A method, apparatus, medium, and program product for control decision making for a mouse-cage environment

Abstract

A method, device, medium and program product for controlling decisions in a mouse-cage environment, relating to the technical field of experimental animal husbandry control. The device transforms real-time carbon dioxide concentration flow into a time-stamped sequence, immediately smoothes out transient spikes with a sliding average, then locks in a tolerable fluctuation bandwidth with a standard deviation, forming a safe steady-state zone that automatically expands and contracts with data drift; when the concentration maximum breaks through the conflict threshold and the circadian rhythm feature is determined to be lost, the zone is activated. Subsequently, the device accumulates the duration of time outside the interval as a recovery duration and obtains a steady-state resilience index. When the index is still within an acceptable range and the rate of change is positive, the device extrapolates the time point, completes the intervention before the index reaches the upper limit by starting the hierarchical strategy in advance, and realizes early steady state.

Description

Technical Field

[0001] This application relates to the field of laboratory animal husbandry and control technology, and in particular to a method, device, medium, and program product for controlling and making decisions about the cage-rearing environment of rats. Background Technology

[0002] With the development of biotechnology, laboratory animals have become indispensable tools in life science research. Among them, mice, due to their short reproductive cycle and clear genetic background, have become one of the most widely used laboratory animals. Therefore, providing mice with a stable and clean breeding environment and effectively controlling that environment are key prerequisites for ensuring the accuracy and reproducibility of experimental results.

[0003] In related technologies, to meet the needs of raising and observing mice, existing mouse rearing systems are typically equipped with a main control unit and a ventilation system. This main control unit uses a preset, programmed control strategy to regulate the environment. A common strategy is based on the mice's circadian rhythm: increasing ventilation at night when the mice are active to remove more metabolic waste, and decreasing ventilation during the day when the mice are resting to conserve energy.

[0004] However, while methods based on fixed time programs or simple feedback mechanisms are generally effective in handling routine husbandry scenarios, unpredictable bursts of carbon dioxide concentration occur during specific experimental procedures involving the co-breeding of unfamiliar male mice to establish a new social hierarchy. These bursts are caused by random, brief fights and chases among the mice. Simultaneously, changes in mouse behavior patterns can disrupt their circadian rhythms, leading to lags in related technologies and difficulty in timely removal of excess carbon dioxide. This further contributes to the continuous deterioration of environmental quality, necessitating artificial adjustments based on mouse activity and metabolic levels, thus reducing the automation level of husbandry management. Summary of the Invention

[0005] This application provides a method, device, medium, and program product for controlling the cage-rearing environment of mice. It is used to automatically identify abnormal behavior patterns and states, select appropriate predictive intervention control methods for environmental control, and improve the automation level and timeliness of environmental regulation in laboratory mouse husbandry and management.

[0006] In a first aspect, this application provides a control decision-making method for a rat cage rearing environment. The method includes: generating concentration time-series data based on acquired carbon dioxide concentration data within the rat cage; determining a safe steady-state zone based on the moving average and standard deviation of the concentration time-series data when the maximum value in the concentration time-series data exceeds a preset conflict threshold and the diurnal rhythm characteristics disappear; the disappearance of the diurnal rhythm characteristics is defined as the discrete value of the concentration time-series data and a preset comparison concentration time-series data within the same time period exceeding a preset threshold; determining the time period during which the concentration time-series data exceeds the safe steady-state zone as a recovery duration within a preset first time period; using the average of a preset value and the recovery duration as a steady-state resilience index; determining a disturbance time point where the steady-state resilience index exceeds the preset acceptable range based on the steady-state resilience index and the rate of change within a preset second time period when the steady-state resilience index is determined to be within a preset acceptable range at the disturbance time point; and switching the control strategy based on the steady-state resilience index at the disturbance time point so that the steady-state resilience index falls within the preset acceptable range after the disturbance time point.

[0007] By adopting the above technical solution, the equipment generates time-series data based on the acquired carbon dioxide concentration data, and determines the safe steady-state range through the moving average and standard deviation of the data, establishing a benchmark judgment mechanism for environmental fluctuations. When environmental fluctuations occur, the time period exceeding the safe steady-state range is defined as the recovery time, and the average of a preset number of recovery times is used as the steady-state resilience index, achieving a quantitative assessment of the environmental fluctuation recovery capability. Based on the value and trend of the steady-state resilience index, the equipment can predict potential disturbance time points before environmental parameters reach dangerous levels and switch control strategies in advance. This predictive control mechanism based on the dynamic changes of environmental parameters enables the equipment to respond early to environmental fluctuations caused by behaviors such as fighting among mice, initiating graded strategies in advance. This avoids the problem of passive adjustment after environmental parameters exceed the standard in traditional solutions, achieving proactive pre-control of environmental parameters and improving the timeliness and accuracy of environmental regulation.

[0008] In conjunction with some embodiments of the first aspect, in some embodiments, the step of switching the control strategy based on the steady-state resilience index at the disturbance time point to bring the steady-state resilience index within the preset acceptable range specifically includes: at the disturbance time point, determining a target control level from a preset hierarchical strategy library containing multiple intervention levels based on the steady-state resilience index and the rate of change; calculating a strategy transition duration based on the difference between the control level of the current control strategy and the target control level, wherein the transition duration is determined based on the larger the level span, the longer the transition duration; adjusting the rate of strategy switching based on the instantaneous change gradient of the real-time monitored carbon dioxide concentration within the strategy transition duration; and determining that the control strategy switching is complete when the steady-state resilience index is within the preset acceptable range within a third time period.

