the main conveyor line, and the remaining source branches are controlled to remain stopped feeding eggs into the main conveyor line. The first egg input detection event after the target source branch resumes input is obtained, and the first actual pass event obtained by the downstream detection node after the post-synchronization boundary is obtained. The first actual pass event is bound to the source attribute of the target source branch, and the bound virtual queue item is used as the starting item of the virtual queue after the post-synchronization boundary. After binding is completed, the next source branch is determined as the target source branch in the predetermined order and the above steps are repeated until all source branches resume feeding eggs into the main conveyor line.

[0081] Specifically, after the post-synchronization boundary is formed, the controller determines the current target source branch according to the pre-stored recovery order and controls the target source branch to input eggs into the main conveyor line; before all source branches complete the binding of the first egg, all other source branches except the current target source branch remain stopped inputting.

[0082] When the input detection node of the target source branch detects the first egg after recovery, the controller records the corresponding first egg input detection event and generates the corresponding predicted pass event based on the transport path and transport status between the source branch and the downstream detection node.

[0083] After the post-synchronization boundary, when the downstream detection node detects the first actual passing event, and the actual passing time of the first actual passing event is within the allowable delivery time range corresponding to the predicted passing event, the controller binds the first actual passing event to the source attribute of the current target source branch, and determines the corresponding virtual queue item as the starting item of the virtual queue after the post-synchronization boundary.

[0084] After completing the first egg binding of the current target source branch, the controller determines the next target source branch according to the recovery order, and repeats the above-mentioned input detection, actual pass event acquisition and source binding process until each source branch completes the source binding of the first egg; then, the controller establishes a virtual queue after the post-synchronization boundary according to the egg input detection events and transmission status signals of each source branch.

[0085] If the first actual pass event is not obtained within the corresponding allowed transmission time range after the target source branch resumes input, the source binding will not be established for the time being.

[0086] Through the above technical solution, this embodiment restores the source branches one by one in a predetermined order after the post-synchronization boundary is formed, and keeps other source branches from inputting, so that the first egg entering the main conveyor line in the recovery phase has a unique source, thereby re-establishing a clear source starting point for the subsequent virtual queue.

[0087] It should be noted that the predetermined order is the original input polling order of each source branch, or arranged in ascending order according to the source branch number, or arranged in ascending order according to the number of inputs of each source branch in the source confusion section so as to prioritize the recovery of branches with fewer inputs and lower matching difficulty; this order is preset in the controller during system initialization and can be modified by administrators through the controller interface according to the actual operation.

[0088] In one embodiment of the present invention, step S5, the constraint matching includes: Feasible matching results between actual passing events and virtual queue items within the source confusion segment are generated according to uniqueness constraints, intra-source order constraints, and delivery time constraints. Specifically, the uniqueness constraint ensures that each actual passing event matches at most one virtual queue item, and each virtual queue item matches at most one actual passing event; the intra-source order constraint ensures that the order of actual passing events matched by virtual queue items belonging to the same source branch is the same as the input order of the virtual queue items; and the delivery time constraint ensures that the actual passing time of an actual passing event falls within the allowed delivery time interval of its matched virtual queue item. When there is only one feasible matching result, the source attribute corresponding to the virtual queue item in the feasible matching result is output as the determined source; when there are two or more feasible matching results, the union of the source attributes corresponding to the same actual passing event in each feasible matching result is taken and output as the candidate source set of the actual passing event.

[0089] Specifically, the controller establishes candidate matching relationships between actual pass events and virtual queue items within the source obfuscation segment.

[0090] The allowed transport time interval is determined based on the predicted passage time interval of the virtual queue item at the corresponding detection node. For virtual queue items that are not determined to have experienced local slippage, the corresponding predicted passage time interval is determined as the allowed transport time interval; for virtual queue items that are determined to have experienced local slippage, the original predicted passage time interval is expanded based on the time residual corresponding to the local slippage, the detection node response error, and the transport speed fluctuation range, and the expanded predicted passage time interval is determined as the allowed transport time interval.

[0091] When the actual passing time of an event falls within the allowed delivery time interval corresponding to a virtual queue item, the candidate matching relationship between the actual passing event and the virtual queue item is retained; otherwise, the candidate matching relationship is deleted.

[0092] When the number of actual passing events within a source obfuscation segment is not equal to the number of virtual queue items, either the actual passing events or the virtual queue items are allowed to be in an unmatched state.

