A method and system for adaptively turning over control of an incubator

Abstract

The present invention belongs to the field of data processing and automatic control technology, and proposes an adaptive flip control method and system for an incubator, which is specifically as follows: firstly, an egg tray is identified in the incubator, and each egg tray is marked as a flip position, and an air source sensor and a flip sensor are arranged at the flip position, and the relative flip amount and air source pressure are read in real time, and then a flip derivation model is constructed through the relative flip amount and air source pressure to obtain the flip failure risk, and finally the flip machine flow valve is adaptively controlled according to the flip failure risk. The adaptability of each flip position in the incubator rack and the current incubation flip action configuration of the incubator rack is effectively quantified, so as to reflect the difference accumulation problem of the actual flip effect, reduce the risk of the flip angle of the incubator egg being too large or too small, avoid the internal damage of the incubator egg, and greatly improve the incubation quality and energy utilization rate of the incubator.

Description

Technical Field

[0001] The invention belongs to the technical field of data processing and automatic control, and in particular relates to an adaptive turning control method and system for an incubator. Background Art

[0002] During the incubation process of hatching eggs, if the hatching eggs remain in a fixed position for a long time, the embryo will adhere to the inner membrane of the eggshell, resulting in abnormal embryonic development or death. Therefore, incubators are equipped with the technology of flipping the hatching eggs. The flipping technology periodically changes the position or angle of the eggs to avoid this adhesion phenomenon and the risk of abnormal embryonic development or death. In medium and large-scale automated hatching scenarios, the flipping function in the incubator mainly uses pneumatic flipping technology. Since its principle is to achieve precise flipping of the hatching eggs through air pressure drive, the flipping action achieved has the advantages of being smooth and highly accurate.

[0003] However, the operation mode of the flipping technology applied by pneumatic flipping technology belongs to collective flipping, that is, a large number of hatching eggs are flipped at the same time. During this flipping process, the same hatching flipping action configuration is performed on all hatching eggs, including air source pressure, air flow, pneumatic actuator stroke and pneumatic valve switching time, etc., and due to differences in weight and eggshell surface friction coefficient of hatching eggs, the actual flipping effects of different batches of hatching eggs during the flipping process will be different. The actual flipping effect usually refers to the degree of match between the actual flipping angle of the hatching eggs during the egg turning process and the preset required flipping angle. The differences in the actual flipping effects of these hatching eggs during the flipping process are easy to accumulate and induce the risk of egg flipping control errors. The risk of egg flipping control errors refers to the increased chance of the embryo adhering to the inside of the eggshell due to the flipping angle being too large or too small, and the higher risk of inducing internal damage to the hatching eggs when flipping again. The reason for the internal damage is that the different positions of the center of gravity of the hatching eggs make the tilt angle or rolling amplitude different when flipping, resulting in insufficient or excessive flipping angles for some hatching eggs. Insufficient flipping angles increase the risk of embryo adhesion, and when the flipping angle is too large, the internal structure may be damaged by shear force. Finally, the difference in flipping effects between different batches leads to a higher hatching failure rate in some batches. Therefore, an adaptive flipping control method and system for incubators is urgently needed to solve this problem. Summary of the invention

[0004] The purpose of the present invention is to provide an adaptive turning control method and system for an incubator to solve one or more technical problems existing in the prior art and at least provide a beneficial choice or create conditions.

[0005] In order to achieve the above object, according to one aspect of the present invention, a method for controlling an incubator to turn over is provided, the method comprising the following steps:

[0006] Identify egg trays in the incubator and mark each egg tray as the turning position;

[0007] Arrange the gas source sensor and the flip sensor at the flip position, and read the relative flip amount and gas source pressure in real time;

[0008] The rollover derivation model is constructed through the relative rollover amount and gas source pressure to obtain the rollover failure risk;

[0009] The flow valve of the overturning machine is adaptively controlled according to the risk of poor overturning.

[0010] Furthermore, the method for identifying an egg tray in an incubator and marking each egg tray as a turning position is as follows: the incubator includes a plurality of egg trays, each egg tray independently performs a turning operation on hatched eggs; and each egg tray on the same egg tray is marked as a turning position.

[0011] The purpose of setting the flip position separately here is that each egg tray needs to collect data separately. The actual meaning of the flip position not only represents the position of an egg tray on the incubation rack, but its essence is the analysis object in computer programming. Therefore, it usually also includes the number of the incubation rack, the type of incubated eggs, the air source pressure data collected by the egg tray, and the relative flipping amount data.