[0009] By adopting the above technical solution, the equipment determines the target control level based on a preset hierarchical strategy library, the steady-state resilience index, and its rate of change, enabling the control measures to have a hierarchical adjustment capability. After determining the target control level, the equipment dynamically calculates the strategy transition duration based on the difference between the current control level and the target control level. During the strategy transition, the equipment adjusts the strategy switching rate based on the real-time monitored gradient of the instantaneous change in carbon dioxide concentration, and determines the transition completion by observing whether the steady-state resilience index returns to the preset acceptable range. This control mechanism based on hierarchical strategies and smooth transition avoids the drastic environmental fluctuations that may be caused by abrupt changes in control intensity, while ensuring the effectiveness of control measures and improving the stability and reliability of environmental regulation.

[0010] In conjunction with some embodiments of the first aspect, in some embodiments, after determining that the control strategy switching is complete if the steady-state resilience index is within the preset acceptable range in the third time period, the method further includes: after performing the control strategy switching, calculating the actual effectiveness value of the control intervention, the actual effectiveness value being the improvement in steady-state resilience index per unit control cost within one or more preset periods; comparing the actual effectiveness value with a benchmark effectiveness value to obtain a comparison deviation, the benchmark effectiveness value being determined by weighted averaging of multiple historical effectiveness values ​​in a historical effectiveness database; when the comparison deviation exceeds a preset effectiveness decay threshold, determining that the hierarchical strategy library has experienced effectiveness drift; if effectiveness drift is determined to have occurred, adjusting the execution parameters corresponding to the control level in the hierarchical strategy library according to the proportion of the comparison deviation to the benchmark effectiveness value.

[0011] By adopting the above technical solution, after the equipment completes the control strategy switch, it establishes a quantitative evaluation standard for control effectiveness by calculating the improvement in steady-state resilience index per unit control cost as the actual performance value. By comparing the actual performance value with a benchmark performance value obtained by weighted averaging of multiple historical performance values ​​in the historical performance database, the equipment can monitor changes in the performance of the control strategy. When the comparison deviation exceeds a preset performance decay threshold, the equipment determines that performance drift has occurred in the hierarchical strategy library and automatically adjusts the execution parameters of the corresponding control level in the hierarchical strategy library according to the proportion of the comparison deviation in the benchmark performance value. This self-correction mechanism based on performance evaluation enables the equipment to continuously monitor and optimize the execution effect of the control strategy, avoiding the problem of gradual decay of control performance over time and ensuring the long-term effectiveness of the control strategy.

[0012] In conjunction with some embodiments of the first aspect, in some embodiments, after the step of switching the control strategy based on the steady-state resilience index at the disturbance time point so that the steady-state resilience index is within the preset acceptable range after the disturbance time point, the method further includes: performing trend analysis on the disturbance integral time series formed based on the disturbance integral value, where the disturbance integral value is the area enclosed between the portion of the concentration time series data curve exceeding the upper limit of the safe steady-state interval and the horizontal baseline representing the upper limit of the safe steady-state interval; if it is determined that the disturbance integral value decreases over time and stabilizes at a preset low level, it is determined that the break-in is successful, and the control strategy of the rat cage is switched to the conventional feeding control mode; if it is determined that the disturbance integral value shows a trend of not decreasing or continuously increasing over time, it is determined that the break-in has failed and an isolation signal is triggered.

[0013] By adopting the above technical solution, the device establishes a quantitative assessment method for the degree of environmental fluctuation by calculating the area between the portion of the concentration time series data curve exceeding the upper limit of the safe steady-state range and the horizontal baseline as the perturbation integral value. Through trend analysis of the perturbation integral time series, the device can determine the establishment status of mouse group relationships. When the perturbation integral value shows a decreasing trend over time and stabilizes at a preset low level, it indicates that the mice have completed the establishment of group hierarchy relationships, and the device will automatically switch to the conventional feeding control mode. When the perturbation integral value continues to rise or does not decrease, the device determines that the break-in process has failed and triggers an isolation signal. This trend analysis mechanism based on perturbation integral values ​​enables automatic monitoring and intelligent judgment of the mouse group relationship establishment process, avoiding the subjectivity and lag that may be caused by manual observation.

[0014] In conjunction with some embodiments of the first aspect, in some embodiments, after the step of generating concentration time series data based on the acquired carbon dioxide concentration data in the rat cage, the method further includes: when the rate of change of the carbon dioxide concentration data in the rat cage is detected to exceed the maximum rate of change threshold determined based on historical data, identifying the rat cage as an event source rat cage; after increasing the ventilation volume of the event source rat cage to the maximum value set by the device, identifying the rat cage in the same ventilation loop as the event source rat cage as a potential impact rat cage; and switching the potential impact rat cage to a high-priority stable maintenance mode, wherein the high-priority stable maintenance mode ignores conventional energy-saving optimization strategies to adjust the rat cage ventilation volume.