[0093] The controller combines the retained candidate matching relationships according to the passing order of actual passing events. Under the conditions of uniqueness constraint, intra-source order constraint, and delivery time constraint, it filters the matching relationship combinations between actual passing events and virtual queue items. Specifically, the uniqueness constraint limits each actual passing event to at most one virtual queue item, and each virtual queue item to at most one actual passing event; the intra-source order constraint limits the actual passing events matched by virtual queue items with earlier input order within the same source branch to no later than the actual passing events matched by virtual queue items with later input order; and the delivery time constraint limits the actual passing time of an actual passing event to fall within the allowed delivery time interval of its matched virtual queue item.

[0094] The controller counts the number of valid matching pairs in each matching relationship combination and takes the matching relationship combination with the largest number of valid matching pairs as the feasible matching result. When there are two or more matching relationship combinations with the same number of valid matching pairs, the controller calculates the sum of the absolute values ​​of the time residuals between the actual passing time of each actual passing event and the predicted passing time of the matched virtual queue item in each matching relationship combination, and takes the matching relationship combination with the smallest sum of the absolute values ​​of the time residuals as the feasible matching result. When there are still two or more matching relationship combinations that meet the above conditions, the multiple matching relationship combinations are retained as multiple feasible matching results. In specific implementation, the above process of screening and optimizing matching relationship combinations can be implemented by computer programs using conventional algorithms in the field, such as enumeration search or establishing a bipartite graph and applying matching algorithms.

[0095] When there is only one feasible match, the controller determines the source attribute carried by the virtual queue item in the feasible match as the source of the corresponding actual pass event.

[0096] When there are two or more feasible matching results, the controller obtains the source attribute of the virtual queue item corresponding to the same actual pass event in each feasible matching result, removes duplicate source attributes, and forms a candidate source set.

[0097] When no feasible matching result satisfies the uniqueness constraint, source order constraint, and delivery time constraint, the corresponding actual passing event and related virtual queue item are marked as having an undetermined source. The actual passing event is not forcibly assigned a source attribute, and the source attribute carried by the virtual queue item that has a candidate correspondence with the actual passing event is determined as a candidate source to be reviewed.

[0098] For example, when the same actual event corresponds to a virtual queue item of the first source branch and the second source branch in different feasible matching results, the output includes a candidate source set containing the first source branch and the second source branch; when multiple feasible matching results correspond to the same source attribute, the candidate source set only contains that source attribute.

[0099] Through the above technical solution, this embodiment simultaneously employs delivery time constraints, uniqueness constraints, and order constraints within the source branch to filter out candidate matching relationships that do not conform to the actual delivery process, and outputs a candidate source set when multiple feasible matching results exist, thereby reducing erroneous matching between actual passing events and virtual queue items.

[0100] It should be noted that if the number of source branches in the candidate source set is equal to the total number of source branches in the system, the actual pass event will be marked as "source undetermined" and will not be included in the abnormal statistics of any source branch, so as to avoid invalid information interfering with subsequent threshold judgment; if the number of source branches in the candidate source set is greater than 1 but less than the total number of source branches, the candidate source set will be retained and used for subsequent abnormal event association.

[0101] In one embodiment of the present invention, step S6, generating the corresponding aquaculture management instruction, includes: Count the number or proportion of abnormal eggs associated with the same source branch, cage column or cage layer within a preset time window; When the number of abnormal eggs or the proportion of abnormal eggs reaches a preset threshold, a separate operation instruction is generated. The separate operation instruction is used to control the source branch to input eggs into the main conveyor line separately in the next egg collection cycle, and to control the other source branches to stop inputting eggs into the main conveyor line in the egg collection cycle. Acquire egg product anomaly detection events during the individual operation period, and associate the egg product anomaly detection events with the source attribute corresponding to the source branch; When the number or proportion of abnormal eggs during the individual operation still reaches the preset threshold, an inspection instruction is generated for the breeding area corresponding to that source branch.

[0102] Specifically, for abnormal eggs whose unique source can be determined by constraint matching results, the controller includes the abnormal egg in the statistics of the number of abnormal eggs corresponding to the unique source and the proportion of abnormal eggs.