[0012] Furthermore, an air source sensor and a flip sensor are arranged at the flip position, and a method for reading the relative flip amount and the air source pressure in real time is as follows: whenever the flip operation starts, the relative flip amount and the air source pressure are read at equal intervals, and the reading interval is 0.05 seconds to 0.2 seconds; wherein the air source sensor includes any one of a piezoelectric air pressure sensor or a gas flow meter, and the air pressure value passing through the flip position is monitored in real time by the air source sensor, which is recorded as the air source pressure; the flip sensor includes any one of a rotary encoder, a tilt sensor or a laser sensor, and the flip angle of the flip position within the reading interval is monitored in real time by the flip sensor, the average value of the flip angles of all flip positions is recorded as the flip level, and the ratio of the flip angle of each flip position to the flip level is recorded as the relative flip amount.

[0013] The relative flipping amount and gas source pressure are both measured during the flipping operation. In the subsequent process of constructing the flipping derivation model to calculate the risk of flipping failure, the relative flipping amount and gas source pressure obtained from each flipping operation should be merged into a continuous analysis sequence in chronological order. That is, the analysis sequence composed of the relative flipping amount at any moment is not limited to the corresponding flipping operation at that moment, but should also be extended to the relative flipping amounts obtained from each historical flipping operation. The same applies to the gas source pressure data in the analysis.

[0014] A flipping operation refers to the process in which the incubator controls all the hatching eggs in the hatching rack to perform a complete flipping operation. After the complete flipping operation, the hatching eggs are flipped 45 degrees to 90 degrees, and the fault tolerance is 2.5 degrees to 5 degrees.

[0015] Furthermore, the flip sensor also includes a CCD camera, which collects images before and after the reading interval through the CCD camera, identifies the corresponding area of ​​each flip position through the edge and records it as the flip area, and uses the image registration algorithm to calculate the flip angle of the flip area before and after the reading interval to obtain the flip angle of the flip position.

[0016] Furthermore, a method for constructing a rollover derivation model and obtaining a rollover failure risk through relative rollover amount and gas source pressure is as follows: a binary group consisting of relative rollover amount and gas source pressure is recorded as a rollover temporary variable array;

[0017] Suppose a time period is used as the monitoring period TVCm, TVCm∈[0.25,2] minutes; within the monitoring period, the time scale of obtaining the flip temporary variable array is recorded as the monitoring point; the monitoring period only adopts the period of flip operation.

[0018] For any flip bit, the relative flip amount of the flip-critical variable array in each monitoring point is obtained, and a flip amount state set Riq is formed; the method of obtaining the step difference evaluation value tRLe using the flip amount state set is as follows:

[0019] tRLe=mean(CRL(Riq))×sigmoid(mean(CRL(Riq)) / max(CRL(Riq)));

[0020] Where CRL(Riq) is the set of differences between any element in the flip state set and its previous element, mean() and max() are functions for taking the average value and the maximum value respectively, and sigmoid() is the activation function; the previous element of any element refers to the element corresponding to the first monitoring point in the reverse time direction of the monitoring point corresponding to the element;

[0021] The principle of obtaining the step difference evaluation value here is actually to analyze the relative flipping amount of the flipping position at each monitoring point to form the flipping amount state set Riq, and calculate the step difference evaluation value based on the flipping amount state set Riq. In the acquisition process, differential operations and sigmoid function applications are involved, which can effectively capture the instantaneous rate of change of the relative flipping amount in the time series. At the same time, the sigmoid activation function is used to smooth the ratio of the mean and the maximum value to compress the data to the range of [0,1], suppress the sensitivity to excessive fluctuations, and reflect a more stable dynamic trend. The step difference evaluation value is a composite indicator that takes into account both the change amplitude of the relative flipping amount and the stability of the change. The purpose is to quantify the continuous stability of the hatching eggs during the flipping process and help identify potential risks of egg turning control errors.

[0022] Calculate the root mean square value of the step difference evaluation value corresponding to all flip positions, and record it as the flip reference level F_tRLe; select any flip position as the current flip position, use the K-means clustering algorithm on the flip temporary variable array of each monitoring point of the current flip position, and divide all monitoring points into several clusters.