[0015] By employing the aforementioned technical solution, the equipment compares the rate of change of carbon dioxide concentration data monitored in real time with the maximum rate of change threshold determined from historical data, enabling rapid localization of the source of abnormal fluctuations. Once the source cage is identified, the equipment immediately intervenes by increasing its ventilation to the maximum value. Simultaneously, it identifies potentially affected cages sharing the ventilation circuit with the source cage and switches these cages to a high-priority stability maintenance mode. This hierarchical response mechanism based on the fluctuation propagation path, by ignoring conventional energy-saving optimization strategies and prioritizing environmental stability, effectively prevents environmental fluctuations in a single cage from affecting other cages through shared ventilation equipment, thus improving the overall resilience of the rearing equipment.

[0016] In conjunction with some embodiments of the first aspect, in some embodiments, after the step of using the average of the preset value and the recovery time as the steady-state resilience index, the method further includes: when it is determined that the steady-state resilience index is within a preset acceptable range, and the rate of change of the steady-state resilience index over time is negative within a preset second time period, maintaining the current control strategy and reducing the intervention intensity of the control strategy according to the degree of negativeness of the rate of change; when it is determined that the steady-state resilience index is not within the preset acceptable range, and the rate of change of the steady-state resilience index over time is positive within a preset second time period, triggering and executing the highest intensity emergency control strategy; when it is determined that the steady-state resilience index is not within the preset acceptable range, and the rate of change of the steady-state resilience index over time is negative within a preset second time period, maintaining the current control strategy and increasing the intervention intensity of the control strategy according to the degree of negativeness of the rate of change until the steady-state resilience index recovers to the preset acceptable range.

[0017] By adopting the above technical solution, the equipment establishes a multi-condition control strategy adjustment mechanism based on the steady-state resilience index value and its changing trend. When the steady-state resilience index is within an acceptable range and the rate of change is negative, the equipment achieves energy-saving optimization by reducing the intervention intensity; when the steady-state resilience index exceeds the acceptable range and the rate of change is positive, the equipment promptly activates the highest-intensity emergency control strategy for rapid intervention; when the steady-state resilience index exceeds the acceptable range but the rate of change is negative, the equipment dynamically adjusts the intervention intensity according to the specific value of the rate of change until the index returns to normal. This classification and control mechanism based on environmental state and changing trend enables the equipment to adopt corresponding control strategies according to different environmental fluctuations, avoiding over- or under-control responses and achieving precise and intelligent environmental control.

[0018] In conjunction with some embodiments of the first aspect, in some embodiments, the step of determining the disturbance time point when the steady-state resilience index exceeds a preset acceptable range based on the steady-state resilience index and the rate of change specifically includes: determining the time point when the steady-state resilience index will reach the upper limit threshold by linear extrapolation based on the steady-state resilience index value, the rate of change and the upper limit threshold of the preset acceptable range, and determining the calculated future time point as the disturbance time point.

[0019] By adopting the above technical solution, the equipment uses the current value and rate of change of the steady-state resilience index, as well as the upper limit threshold of the preset acceptable range, to predict the future time point when the steady-state resilience index will reach the upper limit threshold through linear extrapolation. This allows the system to determine the optimal control strategy switching time point before environmental parameters reach dangerous levels, avoiding the passive approach of waiting for parameters to exceed limits before taking measures in traditional solutions, thus achieving early control of environmental parameters.

[0020] In conjunction with some embodiments of the first aspect, in some embodiments, the device includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory being used to store computer program code including computer instructions, the one or more processors invoking the computer instructions to cause the device to perform the method as claimed in any one of claims 1-7.

[0021] In a second aspect, this application provides a computer program product containing instructions, characterized in that, when the computer program product is run on a device, the device performs the method as described in the first aspect and any possible implementation thereof.

[0022] Thirdly, this application provides a computer-readable storage medium including instructions that, when executed on a device, cause the device to perform the method described in the first aspect and any possible implementation thereof.

[0023] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. By employing time-series analysis based on carbon dioxide concentration data, defining the safe steady-state range, and evaluating the steady-state resilience index, the technology effectively solves the problem that fixed program control in existing technologies is unable to cope with changes in mouse behavior through quantitative analysis of environmental parameter fluctuation characteristics and prediction of disturbance time points, thereby realizing intelligent predictive control of the mouse cage environment.

[0024] 2. By employing a control level selection mechanism based on a hierarchical strategy library and a smooth switching method that dynamically calculates the transition time of the strategy based on the level difference, the technical problem of sudden changes in environmental parameters caused by control strategy switching in existing technologies is effectively solved, thereby achieving stable transition and fine adjustment of environmental regulation. 3. By adopting a performance evaluation system based on unit control cost and a self-correction mechanism that automatically adjusts execution parameters according to performance deviation, the technical problem of performance decay in the long-term operation of control strategies in the prior art is effectively solved, thereby realizing continuous optimization of control strategies and improvement of resource utilization efficiency. Attached Figure Description

[0025] Figure 1 This is a flowchart illustrating a control decision-making method for a rat cage rearing environment in an embodiment of this application; Figure 2 This is another flowchart illustrating a control decision-making method for a rat cage rearing environment in an embodiment of this application; Figure 3 This is a schematic diagram of an exemplary hardware structure of the device in the embodiments of this application. Detailed Implementation

[0026] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification and appended claims of this application, the singular expressions “a,” “an,” “the,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to and includes any or all possible combinations of one or more of the listed items.

[0027] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0028] Please see Figure 1 This is a flowchart illustrating a control decision-making method for a rat cage rearing environment in an embodiment of this application.

