Fibroblast storage device and method

Abstract

The invention provides a fibroblast storage device and method, and relates to the technical field of microbiological devices.The device comprises a monitoring module, a control module and a storage module, the storage prediction module performs cell stage division according to the storage label and the real-time cell state data, establishes stage constraints, performs cell storage environment prediction by using the stage constraints and the real-time cell state data, and establishes a target prediction result; the environment detection module performs real-time acquisition of a storage environment of fibroblasts and establishes a storage environment data set; and the intelligent adjustment module takes the storage environment data set as an environment starting point, takes the target prediction result as a following target, executes search optimization of environment control under the constraint of the adjustment granularity identifier, and performs intelligent adjustment of the storage environment according to a search optimization result. The technical problem that the fibroblast storage reliability is poor in the prior art can be solved, and the technical effect of improving the storage reliability is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of microbiology devices, and particularly relates to a fibroblast storage device and method. BACKGROUND

[0002] With the rapid development of cell engineering, biological sample library and precision medicine technologies, intelligent adjustment and long-term stability of the fibroblast storage environment have become important links affecting the quality and activity of cell preservation.

[0003] At present, existing fibroblast storage regulation technologies mostly rely on static setting and single-point monitoring methods, that is, presetting temperature and humidity parameters and the like, and collecting data through simple sensors to make threshold judgments. When the detected value exceeds the limited range, an alarm is passively sent out or compensation adjustment is performed. This mode lacks real-time prediction and dynamic response capability for environmental change trends, and cannot identify the cumulative effects caused by potential deviations or adjustment lags in time. For example, when the refrigeration system has a slow energy efficiency decay, the temperature may gradually deviate from the target range, but since it does not exceed the alarm threshold, it cannot be responded in advance, ultimately resulting in hidden cold damage to the cells.

[0004] In summary, the existing technology has the technical problem that due to the lack of a multi-dimensional dynamic parameter coordination modeling and trend prediction mechanism in the fibroblast storage environment, it is impossible to perform feedforward regulation and active correction when there is a slight deviation in the environmental parameters, further affecting the activity maintenance stability of the fibroblasts in the long-term storage process, the precision of environmental control, and the overall safety and reliability of the storage system. SUMMARY

[0005] The purpose of the present application is to provide a fibroblast storage device and method to solve the technical problem in the prior art that due to the lack of a multi-dimensional dynamic parameter coordination modeling and trend prediction mechanism in the fibroblast storage environment, it is impossible to perform feedforward regulation and active correction when there is a slight deviation in the environmental parameters, further affecting the activity maintenance stability of the fibroblasts in the long-term storage process, the precision of environmental control, and the overall safety and reliability of the storage system.

[0006] In view of the above problems, the present application provides a fibroblast storage device and method.

[0007] In a first aspect, this application provides a fibroblast storage device, comprising: a monitoring module for monitoring fibroblast growth and establishing real-time cell state data; a storage prediction module for reading storage tags of fibroblasts, dividing cells into stages based on the storage tags and the real-time cell state data, establishing stage constraints, predicting the cell storage environment using the stage constraints and the real-time cell state data, and establishing a target prediction result, wherein the target prediction result is set with an adjustment granularity identifier; an environment detection module for activating a sensor array, performing real-time acquisition of the fibroblast storage environment, and establishing a storage environment dataset, wherein the storage environment dataset includes temperature, humidity, and gas composition; and an intelligent adjustment module for using the storage environment dataset as an environmental starting point, taking the target prediction result as a following target, performing environmental control search and optimization under the constraint of the adjustment granularity identifier, and performing intelligent adjustment of the storage environment based on the search and optimization result.

[0008] Preferably, the fibroblast storage device is further configured to: after reading the stage constraints, construct a stage-aware network based on the stage constraints, the stage-aware network including a stage regulation channel mapped to the stage constraints; select and activate the corresponding stage regulation channel according to the current stage of the fibroblast, and synchronously input the environmental starting point, the target being followed, and the regulation granularity identifier into the activated stage regulation channel; and use the stage regulation channel to perform search optimization under a focusing strategy to establish search optimization results.