[0023] The Euclidean distance between any monitoring point and the flipped temporary variable array of its corresponding cluster center is recorded as the center distance, and the monitoring point with the minimum center distance is recorded as the characteristic monitoring point Kta; the flipped temporary variable arrays of the characteristic monitoring points corresponding to each cluster constitute a set and are recorded as the characteristic point set lst{Kta};

[0024] Calculate the flip failure risk TFRS of the current flip bit based on the flip reference level and characteristic monitoring points:

[0025] ;

[0026] Where i1 is the serial number of the monitoring point in the monitoring period, TFRS(i1) and tRLe(i1) are the rollover risk and step difference evaluation values ​​of the i1-th monitoring point, respectively, PSD(i1,lst{Kta}) is the feature point distance function, and the return value obtained by the feature point distance function PSD(i1,lst{Kta}) is: the average value of the Euclidean distance between the flip-prone array of the i1-th monitoring point and each element in the feature point set; ln() is a logarithmic function with the natural constant e as the base. When the value of i1 is 1, TFRS(i1) represents the current rollover risk.

[0027] Since the calculation of the risk of bad flipping requires the flipping temporary variable array corresponding to the characteristic monitoring point to be processed and obtained, the risk of egg turning control errors accumulated and induced by the difference in actual flipping effects during the flipping process of the hatching eggs can be effectively quantified. However, the acquisition of the characteristic monitoring points is too sensitive to the direct data of the relative flipping amount and the air source pressure, which will lead to a strong sensitivity to the input noise, so that the K-means reference to the initial value and the cluster center deviates, resulting in distorted results or misjudged flipping adjustment risks, causing decision-making deviations, especially in the period when the characteristic monitoring points are not dense or unevenly distributed. This problem is more prominent. However, the existing technology cannot effectively compensate for this deviation. In order to eliminate this influence, the present invention proposes a more preferred solution as follows:

[0028] Preferably, the method of constructing the rollover derivation model and obtaining the rollover failure risk through the relative rollover amount and the gas source pressure is: recording the binary group consisting of the relative rollover amount and the gas source pressure as a rollover temporary variable array;

[0029] Suppose a time period is the monitoring period TVCm, the value range is TVCm∈[0.5,5] minutes, and the time scale of obtaining the flip temporary variable array is recorded as the monitoring point;

[0030] During the monitoring period, any flip position is selected as the current flip position; for the current flip position, the gas source pressures of each monitoring point constitute a pressure monitoring sequence, and the pressure monitoring sequence curve is fitted using the Gaussian process regression method to obtain a gas source fitting line;

[0031] The gas source fitting line here is actually a smooth curve obtained by fitting. The purpose of using Gaussian process regression is to construct a smooth curve that is close to the real data trend based on discrete data points. On the one hand, it will predict the location for real data collection, and on the other hand, it will correct the measured data in the direction of the trend. Because the measurement of gas source pressure often has abnormal values, including sudden pressure peaks or sudden drops in pressure values, the characteristic that the curve passes near the data point rather than strictly passing through it is conducive to processing noisy data and reducing time series complexity to achieve confidence interval recognition effect.

[0032] The fitting value of each monitoring point on the gas source fitting line is used as the pressure correction value of the corresponding monitoring point; the absolute value of the difference between the corresponding pressure correction value of each monitoring point and the gas source pressure is calculated and recorded as the pressure disturbance amount Psstc;

[0033] The principle of obtaining the pressure disturbance here is actually to model the pressure monitoring sequence through Gaussian process regression and obtain the gas source fitting line, and then use the pressure disturbance to quantify the difference between the actual gas source pressure and the fitting value of each monitoring point. The pressure disturbance directly reflects the fluctuation of the gas source pressure during the flipping process, and is a manifestation of the gas source pressure fluctuation and control error risk in the flipping process. If the pressure disturbance is large, it means that the actual gas source pressure at this point is far away from the fitting curve, indicating that the gas source pressure at this point changes more drastically, which means that the gas source pressure is unstable during the flipping process. If the pressure disturbance is small, it means that the fitting curve fits the actual data well, the gas source pressure changes relatively smoothly, and the flipping process is relatively accurate and stable.

[0034] If the pressure disturbance at any monitoring point is greater than the upper quartile of all pressure disturbances, the monitoring point is recorded as a high-pressure disturbance point; the flipping temporary variable array and the pressure disturbance corresponding to a monitoring point are recorded to form a flipping control vector, that is, the flipping control vector is composed of three elements: the relative flipping amount, the gas source pressure and the pressure disturbance at the same monitoring point; the Manhattan distance between the flipping control vector of any monitoring point and any other monitoring point is defined as the control trend difference;

[0035] Any monitoring point is recorded as the current monitoring point, and the monitoring point in the reverse time direction of the current monitoring point is recorded as the search monitoring point;