[0029] S101. Generate concentration time series data based on the obtained carbon dioxide concentration data in the rat cage.

[0030] Carbon dioxide concentration data is obtained from the rat cage and converted into concentration time series data. This can be achieved by sensors installed in the rat cage, which can measure the carbon dioxide concentration in the rat cage in real time or periodically.

[0031] S102. When the maximum value in the concentration time series data is higher than the preset conflict threshold and the diurnal rhythm characteristics disappear, the safe steady-state interval is determined based on the moving average and standard deviation of the concentration time series data.

[0032] Check whether the maximum value in the concentration time series data exceeds a preset conflict threshold and whether the diurnal rhythm characteristics have disappeared. If the conditions are met, determine a relatively stable concentration range, i.e., the safe steady-state zone, by expanding the moving average value upwards and downwards by several times the standard deviation. The disappearance of diurnal rhythm characteristics means that the discrete values ​​of the concentration time series data and the preset comparison concentration time series data within the same time period are greater than the preset threshold.

[0033] Understandably, in the early stages of mouse co-cultivation, the intense exercise and metabolic changes caused by fighting will lead to a significant peak in carbon dioxide concentration. The device determines that the group relationship establishment process has begun by simultaneously monitoring two characteristics (concentration peak exceeding the threshold and disappearance of the diurnal rhythm). The system then uses a moving average and standard deviation to adapt to the phased changes in mouse activity intensity and construct a safe steady-state range.

[0034] Specifically, when calculating the moving average of concentration time series data, the size of the moving window can be selected based on the frequency of data fluctuations. For example, for data collected every minute, a 10-minute moving window can be chosen. The standard deviation within the moving window is calculated to measure the dispersion of the data. Based on the moving average and standard deviation, the upper and lower limits of the safe steady-state region are determined. For example, the upper limit can be the moving average plus twice the standard deviation, and the lower limit can be the moving average minus twice the standard deviation.

[0035] In some embodiments, after generating a concentration time series, the device synchronously monitors the instantaneous rate of change of each new reading. If this rate exceeds the maximum rate of change threshold calculated from data over the past 30 days, the device marks the cage as the source cage and forcibly increases its fan speed to the maximum allowed by the device. Subsequently, it searches the topology table of the same ventilation loop for all potentially affected cages and switches these potentially affected cages to a high-priority stability maintenance mode. In this mode, the device completely ignores energy-saving optimization and adjusts the ventilation volume solely with the goal of quickly restoring environmental steady state until the carbon dioxide concentration in all cages within the loop returns to a safe steady-state range. Simultaneously, isolation alerts for these potentially affected cages are sent to the administrator.

[0036] S103. Within a preset first time period, the time period during which the concentration time series data exceeds the safe steady-state range is determined as the recovery time.

[0037] Among them, the preset first time refers to the total observation window used by the device to limit the statistical recovery time.

[0038] As group relationships develop and enter different stages, the fluctuation characteristics of environmental parameters will change accordingly. After each update of the safe steady-state range (such as boundary movement, arrival of new readings, or timed task scenarios), the device first reads the upper and lower limits of the latest safe steady-state range, and then sequentially scans the concentration sequence within a preset first time interval. Whenever a reading is found to have crossed the boundary for the first time, the device records that moment as the starting point and continues scanning until the first reading returns to the range, recording that moment as the ending point. The duration between the starting point and the ending point is recorded as a recovery time. If the boundary-crossing range is not closed by the end of the observation window, the event is marked as incomplete and is not included in the statistics. Since the air supply parameters of the central air supply system include changes in fan speed, temperature and humidity correction, and duct resistance compensation, this process includes the air supply system response time, duct transmission delay, and the influence of filter components. By using the most recent set of recovery time data and combining the dynamic characteristics of each component of the system (such as weighting the recovery time data based on the system response model, where the pipeline transmission delay weight is related to the pipeline length and airflow velocity; the filter influence weight is related to the usage time and resistance changes; and the temperature and humidity regulation weight is related to the controller parameters and environmental inertia), the original recovery time is corrected using a Kalman filter. The filter parameters are dynamically adjusted according to the real-time status of the system to eliminate the influence of the system's inherent delay.

[0039] S104. The average value of the preset recovery time is used as the steady-state resilience index.

[0040] Upon acquiring a new closed-loop recovery time, the device maintains a circular queue of a preset length to represent the overall performance of the environmental control system in responding to the current stage of mouse behavior. Each new recovery time is pushed to the end of the queue; if the queue is full, the oldest recovery time is automatically removed. The device calculates the arithmetic mean of all recovery times in the queue in real time and uses this average as the latest steady-state resilience index. The index is updated in real time as the queue content changes, and is written to the database or memory-mapped file along with the current timestamp.

[0041] In some embodiments, the steady-state resilience index can be generated in a variety of ways: Optionally, the device maintains a floating-point array in the microcontroller using DMA, and the array index is advanced cyclically by modulo operation, and the average value is updated using an incremental method each time a new duration is added; Optionally, the device stores the most recent recovery durations in a List structure in Redis, and calculates the average value atomically on the server side using Lua scripts and publishes it to the message bus, which is not limited here.

[0042] S105. When it is determined that the steady-state resilience index is within a preset acceptable range, and the rate of change of the steady-state resilience index with time is positive within a preset second time period, the disturbance time point when the steady-state resilience index exceeds the preset acceptable range is determined based on the steady-state resilience index and the rate of change.