[0009] Preferably, the fibroblast storage device is further configured to: configure a focusing strategy using the stage regulation channel, wherein the focusing strategy is a sensitive environmental factor focusing strategy mapped to the stage; initialize and configure an initial solution set based on the focusing strategy and the regulation granularity identifier, and perform evolutionary simulation of the initial solution set in the stage regulation channel, wherein the evolutionary simulation is performed with the environmental starting point as the simulation starting point and the initial solution set as the evolutionary scheme, and the evolutionary process and the deviation between the evolutionary result and the target being followed are recorded; evaluate the fitness of the initial solution set based on the evolutionary process and the target deviation, and establish a fitness evaluation result; perform iterative updates of the initial solution set based on the fitness evaluation result and the focusing strategy to complete the search and optimization.

[0010] Preferably, the fibroblast storage device is further configured to: after locating key environmental factors according to the focusing strategy, set a first update adjustment bias for parameters based on the location results, the first update adjustment bias including parameter priority bias and adjustment coverage bias; select optimal solutions within the initial solution set according to the fitness evaluation results, and establish a second update adjustment bias using the optimal solution selection results; configure a random search function, and after performing a random search of the initial solution set using the random search function, complete iterative updates based on the random search results, the first update adjustment bias, and the second update adjustment bias.

[0011] Preferably, the fibroblast storage device is further configured to: collect iteration round data and adaptation evaluation data of multiple iterations during the iterative update process; perform convergence phase evaluation based on the iteration round data and adaptation evaluation data; generate phase constraints; perform random adjustment of the random search function based on the phase constraints; and simultaneously compensate the bias weights of the first update adjustment bias and the second update adjustment bias to complete the iterative update.

[0012] Preferably, the fibroblast storage device is further configured to: control the storage environment adjustment according to the search optimization results, and read the monitoring response through a monitoring module, wherein the monitoring response includes the environmental change monitoring response and the biological response of the fibroblasts; construct interactive correction terms using the monitoring response, and perform control optimization management of the search optimization results.

[0013] Preferably, the fibroblast storage device is further configured to: establish a first interactive correction term based on the environmental change monitoring response and the search optimization control result, wherein the first interactive correction term is a control response deviation correction term; perform biological response adaptation analysis based on the biological response and the environmental change monitoring response, and establish a second interactive correction term, wherein the second interactive correction term is a biological adaptation feedback correction term; and establish an interactive correction term based on the first interactive correction term and the second interactive correction term.

[0014] Preferably, the fibroblast storage device is further configured to: input the stage constraints and the real-time cell state data into a multimodal prediction model, perform environmental prediction of the growth stage using the temporal analysis layer in the multimodal prediction model, and establish the target prediction result.

[0015] Preferably, the fibroblast storage device is further configured with: an intelligent early warning module, used to create an early warning strategy based on the adjustment granularity identifier and the search optimization results, the early warning strategy including an adjustment deviation early warning strategy and an adjustment gradient early warning strategy, and using the early warning strategy to monitor and verify the cell storage environment adjustment process by a monitoring module, and to issue an early warning.

[0016] Secondly, this application also provides a method for storing fibroblasts, comprising: performing fibroblast growth monitoring to establish real-time cell state data; reading the storage tags of fibroblasts, dividing the cells into stages based on the storage tags and the real-time cell state data, establishing stage constraints, using the stage constraints and the real-time cell state data to predict the cell storage environment, and establishing a target prediction result, wherein the target prediction result is set with an adjustment granularity identifier; activating a sensor array to perform real-time acquisition of the fibroblast storage environment, establishing a storage environment dataset, wherein the storage environment dataset includes temperature, humidity, and gas composition; using the storage environment dataset as the starting point of the environment, taking the target prediction result as the target to follow, performing environmental control search and optimization under the constraint of the adjustment granularity identifier, and performing intelligent adjustment of the storage environment based on the search and optimization result.