[0036] The first reference point selection condition is set as follows: the control trend difference of the search monitoring point has a minimum value and is discontinuous with the current monitoring point; the second reference point selection condition is set as follows: there is at least one high pressure disturbance point between the current monitoring point and the search monitoring point; the monitoring point that is discontinuous with the current monitoring point means that the reference point cannot select the first search monitoring point in the reverse time direction with the current monitoring point;

[0037] Searching for a search monitoring point that satisfies both the first reference point selection condition and the second reference point selection condition from the current monitoring point in the reverse time direction, recording it as the reference point, and recording the period between the current monitoring point and the reference point as the fluctuation analysis domain;

[0038] According to the fluctuation analysis domain and the pressure high disturbance point, the flip failure risk TFRS of the current flip position is calculated. TFRS = mean{e -j1 grpu(j1)·log(Rlao(j1)·intSqrt(Psstc(j1)) -1}; where mean{} is the mean value function, j1 is the serial number of the monitoring point in the monitoring period TVCm, TFRS(j1) represents the bad rollover risk of the j1-th monitoring point, grpu(j1) is the relative rollover trend function, which is specifically defined as: the ratio of the number of monitoring points in the fluctuation analysis domain corresponding to the j1-th monitoring point whose relative rollover amount is greater than or equal to the j1-th monitoring point to the number of monitoring points whose relative rollover amount is less than the j1-th monitoring point, and the obtained ratio is used as the return value of the relative rollover trend function; Psstc(j1) is the pressure disturbance amount of the j1-th monitoring point, Rlao(j1) is the number of all high-pressure disturbance points in the fluctuation analysis domain corresponding to the j1-th monitoring point, intSqrt() is the square root rounding function, and the square root of the calling element is rounded up to obtain the return value by the square root rounding function, and lg() is a logarithmic function with base 10;

[0039] Beneficial effects: Since the risk of poor turning is calculated by the relative turning amount and the air source pressure, it can effectively quantify the adaptability of each turning position in the incubator and the current incubation turning action configuration of the incubator, thereby reflecting the difference accumulation problem of the actual turning effect, and evaluating the degree of risk of egg turning control error in the current turning position in real time and dynamically, thereby providing a mathematical interface for subsequent adaptive control and adjusting the flow valve of the turning machine in the turning position, thereby reducing the risk of the incubation egg turning angle being too large or too small, avoiding the internal damage of the incubation egg, and greatly improving the incubation quality and energy utilization of the incubator.

[0040] Furthermore, the method for adaptively controlling the flow valve of the flipping machine according to the flipping failure risk is: at any time, the flipping failure risk writing sequence of each incubation position is collected and recorded as a risk monitoring sequence, and a box plot method is used to detect whether an abnormal value appears in the risk monitoring sequence. If an abnormal value exists, the moment is marked as a bad point, and each flipping position where the failure occurs is recorded as a bad flipping position;

[0041] In the current flipping operation, the average value of the risk monitoring sequence corresponding to each non-defective point is the flipping reference value, and the average value of each flipping defect risk of any defective flipping position is the flipping performance value; if the flipping performance value of the defective flipping position is greater than the flipping reference value, the flow valve corresponding to the flipping position will be reduced by 0.1%-5%; if the flipping performance value of the defective flipping position is less than the flipping reference value, the flow valve corresponding to the flipping position will be increased by 0.1%-5%.

[0042] The flow valve is controlled based on the flipping operation as a cycle, and is adjusted after each flipping operation is completed, rather than during the flipping operation.

[0043] Preferably, all undefined variables in the present invention, if not clearly defined, can be manually set thresholds.

[0044] The present invention also provides an incubator adaptive flip control system, which comprises: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps in the incubator adaptive flip control method are implemented. The incubator adaptive flip control system can be run in computing devices such as desktop computers, notebook computers, PDAs, and cloud data centers. The executable system may include, but is not limited to, a processor, a memory, and a server cluster. The processor executes the computer program and runs it in the following system units:

[0045] A turning position identification unit is used to identify the egg tray in the incubator and mark each egg tray as a turning position;

[0046] A real-time monitoring unit is used to arrange the gas source sensor and the flip sensor at the flip position, and read the relative flip amount and gas source pressure in real time;

[0047] A derivation model building unit, used to build a rollover derivation model through relative rollover amount and gas source pressure and obtain the rollover failure risk;

[0048] The flow valve dynamic control unit is used to adaptively control the flow valve of the turnover machine according to the risk of improper turnover.