[0043] Among them, the preset acceptable range refers to the interval between the lower limit and the upper limit set by the device for the steady-state resilience index; the rate of change of the steady-state resilience index over time refers to the speed at which the index increases within a preset second time period; the preset second time period refers to the backtracking window used to calculate the rate of change; and the disturbance time point refers to the future moment when the device predicts that the index will touch the upper limit.

[0044] After updating the steady-state resilience index, the changing trend of the steady-state resilience index is analyzed to predict possible behavioral changes during the establishment of mouse group relationships. First, it is determined whether the current index is within a preset acceptable range. If it is within the range, the index sequence is backtracked to a preset second time period, and the slope of the index increase is obtained through linear fitting. If the slope is positive, the device uses the current index value, slope, and upper limit to perform linear extrapolation to calculate the expected future time when the upper limit is reached. This future time is the disturbance time point. If the slope is not positive or the extrapolation time exceeds the maximum look-ahead time set by the device, the prediction is skipped. The device writes the disturbance time point into the scheduling table and sets a one-time timed task.

[0045] Understandably, if the index has fallen within an acceptable range and the rate of change is negative within the second time window, the current strategy should be maintained and the intervention intensity should be gradually reduced according to the degree of negative change. If the index is still outside the range and the rate of change is positive, the device should immediately trigger the highest intensity emergency control strategy to pull the environment back as quickly as possible. If the index is outside the range and the rate of change is negative, the device should maintain the current strategy framework, but increase the intervention intensity according to the magnitude of the negative change until the index re-enters the acceptable range.

[0046] Understandably, the equipment needs to adjust the fan parameters in advance, optimize the filter's working status, and pre-adjust the temperature and humidity control unit, taking into account pipeline transmission delays and system response characteristics, and reserving sufficient margin for control.

[0047] In some embodiments, when the device determines the disturbance time point, it uses a steady-state resilience index, its rate of change, and an upper limit threshold of the acceptable range for linear extrapolation: the difference between the current resilience index and the upper limit threshold is divided by the rate of change to obtain a remaining time, which is mapped to a future time as the disturbance time point that triggers the policy switching.

[0048] S106. At the disturbance time point, switch the control strategy based on the steady-state resilience index so that the steady-state resilience index is within a preset acceptable range after the disturbance time point.

[0049] When the device's real-time clock reaches the disturbance time point, the device first reads the current policy level and the target level mapped from the hierarchical policy library, calculates the level span, and determines the transition duration based on the predetermined transition rate. During the transition duration, the device gradually adjusts the control parameters, such as the PWM duty cycle of the ventilation fan and the power percentage of the heater, using linear or exponential interpolation. At the same time, it monitors the instantaneous change gradient of carbon dioxide every 5 seconds. If the gradient exceeds the safety limit three times in a row, the device pauses the transition and maintains the current level until the gradient falls back. When the resilience index is detected to have returned to an acceptable range for the third consecutive time, the device marks the policy switch as complete and persists the new policy to the local flash memory.

[0050] In some embodiments, control strategy switching can be implemented in a variety of ways: Optionally, the device executes a PID iteration every 100 milliseconds in the microcontroller with a timer interrupt, gradually adjusts the PWM duty cycle according to the level difference, and records the current level in the EEPROM; Optionally, the device uses a custom controller in the cloud to receive disturbance events, and sends new strategy parameters to the edge nodes through the interface. The nodes ensure that the parameters are executed with linear interpolation at a 10-second granularity after they arrive, which is not limited here.

[0051] In some embodiments, the device invokes a mouse socialization trajectory database before the disturbance time point. This database stores typical conflict and remission timelines for different strains, densities, and ages. Based on the current cage registration information, the device automatically matches the corresponding trajectory, predicts the next critical behavioral transition point, and preloads the control switching strategy into a buffer queue. When the real-time fluctuation characteristics match the trajectory and the critical behavioral transition point is before the disturbance time point, the strategy takes effect immediately at the critical behavioral transition point.

[0052] In some embodiments, after switching control strategies, the device can also perform trend analysis on the perturbation integral time series based on the perturbation integral value, which is defined as the area enclosed between the portion of the concentration time series data curve that exceeds the upper limit of the safe steady-state range within an observation period and the horizontal baseline representing the upper limit of that range. This area value can be obtained by integrating the concentration value of the excess portion as a function of time. The device periodically calculates this integral value to form a time series and performs trend analysis on the perturbation integral time series, for example, by analyzing its slope through linear regression. If the slope is determined to be negative or close to zero for several consecutive periods, and the integral value is stable at a preset low level (e.g., below 10% of the initial peak), it is determined to be a successful integration, indicating that the mice have established a stable social hierarchy. At this time, the control strategy of the mouse cage can be switched to a conventional energy-saving feeding control mode based on diurnal rhythm. If the integral value shows a trend of not decreasing or continuously increasing over time, it is determined to be a failed integration, indicating that there is continuous intense conflict among the mice. At this time, the device triggers an isolation signal to remind the keeper to intervene manually.

[0053] In this embodiment, by employing a steady-state resilience assessment and predictive control mechanism based on concentration data, the problem of fixed program control in the prior art being unable to cope with changes in mouse behavior is effectively solved by analyzing environmental parameter trends and predicting state, combined with an adaptive control strategy, thereby realizing intelligent predictive control of the mouse cage environment.