[0017] The technical solution provided in this application has at least the following technical effects or advantages: by achieving the technical goals of multimodal perception, dynamic prediction and intelligent adaptive regulation of the fibroblast storage environment, it can achieve the technical effects of real-time identification of environmental deviations, early correction of regulatory behavior and continuous optimization of storage conditions, thereby ensuring the long-term activity of fibroblasts and the operational stability of the storage system.

[0018] The above description is merely an overview of the technical solution of this application. To enable a clearer understanding of the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the structure of a fibroblast storage device according to this application.

[0021] Figure 2 This is a schematic diagram of a fibroblast storage method according to this application.

[0022] Explanation of reference numerals in the attached diagram: Monitoring module 1, Storage and prediction module 2, Environmental detection module 3, Intelligent adjustment module 4. Detailed Implementation

[0023] This application provides a fibroblast storage device and method, solving the technical problem in existing technologies where the collaborative modeling and trend prediction mechanism of multidimensional dynamic parameters in the fibroblast storage environment fails to provide feedforward adjustment and proactive correction when environmental parameters deviate slightly. This further affects the stability of fibroblast activity maintenance, the accuracy of environmental control, and the overall safety and reliability of the storage system during long-term storage. The application achieves the technical goals of multimodal perception, dynamic prediction, and intelligent adaptive regulation of the fibroblast storage environment, enabling real-time identification of environmental deviations, early correction of regulatory behavior, and continuous optimization of storage conditions, thereby ensuring the long-term activity of fibroblasts and the operational stability of the storage system.

[0024] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0025] Example 1, please refer to the appendix. Figure 1 This application provides a fibroblast storage device, specifically comprising: Monitoring module 1 is used to perform fibroblast growth monitoring and establish real-time cell status data.

[0026] Specifically, monitoring fibroblast growth via monitoring module 1 involves continuously observing and recording fibroblasts at different culture stages through methods such as microscopic image acquisition, fluorescent labeling detection, or biochemical sensing to understand their growth changes. Fibroblasts are cells found in connective tissue, responsible for synthesizing collagen and repairing tissue damage. Their growth status typically includes multi-dimensional information such as cell number, morphology, activity, and metabolic level. The data acquisition module integrates information such as images, temperature, pH, and nutrient concentration obtained during monitoring into a dynamically updated dataset, establishing real-time cell state data to reflect the physiological state of cells at each time point.

[0027] Storage prediction module 2 is used to read the storage tags of fibroblasts, divide the cells into stages based on the storage tags and the real-time cell state data, establish stage constraints, use the stage constraints and the real-time cell state data to predict the cell storage environment, and establish a target prediction result, wherein the target prediction result is set with an adjustment granularity identifier.

[0028] Furthermore, this application also includes: inputting the stage constraints and the real-time cell state data into a multimodal prediction model, using the time-series analysis layer in the multimodal prediction model to perform environmental prediction of the growth stage, and establishing the target prediction result.

[0029] Specifically, reading the storage tags of fibroblasts refers to assigning a unique identification information to each batch of fibroblasts before they are stored or cultured. This information includes cell origin, culture time, passage number, culture medium type, and storage temperature. Storage tags are used to identify the historical characteristics of cells, accurately identifying cell populations from different batches or at different stages. Cell stage segmentation based on storage tags and real-time cell state data involves comprehensively analyzing parameters such as fibroblast growth rate, metabolic level, and morphological changes after acquiring tag information and current cell state data. This allows for the determination of different physiological stages, such as early growth, proliferative phase, or stationary phase, thus establishing stage constraints. Stage segmentation can be accomplished using machine learning models. For example, when the fibroblast activity index continuously increases and the metabolic rate exceeds a certain threshold, the fibroblasts are classified as being in a rapid proliferation phase. Stage constraints refer to setting ranges and constraints for subsequent environmental control based on the differences in sensitivity to environmental conditions at different stages.