[0049] The beneficial effects of the present invention are as follows: the present invention provides an adaptive turning control method and system for an incubator, which calculates the risk of poor turning by the relative turning amount and the air source pressure, and can effectively quantify the adaptability of each turning position in the incubator rack to the current incubation turning action configuration of the incubator rack, thereby reflecting the difference accumulation problem of the actual turning effect, and dynamically evaluating the degree of risk of egg turning control error at the current turning position in real time, thereby reducing the risk of the incubation egg turning angle being too large or too small. From the bottom logic, it can prevent the risk of embryo adhesion being increased due to insufficient turning angle, or the risk of the internal structure being damaged due to shear force when the turning angle is too large, thereby avoiding the internal damage of the incubation egg, and greatly improving the incubation quality and energy utilization rate of the incubator. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The above and other features of the present invention will become more obvious by describing in detail the embodiments shown in the accompanying drawings. The same reference numerals in the accompanying drawings of the present invention represent the same or similar elements. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other accompanying drawings can be obtained based on these accompanying drawings without creative work. In the accompanying drawings:

[0051] Figure 1 Shown is a flow chart of an adaptive flipping control method for an incubator;

[0052] Figure 2 Shown is a structural diagram of an adaptive flipping control system for an incubator. DETAILED DESCRIPTION

[0053] The following will be combined with the embodiments and drawings to clearly and completely describe the concept, specific structure and technical effects of the present invention, so as to fully understand the purpose, scheme and effect of the present invention. It should be noted that the embodiments and features in the embodiments of this application can be combined with each other without conflict.

[0054] like Figure 1 The flowchart of a method for controlling an incubator's self-adaptive flipping is shown below. Figure 1 To illustrate an adaptive flipping control method for an incubator according to an embodiment of the present invention, the method comprises the following steps:

[0055] Example 1

[0056] Identify egg trays in the incubator and mark each egg tray as the turning position;

[0057] Arrange the gas source sensor and the flip sensor at the flip position, and read the relative flip amount and gas source pressure in real time;

[0058] The rollover derivation model is constructed through the relative rollover amount and gas source pressure to obtain the rollover failure risk;

[0059] The flow valve of the overturning machine is adaptively controlled according to the risk of poor overturning.

[0060] Furthermore, the method for identifying an egg tray in an incubator and marking each egg tray as a turning position is as follows: the incubator includes a plurality of egg trays, each egg tray independently performs a turning operation on hatched eggs; and each egg tray on the same egg tray is marked as a turning position.

[0061] The purpose of setting the flip position separately here is that each egg tray needs to collect data separately. The actual meaning of the flip position not only represents the position of an egg tray on the incubation rack, but its essence is the analysis object in computer programming. Therefore, it usually also includes the number of the incubation rack, the type of incubated eggs, the air source pressure data collected by the egg tray, and the relative flipping amount data.

[0062] Furthermore, an air source sensor and a flip sensor are arranged at the flip position, and the method for reading the relative flip amount and the air source pressure in real time is: whenever the flip operation starts, the relative flip amount and the air source pressure are read at equal intervals, and the reading interval is 0.1 second; wherein the air source sensor is a piezoelectric air pressure sensor, and the air pressure value passing through the flip position is monitored in real time by the air source sensor, which is recorded as the air source pressure; the flip sensor is a tilt sensor, and the flip angle of the flip position within the reading interval is monitored in real time by the flip sensor, the average value of the flip angles of all flip positions is recorded as the flip level, and the ratio of the flip angle of each flip position to the flip level is recorded as the relative flip amount.

[0063] A flipping operation refers to the process in which the incubator controls all the hatching eggs in the hatching rack to perform a complete flipping operation. After the complete flipping operation, the hatching eggs are flipped 45 degrees, and the fault tolerance is 2.5 degrees.

[0064] Furthermore, the flip sensor also includes a CCD camera, which collects images before and after the reading interval through the CCD camera, identifies the corresponding area of ​​each flip position through the edge and records it as the flip area, and uses the image registration algorithm to calculate the flip angle of the flip area before and after the reading interval to obtain the flip angle of the flip position.