[0054] In practical applications, once multiple male mice establish a hierarchy by being housed together, subsequent periodic status confirmation behaviors emerge, replacing the initial environmental fluctuations caused by fighting. Under this low-intensity but persistent disturbance, relying solely on the steady-state resilience index for control is insufficient to distinguish the source and characteristics of these environmental fluctuations. This proposed control decision-making method for caged mouse environments can achieve accurate identification and differentiated control of different types of environmental fluctuations by constructing an effectiveness evaluation system based on unit control cost and a hierarchical strategy library parameter self-correction mechanism.

[0055] Please see Figure 2 This is another flowchart illustrating a control decision-making method for a rat cage rearing environment in an embodiment of this application.

[0056] S201. Generate concentration time series data based on the obtained carbon dioxide concentration data in the rat cage.

[0057] S202. When the maximum value in the concentration time series data is higher than the preset conflict threshold and the diurnal rhythm characteristics disappear, the safe steady-state interval is determined based on the moving average and standard deviation of the concentration time series data.

[0058] S203. Within a preset first time period, the time period during which the concentration time series data exceeds the safe steady-state range is determined as the recovery time.

[0059] S204. The average value of the preset recovery time is used as the steady-state resilience index.

[0060] S205. When it is determined that the steady-state resilience index is within a preset acceptable range, and the rate of change of the steady-state resilience index with time is positive within a preset second time period, the disturbance time point when the steady-state resilience index exceeds the preset acceptable range is determined based on the steady-state resilience index and the rate of change.

[0061] Steps S201 to S205 are similar to steps S101 to S105, and will not be described in detail here.

[0062] S206. At the disturbance time point, the target control level is determined from a pre-defined hierarchical strategy library containing multiple intervention levels, based on the steady-state resilience index and the rate of change.

[0063] Among them, the preset hierarchical strategy library refers to a set of control strategies with multiple intervention levels predefined by the device; the target control level refers to the control strategy level selected from the hierarchical strategy library based on the current steady-state resilience index and rate of change.

[0064] Upon reaching the disturbance point, the device first acquires the current steady-state resilience index and its rate of change, then selects the target control level from a pre-defined hierarchical strategy library based on these parameters. Each intervention level in the library corresponds to a different control strategy, such as ventilation volume, temperature regulation, or light intensity. Using a mapping table or algorithm, the device makes an initial assumption about the source and intensity of the disturbance based on the combination of the steady-state resilience index and the rate of change, and selects the most suitable strategy. For example, a high-level strategy (such as "emergency intervention") is designed to address a sharp, violent increase in CO2 concentration, similar to an initial fight; while a medium-to-low-level strategy (such as "active regulation") might be pre-designed to address smaller but persistent fluctuations, such as the periodic status confirmation behavior of male mice.

[0065] In some embodiments, the target control level can be determined in a variety of ways: Optionally, the device runs a decision algorithm on a local server, which maps the steady-state resilience index and rate of change to the corresponding target control level according to a preset rule table; Optionally, the device uses a machine learning model on a cloud platform to predict the most suitable control level through a trained classifier, which is not limited here.

[0066] S207. Calculate the policy transition time based on the gap between the current control level and the target control level.

[0067] After determining the target control level, the equipment calculates the gap between the current control strategy's control level and the target control level. A larger gap indicates a greater need for adjustment. Based on a preset transition rate rule, the equipment calculates the time required to transition from the current level to the target level. The transition rate can be adjusted according to the type of control strategy and the equipment's design requirements. For example, if the control strategy involves adjusting ventilation volume, the transition rate can be set to increase or decrease the ventilation volume by a certain percentage per minute.

[0068] S208. During the strategy transition period, monitor the instantaneous change gradient of carbon dioxide concentration in real time. If the gradient exceeds the safety threshold, dynamically reduce the rate of strategy switching.

[0069] The device continuously monitors the instantaneous gradient of carbon dioxide concentration during the strategy transition period. During this transition, the device monitors the CO2 concentration gradient in real time, and if the gradient exceeds a safe threshold, mouse behavior may change, prompting the device to dynamically adjust the rate of strategy switching. For example, if the concentration change is too rapid, the device may slow the increase or decrease in ventilation volume, or adjust the rate of change of other control parameters.

[0070] S209. If the steady-state resilience index is within the preset acceptable range within the third time period, then the control strategy switching is determined to be complete.

[0071] This process is executed after the device performs a policy transition. The device continuously monitors the steady-state resilience index for a preset third time interval to verify the effectiveness of the new control policy. If the steady-state resilience index remains within a preset acceptable range during the preset third time interval, the device confirms the control policy switch is complete. If the steady-state resilience index fails to remain within the acceptable range, the control policy is reassessed or an alarm is triggered.

[0072] S210. Calculate the actual effectiveness value of the control intervention. The actual effectiveness value is the improvement in the steady-state resilience index per unit control cost within one or more preset periods.

[0073] Among them, control intervention refers to the adjustment of the environment by the equipment according to the control strategy; actual effectiveness value refers to the ratio of the improvement in steady-state resilience index achieved by control intervention within a specific period to the unit control cost; unit control cost refers to the resource consumption required to implement control intervention, such as energy cost or equipment wear and tear cost.