[0030] Furthermore, inputting stage constraints and real-time cell state data into the multimodal prediction model refers to feeding the constraints corresponding to the cell growth stage and the real-time collected cell physiological data as input information into a prediction system capable of processing multiple types of data. A multimodal prediction model is a computational model that can simultaneously process data from different sources and in different formats; for example, it can simultaneously perform feature extraction and correlation analysis on image information, numerical signals, and text labels. Utilizing the temporal analysis layer within the multimodal prediction model to perform environmental prediction for the growth stage refers to incorporating a structural layer within the model capable of identifying patterns of temporal change. This allows for the prediction of optimal environmental parameters such as temperature, humidity, and gas concentration for future time periods by analyzing the continuous changing trends of data during cell growth. Based on the output calculated by the multimodal prediction model, reference targets are generated to guide subsequent environmental control, establishing target prediction results, including predicted values ​​and trends of various environmental factors. The construction process of the multimodal prediction model includes multi-source fusion of input data, construction of a feature extraction network, and modeling of temporal dependencies. The multimodal prediction model first performs modal alignment on different types of data, mapping information such as images, numerical values, and text to a unified feature space through feature encoding. After modal alignment, a feature fusion mechanism is employed to weight information from different sources or apply self-attention weights, enabling the multimodal prediction model to identify features that are more critical for prediction at specific growth stages. Following fusion, a temporal analysis layer is added to capture patterns of change over time. This temporal analysis layer can employ structures such as Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), or one-dimensional convolutional temporal networks to record the evolutionary path of cell states at different time points and identify periodicity and mutation trends in the data. For example, when cell metabolic rate, oxygen consumption, and morphological changes show a synchronous upward trend over a period of time, the temporal analysis layer will convert the correlation into predictive weights for future environmental temperature and nutrient supply.

[0031] The target prediction result is set with an adjustment granularity identifier, which means that a parameter describing the control precision is added to the output result. This parameter indicates the step size or sensitivity level that the environmental control should take. For example, when the adjustment granularity identifier value is 0.1, it means that the temperature adjustment step is 0.1 degrees Celsius.

[0032] The environmental detection module 3 is used to activate the sensor array, perform real-time acquisition of the storage environment of fibroblasts, and establish a storage environment dataset, which includes temperature, humidity, and gas composition.

[0033] Specifically, the sensor array is activated, initiating the detection module composed of multiple sensors of different types to enter operational status, enabling simultaneous sensing and acquisition of various physical and chemical parameters of the cell storage environment. The sensor array is a multi-sensor system, including temperature sensors, humidity sensors, gas sensors, and optical sensors, which, through a spatially distributed layout, achieve comprehensive monitoring of environmental conditions at various locations within the incubator or storage chamber. Real-time acquisition of the fibroblast storage environment refers to continuously recording environmental changes at set time intervals during cell storage, ensuring the continuity and immediacy of monitoring data. Real-time acquisition not only reflects the dynamic changes of the environment at different points in time but also captures sudden deviations, establishing a storage environment dataset, thereby providing immediate feedback for subsequent environmental adjustments. The storage environment dataset includes temperature, humidity, and gas composition, signifying a structured record of key parameters for storage environment control. For example, temperature describes the thermal state, humidity reflects water evaporation and the risk of cell desiccation, and gas composition includes the proportions of oxygen, carbon dioxide, and nitrogen, used to assess respiration and metabolic suitability. For example, when the temperature is maintained at 37 degrees Celsius, the humidity at 85%, and the oxygen concentration at 20%, the cells are in a relatively stable metabolic state. However, if the humidity drops to 70%, it may lead to increased cell membrane permeability and cause metabolic stress response.