[0065] Furthermore, a method for constructing a rollover derivation model and obtaining a rollover failure risk through relative rollover amount and gas source pressure is as follows: a binary group consisting of relative rollover amount and gas source pressure is recorded as a rollover temporary variable array;

[0066] Suppose a time period is used as the monitoring period TVCm, and the value is 1 minute; within the monitoring period, the time scale of obtaining the flip temporary variable array is recorded as the monitoring point;

[0067] For any flipping bit, obtain the relative flipping amount of the flipping array in each monitoring point and form a flipping amount state set Riq;

[0068] The method of obtaining the step difference evaluation value tRLe using the flip state set is as follows:

[0069] tRLe=mean(CRL(Riq))×sigmoid(mean(CRL(Riq)) / max(CRL(Riq)));

[0070] Where CRL(Riq) is the set of differences between any element in the flip state set and its previous element, mean() and max() are functions for taking the average value and the maximum value respectively, and sigmoid() is the activation function; the previous element of any element refers to the element corresponding to the first monitoring point in the reverse time direction of the monitoring point corresponding to the element;

[0071] Calculate the root mean square value of the step difference evaluation value corresponding to all flip positions, and record it as the flip reference level F_tRLe; select any flip position as the current flip position, use the K-means clustering algorithm on the flip temporary variable array of each monitoring point of the current flip position, and divide all monitoring points into several clusters.

[0072] The Euclidean distance between any monitoring point and the flipped temporary variable array of its corresponding cluster center is recorded as the center distance, and the monitoring point with the minimum center distance is recorded as the characteristic monitoring point Kta; the flipped temporary variable arrays of the characteristic monitoring points corresponding to each cluster constitute a set and are recorded as the characteristic point set lst{Kta};

[0073] Calculate the flip failure risk TFRS of the current flip bit based on the flip reference level and characteristic monitoring points:

[0074] ;

[0075] Where i1 is the serial number of the monitoring point in the monitoring period, TFRS(i1) and tRLe(i1) are the flipping risk and step error assessment values ​​of the i1-th monitoring point respectively, PSD(i1,lst{Kta}) is the feature point distance function, and the return value obtained by the feature point distance function PSD(i1,lst{Kta}) is: the average value of the Euclidean distance between the flipping temporary variable array of the i1-th monitoring point and each element in the feature point set; ln() is a logarithmic function with the natural constant e as the base.

[0076] Furthermore, the method for adaptively controlling the flow valve of the flipping machine according to the flipping failure risk is: at any time, the flipping failure risk writing sequence of each incubation position is collected and recorded as a risk monitoring sequence, and a box plot method is used to detect whether an abnormal value appears in the risk monitoring sequence. If an abnormal value exists, the moment is marked as a bad point, and each flipping position where the failure occurs is recorded as a bad flipping position;

[0077] In the current flipping operation, the average value of the risk monitoring sequence corresponding to each non-defective point is the flipping reference value, and the average value of each flipping defect risk of any defective flipping position is the flipping performance value; if the flipping performance value of the defective flipping position is greater than the flipping reference value, the flow valve corresponding to the flipping position will be reduced by 1%, and if the flipping performance value of the defective flipping position is less than the flipping reference value, the flow valve corresponding to the flipping position will be increased by 1%.

[0078] Example 2

[0079] The method adjustment of Example 2 is the same as that of Example 1, and the difference is that the method of constructing the flip derivation model and obtaining the flip failure risk through the relative flip amount and the gas source pressure is: the binary group consisting of the relative flip amount and the gas source pressure is recorded as the flip temporary variable array;

[0080] Assume a time period as the monitoring period TVCm, the value is 1 minute, and record the time scale of obtaining the flip temporary variable array as the monitoring point;

[0081] During the monitoring period, any flip position is selected as the current flip position; for the current flip position, the gas source pressures of each monitoring point constitute a pressure monitoring sequence, and the pressure monitoring sequence curve is fitted using the Gaussian process regression method to obtain a gas source fitting line;

[0082] The fitting value of each monitoring point on the gas source fitting line is used as the pressure correction value of the corresponding monitoring point; the absolute value of the difference between the corresponding pressure correction value of each monitoring point and the gas source pressure is calculated and recorded as the pressure disturbance amount Psstc;

[0083] If the pressure disturbance at any monitoring point is greater than the upper quartile of all pressure disturbances, the monitoring point is recorded as a high-pressure disturbance point; the flipping temporary variable array and the pressure disturbance corresponding to a monitoring point constitute a flipping control vector, and the Manhattan distance between the flipping control vector of any monitoring point and any other monitoring point is defined as the control trend difference;

[0084] Any monitoring point is recorded as the current monitoring point, and the monitoring point in the reverse time direction of the current monitoring point is recorded as the search monitoring point;