[0074] This process is executed after the equipment confirms the control strategy switchover. The equipment first sets one or more preset cycles to evaluate the effectiveness of the control intervention. The equipment calculates the improvement in the steady-state resilience index, i.e., the difference in the steady-state resilience index before and after the control intervention. Simultaneously, the total additional operating time of the fan, the additional electricity consumed by the heater, and the energy consumption resulting from the increased brightness are all converted into a unified unit of yuan per cycle, including energy consumption or equipment operating time, such as the energy cost consumed by the fan and other equipment.

[0075] S211. Compare the actual performance value with the benchmark performance value to obtain the comparison deviation.

[0076] After calculating the actual performance value, the device retrieves the actual performance value under the current control strategy from the database and reads the corresponding baseline performance value from the historical performance database.

[0077] The baseline performance value can be obtained by weighted averaging of historical performance data or by calculation using a theoretical model. The equipment compares the actual performance value with the baseline performance value and calculates the difference between the two, i.e., the comparison deviation.

[0078] S212. Determine whether the hierarchical strategy library algorithm has experienced performance drift based on the comparison bias.

[0079] The device first reads the preset drift judgment threshold, which is set by the experimenter and can be adjusted independently for different cage positions. If the absolute value of the deviation is higher than this threshold for three consecutive samples, the device records a drift event and writes the drift flag into the real-time status table in memory. If the drift flag remains valid for two consecutive monitoring cycles, the device further generates a drift alarm and pushes it to the operation and maintenance terminal. At the same time, the timestamp of the drift event, the deviation value, and the triggering conditions are packaged and stored in the historical log.

[0080] S213. If performance drift is determined to have occurred, the execution parameters corresponding to the control level in the hierarchical strategy library shall be adjusted according to the proportion of the comparison deviation in the baseline performance value.

[0081] Execution begins immediately after the drift flag is set. The device first calculates the percentage, which is calculated by dividing the comparison deviation by the baseline performance value, and the result is expressed as a percentage with one decimal place. Then, the device linearly scales the execution parameters of the corresponding control level according to this percentage: if the percentage reaches 10%, all execution parameters of that level are reduced by 10%; if the percentage reaches 20%, they are reduced by 20%, and so on, until the absolute value of the adjusted deviation is lower than the drift threshold. The minimum adjustment step size is 1% to avoid excessive oscillation. After each adjustment, the device immediately writes the new parameters back to the local policy library and ensures complete writing through CRC verification. If parameter writing fails, the device automatically rolls back to the previous version and records the reason for the failure in the log.

[0082] In some embodiments, the device categorizes the execution parameters into three sensitivity levels—high, medium, and low—based on their impact on mouse welfare. After a drift occurs, the percentage of deviation is first calculated, and then power-law, linear, or stepwise adjustment functions are applied according to the sensitivity level. High-sensitivity parameters such as ventilation frequency are fine-tuned for small deviations and sharply increased for large deviations; low-sensitivity parameters such as sampling period only change when the percentage crosses the threshold of the entire level, allowing different parameters to be corrected differently for the same drift event.

[0083] In some embodiments, adaptive adjustment of execution parameters can be achieved in a variety of ways: Optionally, the device performs closed-loop calibration within the MCU, and after each drift event is triggered, the execution parameters are gradually corrected in steps of 5%, and three backups are stored in the EEPROM to prevent loss due to power failure; Optionally, the device generates new parameters in the cloud using a linear mapping function based on proportion, and pushes them to the edge node via OTA differential packets. After receiving the parameters, the node verifies the hash value and atomically replaces the old parameters. This is not limited here.

[0084] This method first extracts a steady-state resilience index during the period when the diurnal rhythm disappears due to intense fighting. Then, it periodically confirms the index and its rate of change to match the target control level from the hierarchical strategy library. During the transition period, it dynamically limits the speed based on the instantaneous concentration gradient. After the switch is completed, the actual effectiveness is evaluated by the resilience improvement per unit energy consumption. If there is a drift relative to the benchmark value, the execution parameters are adaptively written back according to the deviation ratio. This avoids both over-intervention and control lag, and ultimately keeps the rat cage environment in a steady state for a long time.

[0085] The exemplary device 300 provided in the embodiments of this application is described below. Figure 3 This is an exemplary hardware structure diagram of the device 300 provided in the embodiments of this application.

[0086] In some embodiments, the device 300 is a computer device or includes a computer device. The computer device includes a processor, memory, and a network interface connected via a system bus. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the computer device stores data. The network interface of the computer device is used to communicate with other external terminals or servers via a network connection. In some embodiments, the network interface can be a wired network interface; in some embodiments, the network interface can also be a wireless network interface. When the computer program is executed by the processor, it implements the methods in the embodiments of this application.

[0087] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0088] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0089] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as meaning "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as meaning "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

[0090] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive), etc.