[0034] The intelligent adjustment module 4 is used to take the storage environment dataset as the starting point, take the target prediction result as the target to follow, perform environmental control search and optimization under the constraint of adjustment granularity identifier, and perform intelligent adjustment of storage environment according to the search and optimization result.

[0035] Furthermore, this application also includes: after reading the stage constraints, constructing a stage-aware network based on the stage constraints, the stage-aware network including a stage regulation channel mapped to the stage constraints; selecting and activating the corresponding stage regulation channel according to the current stage of the fibroblast, and synchronously inputting the environmental starting point, the target being followed, and the regulation granularity identifier into the activated stage regulation channel; and using the stage regulation channel to perform search optimization under a focusing strategy and establish search optimization results.

[0036] Furthermore, this application also includes: configuring a focusing strategy using the stage adjustment channel, wherein the focusing strategy is a sensitive environmental factor focusing strategy mapped to the stage; initializing and configuring an initial solution set based on the focusing strategy and adjustment granularity identifier, and performing evolutionary simulation of the initial solution set in the stage adjustment channel, wherein the evolutionary simulation is performed with the environmental starting point as the simulation starting point, the initial solution set as the evolution scheme, and recording the evolution process and the deviation between the evolution results and the target being followed; evaluating the fitness of the initial solution set based on the evolution process and the target deviation, and establishing a fitness evaluation result; performing iterative updates of the initial solution set based on the fitness evaluation result and the focusing strategy to complete the search optimization.

[0037] Furthermore, this application also includes: after locating key environmental factors according to the focusing strategy, setting a first update adjustment bias for parameters based on the location results, the first update adjustment bias including parameter priority bias and adjustment coverage bias; selecting optimal solutions within the initial solution set according to the fitness evaluation results, and establishing a second update adjustment bias using the optimal solution selection results; configuring a random search function, performing a random search of the initial solution set using the random search function, and completing iterative updates based on the random search results, the first update adjustment bias, and the second update adjustment bias.

[0038] Furthermore, this application also includes: collecting iteration round data and adaptation evaluation data from multiple iterations; performing convergence phase evaluation based on the iteration round data and adaptation evaluation data; generating phase constraints; performing random adjustment of the random search function based on the phase constraints; and simultaneously compensating the bias weights of the first update adjustment bias and the second update adjustment bias to complete the iterative update.

[0039] Furthermore, this application also includes: controlling the adjustment of the storage environment according to the search optimization results, and reading the monitoring response through the monitoring module, wherein the monitoring response includes the environmental change monitoring response and the biological response of fibroblasts; constructing interactive correction terms using the monitoring response, and performing control optimization management of the search optimization results.

[0040] Furthermore, this application also includes: establishing a first interactive correction term based on the environmental change monitoring response and the search optimization control result of the monitoring response, wherein the first interactive correction term is a control response deviation correction term; performing an adaptation analysis of the biological response based on the biological response and the environmental change monitoring response to establish a second interactive correction term, wherein the second interactive correction term is a biological adaptation feedback correction term; and establishing an interactive correction term based on the first interactive correction term and the second interactive correction term.

[0041] Specifically, using the storage environment dataset as the environmental starting point means using the currently collected storage environment data as the initial reference state before performing environmental regulation. The environmental starting point represents a comprehensive understanding of the current cell storage environment, including real-time parameters across multiple dimensions such as temperature, humidity, oxygen, and carbon dioxide concentrations, collectively describing the actual environmental state of the cells. Using the environmental starting point avoids blind adjustments in control calculations, ensuring that each adjustment is based on the true and up-to-date environmental state. Using the target prediction result as the target to follow means using the optimal environmental conditions output by the multimodal prediction model as the reference direction for the next control step, gradually bringing the current environment closer to the ideal state calculated by the model. Target following is a closed-loop tracking mechanism that compares the difference between the current environmental state and the target prediction value in real time and dynamically adjusts the regulation strategy based on the magnitude of the deviation.