[0085] The first reference point selection condition is set as follows: the control trend difference of the search monitoring point has a minimum value and is discontinuous with the current monitoring point; the second reference point selection condition is set as follows: there is at least one high pressure disturbance point between the current monitoring point and the search monitoring point; the monitoring point that is discontinuous with the current monitoring point means that the reference point cannot select the first search monitoring point in the reverse time direction with the current monitoring point;

[0086] Searching for a search monitoring point that satisfies both the first reference point selection condition and the second reference point selection condition from the current monitoring point in the reverse time direction, recording it as the reference point, and recording the period between the current monitoring point and the reference point as the fluctuation analysis domain;

[0087] According to the fluctuation analysis domain and the pressure high disturbance point, the flip failure risk TFRS of the current flip position is calculated. TFRS = mean{e -j1 grpu(j1)·lg(Rlao(j1)·intSqrt(Psstc(j1)) -1}; where mean{} is the mean value function, j1 is the serial number of the monitoring point in the monitoring period TVCm, TFRS(j1) represents the bad rollover risk of the j1-th monitoring point, grpu(j1) is the relative rollover trend function, which is specifically defined as: the ratio of the number of monitoring points in the fluctuation analysis domain corresponding to the j1-th monitoring point whose relative rollover amount is greater than or equal to the j1-th monitoring point to the number of monitoring points whose relative rollover amount is less than the j1-th monitoring point, and the obtained ratio is used as the return value of the relative rollover trend function; Psstc(j1) is the pressure disturbance amount of the j1-th monitoring point, Rlao(j1) is the number of all high-pressure disturbance points in the fluctuation analysis domain corresponding to the j1-th monitoring point, intSqrt() is the square root rounding function, and the square root of the calling element is rounded up by the square root rounding function to obtain the return value;

[0088] An embodiment of the present invention provides an adaptive incubator flip control system, such as Figure 2 The figure shows a structural diagram of an adaptive flipping control system for an incubator of the present invention. An adaptive flipping control system for an incubator of this embodiment comprises: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps in the above-mentioned adaptive flipping control method for an incubator are implemented.

[0089] The system comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to run in the following units of the system:

[0090] A turning position identification unit is used to identify the egg tray in the incubator and mark each egg tray as a turning position;

[0091] A real-time monitoring unit is used to arrange the gas source sensor and the flip sensor at the flip position, and read the relative flip amount and gas source pressure in real time;

[0092] A derivation model building unit, used to build a rollover derivation model through relative rollover amount and gas source pressure and obtain the rollover failure risk;

[0093] The flow valve dynamic control unit is used to adaptively control the flow valve of the turnover machine according to the risk of improper turnover.

[0094] The adaptive flip control system for an incubator can be run on computing devices such as desktop computers, laptop computers, PDAs, and cloud servers. The adaptive flip control system for an incubator can include, but is not limited to, a processor and a memory. Those skilled in the art will appreciate that the example is merely an example of an adaptive flip control system for an incubator, and does not constitute a limitation on an adaptive flip control system for an incubator. The example may include more or fewer components than the example, or a combination of certain components, or different components. For example, the adaptive flip control system for an incubator may also include input and output devices, network access devices, buses, and the like.

[0095] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the operating system of the incubator adaptive flip control system, and various interfaces and lines are used to connect the various parts of the operating system of the incubator adaptive flip control system.

[0096] The memory can be used to store the computer program and / or module, and the processor realizes various functions of the incubator adaptive flip control system by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0097] Although the description of the present invention has been quite detailed and has been described in particular with respect to several described embodiments, it is not intended to be limited to any of these details or embodiments or any particular embodiment, so as to effectively cover the intended scope of the present invention. In addition, the present invention is described above with the embodiments foreseeable by the inventors, and its purpose is to provide a useful description, and those non-substantial changes to the present invention that are not currently foreseen may still represent equivalent changes of the present invention.