[0091] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A method for controlling the decision-making process of a caged rodent rearing environment, characterized in that, The method includes: Concentration time series data are generated based on the obtained carbon dioxide concentration data in the rat cages; When the maximum value in the concentration time series data is higher than the preset conflict threshold and the diurnal rhythm feature disappears, a safe steady-state interval is determined based on the moving average and standard deviation of the concentration time series data. The disappearance of the diurnal rhythm feature means that the discrete value of the concentration time series data and the preset comparison concentration time series data in the same time period is greater than the preset threshold. Within a preset first time period, the time period during which the concentration time series data exceeds the safe steady-state range is determined as the recovery time. The average of the preset recovery times is used as the steady-state resilience index; When it is determined that the steady-state resilience index is within a preset acceptable range, and the rate of change of the steady-state resilience index with time is positive within a preset second time period, the disturbance time point when the steady-state resilience index exceeds the preset acceptable range is determined based on the steady-state resilience index and the rate of change. At the disturbance time point, the control strategy is switched based on the steady-state resilience index so that the steady-state resilience index is within the preset acceptable range after the disturbance time point.

2. The method according to claim 1, characterized in that, The step of switching the control strategy based on the steady-state resilience index at the disturbance time point to bring the steady-state resilience index within the preset acceptable range specifically includes: At the disturbance time point, based on the steady-state resilience index and the rate of change, the target control level is determined from a preset hierarchical strategy library containing multiple intervention levels; Based on the difference between the control level of the current control strategy and the target control level, the strategy transition time is calculated. The transition time is determined based on the fact that the larger the level span, the longer the transition time. During the strategy transition period, the rate of strategy switching is adjusted based on the instantaneous change gradient of carbon dioxide concentration monitored in real time. If the steady-state resilience index is within the preset acceptable range within the third time period, the control strategy switching is determined to be complete.

3. The method according to claim 2, characterized in that, After determining that the control strategy switching is complete when the steady-state resilience index is within the preset acceptable range for a third time period, the method further includes: After the control strategy switch is executed, the actual effectiveness value of the control intervention is calculated. The actual effectiveness value is the improvement in steady-state resilience index per unit control cost within one or more preset periods. The actual performance value is compared with a benchmark performance value to obtain the comparison deviation. The benchmark performance value is determined by weighted averaging of multiple historical performance values ​​in the historical performance database. When the comparison deviation exceeds a preset performance decay threshold, it is determined that the hierarchical strategy library has experienced performance drift. If a performance drift is determined to have occurred, the execution parameters corresponding to the control level in the hierarchical strategy library are adjusted according to the proportion of the comparison deviation to the baseline performance value.

4. The method according to claim 1, characterized in that, After the step of switching the control strategy based on the steady-state resilience index at the disturbance time point, so that the steady-state resilience index is within the preset acceptable range after the disturbance time point, the method further includes: A trend analysis is performed on the perturbation integral time series formed based on the perturbation integral value, wherein the perturbation integral value is the area enclosed between the portion of the concentration time series data that exceeds the upper limit of the safe steady-state interval and the horizontal baseline representing the upper limit of the safe steady-state interval; If it is determined that the disturbance integral value decreases over time and stabilizes at a preset low level, it is determined that the break-in is successful, and the control strategy of the rat cage is switched to the conventional feeding control mode. If the disturbance integral value is determined to show a trend of not decreasing or continuously increasing over time, it is determined to be a break-in failure and an isolation signal is triggered.

5. The method according to claim 1, characterized in that, After the step of generating concentration time series data based on the acquired carbon dioxide concentration data in the rat cage, the method further includes: When the rate of change of carbon dioxide concentration data in the rat cage exceeds the maximum rate of change threshold determined based on historical data, the rat cage is identified as the source rat cage of the event. After increasing the ventilation volume of the rat cage that is the source of the incident to the maximum value set by the equipment, the rat cage that is in the same ventilation circuit as the rat cage that is the source of the incident is identified as a potential impact rat cage. The potentially affected rat cages are switched to a high-priority stability maintenance mode, which ignores conventional energy-saving optimization strategies to adjust the rat cage ventilation.

6. The method according to claim 1, characterized in that, After the step of using the average of the preset value and the recovery time as the steady-state resilience index, the method further includes: When it is determined that the steady-state resilience index is within a preset acceptable range, and the rate of change of the steady-state resilience index over time is negative within a preset second time period, the current control strategy is maintained and the intervention intensity of the control strategy is reduced according to the degree to which the rate of change is negative. When it is determined that the steady-state resilience index is not within the preset acceptable range, and the rate of change of the steady-state resilience index over time is positive within a preset second time period, the highest intensity emergency control strategy is triggered and executed. When it is determined that the steady-state resilience index is not within the preset acceptable range, and the rate of change of the steady-state resilience index over time is negative within a preset second time period, the current control strategy is maintained and the intervention intensity of the control strategy is increased according to the degree of negativeness of the rate of change until the steady-state resilience index recovers to the preset acceptable range.

7. The method according to claim 1, characterized in that, The step of determining the disturbance time point where the steady-state resilience index exceeds a preset acceptable range based on the steady-state resilience index and the rate of change specifically includes: Based on the steady-state resilience index, the rate of change, and the upper limit threshold of the preset acceptable range, the time point at which the steady-state resilience index will reach the upper limit threshold is determined by linear extrapolation, and the calculated future time point is determined as the disturbance time point.

8. A device, characterized in that, The device includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory being used to store computer program code, the computer program code including computer instructions, the one or more processors invoking the computer instructions to cause the device to perform the method as described in any one of claims 1-7.

9. A computer program product containing instructions, characterized in that, When the computer program product is run on the device, the device causes the device to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium comprising instructions, characterized in that, When the instructions are executed on the device, the device causes the device to perform the method as described in any one of claims 1-7.

Images