[0103] For candidate abnormal events, the controller counts the candidate abnormal events that have established a pending review association with the same candidate source in a weighted manner within a preset time window; where, when each candidate abnormal event is counted in the candidate abnormal event count of its respective candidate source, the weight is 1 / N, where N is the number of candidate sources corresponding to the candidate abnormal event; the weighting ratio of the candidate abnormal event is the ratio of the weighted number of candidate abnormal events of the candidate source to the total number of eggs involved in all candidate abnormal events corresponding to the candidate source within the same time window (calculated in the same weighting method); When the weighted number or weighted ratio of candidate abnormal events of a certain candidate source reaches the preset candidate review threshold, the candidate source is determined as the source to be reviewed, and a separate running instruction is generated for the corresponding source branch of the candidate source.

[0104] The separate operation command controls the source branch corresponding to the source to be reviewed to input eggs into the main conveyor line in the next egg collection cycle, and controls the other source branches to keep inputting eggs stopped.

[0105] The controller acquires egg anomaly detection events during standalone operation and associates them with the source attributes of the source to be verified. When the number or proportion of abnormal eggs during standalone operation still reaches the preset threshold, an inspection instruction is generated for the breeding area corresponding to the source to be verified. The inspection instruction includes the source branch identifier, cage row or cage layer identifier, the number or proportion of abnormal eggs, and the time of instruction generation.

[0106] Through the above technical solution, this embodiment counts the number and proportion of abnormal eggs in each source branch, cage row or cage layer, and when the abnormality reaches the threshold, the corresponding source branch is run separately. The detection results during the separate operation are used to verify the source of the abnormality, thereby reducing misjudgment caused by incorrect transport convergence or source matching.

[0107] It should be noted that the preset time window is preferably one complete egg collection cycle (i.e., the complete cycle from the start to the end of egg collection) or 2 to 4 hours continuously. The specific duration can be adjusted by on-site technicians according to the average daily egg production of the farm, the operating speed of the main conveyor line, and the actual sample size. If the window duration is too short, the sample size will be insufficient, resulting in statistically insignificant results. If it is too long, it will dilute short-term abnormal signals and cause a delayed response. Those skilled in the art can determine the duration within the range of 1 to 6 hours based on the egg production frequency.

[0108] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A method for automated and intelligent management of laying hen farming, characterized in that, The method includes: S1: Obtain egg input detection events from multiple source branches and the conveying status signals of the source branches and the main conveyor line, establish a virtual queue carrying source attributes, and generate predicted pass events for each virtual queue item at multiple detection nodes; S2: Based on the distribution of time residuals, quantity residuals, and sequence residuals between actual passing events and predicted passing events, identify overall shift, local slippage shift, or queue count mismatch of the transport section; S3: When queue counting mismatch is identified, determine the pre-synchronization boundary and source confusion section, control the source branch to stop inputting eggs into the main conveyor line, and keep the main conveyor line running until the downstream detection node detects a continuous state of no eggs; S4: Determine the blank conveying interval corresponding to the continuous egg-free state as the post-synchronization boundary, and restore the source branch line to input eggs into the main conveying line in a predetermined order; S5: Based on the number of source branch inputs between the pre-synchronization boundary and the post-synchronization boundary, the number of actual passing events, the input order, and the allowed transmission time interval, perform constraint matching on the actual passing events and virtual queue items in the source confusion section; S6: Associate the egg product anomaly detection event with the constraint matching result to determine the source information of the abnormal eggs and generate the corresponding breeding control instructions.

2. The automated intelligent breeding and management method for laying hens according to claim 1, characterized in that, In step S1, each virtual queue item includes at least the source branch identifier, the corresponding cage or cage layer identifier, the input detection time, the input sequence within the branch, the initial encoder reading, the predicted position, and the predicted passage time interval; the transport displacement is calculated based on the start / stop status in the transport status signal, the encoder cumulative pulse, and the pulse displacement conversion coefficient, and the predicted position and the predicted passage time interval are updated based on the transport displacement.

3. The automated intelligent breeding and management method for laying hens according to claim 1, characterized in that, In step S2, identifying the overall offset or local slippage of the conveyor section includes: When the time residuals of at least a set number of consecutive actual passing events at the same detection node have the same sign, the absolute value of the difference between each time residual and the average value of the time residuals of the set number of consecutive actual passing events is less than the first threshold and the quantity residual is zero, it is determined that the overall displacement of the conveying section is corrected according to the average value of the time residuals. When the absolute value of the time residual of the actual passing event of the target is greater than the second threshold, and the absolute values ​​of the time residuals of the preceding and following actual passing events that are immediately adjacent to the actual passing event of the target are not greater than the second threshold and the quantity residual is zero, it is determined to be a local slip offset. The time offset and pulse displacement conversion coefficient of the corresponding transport segment are kept unchanged, the predicted passing time interval of the virtual queue item corresponding to the actual passing event of the target is expanded, and the virtual queue item and its adjacent virtual queue items in the same source branch are used as the candidate virtual queue item set of the actual passing event of the target.