Claims

1. An adaptive turning control method for an incubator, characterized in that: The method comprises the following steps: Identify the egg tray in the incubator and mark each egg tray as a flip position; arrange the air source sensor and flip sensor at the flip position, and read the relative flip amount and air source pressure in real time; construct a flip derivation model based on the relative flip amount and air source pressure and obtain the flip failure risk; adaptively control the flip machine flow valve according to the flip failure risk; The method of constructing a rollover derivation model and obtaining the rollover failure risk through relative rollover amount and gas source pressure is as follows: recording the binary group consisting of relative rollover amount and gas source pressure as a rollover temporary variable array; Assume a time period as the monitoring period TVCm, TVCm∈[0.25,2] minutes; within the monitoring period, the time scale of obtaining the flip-imminent variable array is recorded as the monitoring point; for any flip bit, the relative flip amount of the flip-imminent variable array in each monitoring point is obtained, and a flip amount state set Riq is formed; the method of obtaining the step difference evaluation value tRLe using the flip amount state set is as follows: tRLe=mean(CRL(Riq))×sigmoid(mean(CRL(Riq)) / max(CRL(Riq))); Where CRL(Riq) is the set of differences between any element in the flip state set and its previous element; Calculate the root mean square value of the step difference evaluation value corresponding to all flip positions, and record it as the flip reference level F_tRLe; select any flip position as the current flip position, use the K-means clustering algorithm on the flip temporary variable array of each monitoring point of the current flip position, and divide all monitoring points into several clusters; The Euclidean distance between any monitoring point and the flipped temporary variable array of its corresponding cluster center is recorded as the center distance, and the monitoring point with the minimum center distance is recorded as the characteristic monitoring point Kta; the flipped temporary variable arrays of the characteristic monitoring points corresponding to each cluster constitute a set and are recorded as the characteristic point set lst{Kta}; Calculate the flip failure risk TFRS of the current flip bit based on the flip reference level and characteristic monitoring points: ； Where i1 is the serial number of the monitoring point in the monitoring period, TFRS(i1) and tRLe(i1) are the rollover risk and step error assessment values ​​of the i1-th monitoring point respectively, PSD(i1,lst{Kta}) is the feature point distance function, and the return value obtained by the feature point distance function PSD(i1,lst{Kta}) is: the average value of the Euclidean distance between the flipping array of the i1-th monitoring point and each element in the feature point set; ln() is a logarithmic function with the natural constant e as the base. When the value of i1 is 1, TFRS(i1) represents the current rollover risk.

2. The method for controlling an incubator's self-adaptive turning according to claim 1, characterized in that: The method for identifying an egg tray in an incubator and marking each egg tray as a turning position is as follows: the incubator includes a plurality of egg trays, each egg tray independently performs a hatching egg turning operation; and each egg tray on the same egg tray is marked as a turning position.

3. The method for controlling an incubator's self-adaptive turning according to claim 1, characterized in that: The method of arranging the air source sensor and the flip sensor at the flip position and reading the relative flip amount and the air source pressure in real time is: whenever the flip operation starts, the relative flip amount and the air source pressure are read at equal intervals, and the reading interval is 0.05 seconds to 0.2 seconds; wherein the air source sensor includes any one of a piezoelectric air pressure sensor or a gas flow meter, and the air pressure value passing through the flip position is monitored in real time by the air source sensor, which is recorded as the air source pressure; the flip sensor includes any one of a rotary encoder, a tilt sensor or a laser sensor, and the flip angle of the flip position in the reading interval is monitored in real time by the flip sensor, the average value of the flip angles of all flip positions is recorded as the flip level, and the ratio of the flip angle of each flip position to the flip level is recorded as the relative flip amount.

4. The method for controlling an incubator's self-adaptive turning according to claim 3, characterized in that: The flip sensor also includes a CCD camera, which collects images before and after the reading interval. The corresponding area of ​​each flip position is identified by edge recognition and recorded as the flip area. The image registration algorithm is used to calculate the flip angle of the flip area before and after the reading interval to obtain the flip angle of the flip position.

5. The method for controlling an incubator's adaptive turning according to claim 1, characterized in that: The method for adaptively controlling the flow valve of the flipping machine according to the flipping failure risk is as follows: at any time, the flipping failure risk writing sequence of each hatching position is collected and recorded as the risk monitoring sequence, and the box plot method is used to detect whether an abnormal value appears in the risk monitoring sequence. If an abnormal value exists, the moment is marked as a bad point, and each flipping position where the failure occurs is recorded as a bad flipping position; In the current flipping operation, the average value of the risk monitoring sequence corresponding to each non-defective point is the flipping reference value, and the average value of each flipping defect risk of any defective flipping position is the flipping performance value; if the flipping performance value of the defective flipping position is greater than the flipping reference value, the flow valve corresponding to the flipping position will be reduced by 0.1%-5%; if the flipping performance value of the defective flipping position is less than the flipping reference value, the flow valve corresponding to the flipping position will be increased by 0.1%-5%.

6. An adaptive incubator flip control system, characterized in that: The adaptive flipping control system for an incubator comprises: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the adaptive flipping control method for an incubator described in any one of claims 1 to 5 are implemented. The adaptive flipping control system for an incubator runs on a desktop computer, a laptop computer, a PDA, and a computing device in a cloud data center.

Images