4. The automated intelligent breeding and management method for laying hens according to claim 1, characterized in that, In step S2, when the correspondence between the actual passing events and the predicted passing events satisfies at least one of the following conditions, it is determined that the queue count is mismatched: within a set statistical interval, the number of actual passing events is not equal to the number of predicted passing events; The actual passing time of an actual passing event falls within the predicted passing time interval of two or more virtual queue items; no corresponding actual passing event exists for the same virtual queue item at two or more consecutive detection nodes; the order of actual passing events corresponding to each virtual queue item within the same source branch is inconsistent with the input order of each virtual queue item; After determining that the queue count mismatch is true, the assignment of source values ​​to each virtual queue item in the source confusion section is stopped, and all candidate correspondences are retained until the source attribute is re-determined through constraint matching in step S5.

5. The automated intelligent breeding and management method for laying hens according to claim 1, characterized in that, In steps S3 and S4, the continuous egg-free state is determined by the continuous egg-free time of the downstream detection node or the continuous idle encoder displacement of the main conveyor line; when the continuous idle encoder displacement is not less than the sum of the maximum occupied length of the egg in the conveying direction and twice the maximum allowable position error, it is determined that the blank conveying interval is formed, so as to isolate the source confusion section and subsequent eggs through the blank conveying interval.

6. The automated intelligent breeding and management method for laying hens according to claim 1, characterized in that, In step S4, restoring the input of eggs from the source branch line to the main conveyor line in a predetermined order includes: After the post-synchronization boundary is formed, the target source branch to be restored is determined in a predetermined order, the target source branch is controlled to resume feeding eggs into the main conveyor line, and the remaining source branches are controlled to remain stopped feeding eggs into the main conveyor line. The first egg input detection event after the target source branch resumes input is obtained, and the first actual pass event obtained by the downstream detection node after the post-synchronization boundary is obtained. The first actual pass event is bound to the source attribute of the target source branch, and the bound virtual queue item is used as the starting item of the virtual queue after the post-synchronization boundary. After binding is completed, the next source branch is determined as the target source branch in the predetermined order and the above steps are repeated until all source branches resume feeding eggs into the main conveyor line.

7. The automated intelligent breeding and management method for laying hens according to claim 1, characterized in that, In step S5, the constraint matching includes: Feasible matching results between actual passing events and virtual queue items within the source confusion segment are generated according to uniqueness constraints, intra-source order constraints, and delivery time constraints. Specifically, the uniqueness constraint ensures that each actual passing event matches at most one virtual queue item, and each virtual queue item matches at most one actual passing event; the intra-source order constraint ensures that the order of actual passing events matched by virtual queue items belonging to the same source branch is the same as the input order of the virtual queue items; and the delivery time constraint ensures that the actual passing time of an actual passing event falls within the allowed delivery time interval of its matched virtual queue item. When there is only one feasible matching result, the source attribute corresponding to the virtual queue item in the feasible matching result is output as the determined source; when there are two or more feasible matching results, the union of the source attributes corresponding to the same actual passing event in each feasible matching result is taken and output as the candidate source set of the actual passing event.

8. The automated intelligent breeding and management method for laying hens according to claim 1, characterized in that, In step S6, generating the corresponding aquaculture control instructions includes: Count the number or proportion of abnormal eggs associated with the same source branch, cage column or cage layer within a preset time window; When the number of abnormal eggs or the proportion of abnormal eggs reaches a preset threshold, a separate operation instruction is generated. The separate operation instruction is used to control the source branch to input eggs into the main conveyor line separately in the next egg collection cycle, and to control the other source branches to stop inputting eggs into the main conveyor line in the egg collection cycle. Obtain egg product anomaly detection events during the individual operation period, and associate the egg product anomaly detection events with the source attribute corresponding to the source branch; When the number or proportion of abnormal eggs during the individual operation still reaches the preset threshold, an inspection instruction is generated for the breeding area corresponding to that source branch.