[0042] Furthermore, after reading the stage constraints, a stage-aware network is constructed based on these constraints. This refers to establishing a neural network model capable of sensing and distinguishing the characteristics of each stage after acquiring the constraints for different cell growth stages. The stage-aware network is a structured computational network that can identify the current stage of the cell based on these constraints and predict the sensitivity of that stage to environmental factors. The stage-aware network includes stage regulation channels mapped to the stage constraints, meaning each channel corresponds to a specific growth stage and is responsible for performing personalized environmental regulation calculations based on the characteristics of that stage.

[0043] Selecting and activating the corresponding stage-specific regulatory channels based on the current stage of fibroblasts refers to analyzing real-time monitoring data to determine the cell's current stage and then activating only the regulatory channels relevant to that stage. This improves computational efficiency and avoids interference from irrelevant parameters. For example, when cells are in a rapid growth phase, the "proliferation channel" is automatically activated while the "dormant channel" is disabled, focusing on key variables such as temperature and carbon dioxide concentration. The environmental starting point, target, and regulatory granularity flag are simultaneously input into the activated stage-specific regulatory channel. This means inputting the current environmental state, the predicted ideal environmental target, and the regulatory precision parameters into the channel, enabling targeted optimization within a specific stage. The regulatory granularity flag controls the fineness of the regulatory action; for example, when the granularity flag value is small, smoother and more precise fine-tuning operations are performed.

[0044] Furthermore, utilizing stage-based regulatory channels to configure focusing strategies means establishing corresponding control channels based on the changing patterns of different biological stages. This allows for selective focus on environmental factors that have the greatest impact on fibroblast status. A focusing strategy is an optimization guidance mechanism aimed at screening for the most sensitive regulatory factors among multidimensional environmental variables, concentrating the search space on high-influence regions. For example, when fibroblasts are in an active division phase, temperature and oxygen concentration are likely the most important sensitive factors, and the focusing strategy will prioritize optimizing these two parameters. Stage mapping refers to mapping different stages of the cell life cycle to corresponding environmental response patterns, thereby automatically activating appropriate control logic at specific stages.

[0045] Next, the initial solution set is initialized based on the focusing strategy and adjustment granularity identifier, meaning a set of candidate control schemes is generated according to the set optimization range and control precision. The initial solution set consists of several hypothetical schemes at the start of the search, each containing different combinations of temperature, humidity, and gas ratios. The evolution of the initial solution set is simulated in the staged adjustment channel, allowing the initial schemes to evolve in the model environment through simulation, observing their gradual approach to the target state. Using the environmental starting point as the simulation starting point illustrates that dynamic adjustments are performed to approximate the target starting from the current real environmental conditions; using the initial solution set as the evolution scheme means that each candidate solution will undergo independent experimental evolution. The deviation between the evolution process and the result and the target is used to quantify the degree of optimization. For example, if the target is to maintain cell viability at 90%, the deviation can be expressed as absolute error or relative error.

[0046] Then, the fitness evaluation of the initial solution set is performed based on the evolution process and the target deviation. This refers to judging the merits of each solution based on its performance during the simulation. Fitness evaluation is a quantitative method used to measure the effectiveness of a solution in achieving the goal. Solutions with smaller target deviations and more stable evolution processes have higher fitness. The fitness evaluation results can be obtained through a scoring function or a normalization algorithm, such as converting the errors of all solutions into scores between 0 and 1, where 1 represents the optimal solution. This ensures that high-quality solutions have a greater chance of being retained and improved in the next iteration.

[0047] Furthermore, after locating key environmental factors according to the focusing strategy, the first update adjustment bias based on the location results refers to determining the direction and weight of parameter updates based on the identified key factors. The first update adjustment bias includes parameter priority bias and adjustment coverage bias. Parameter priority bias means that key environmental factors have higher adjustment weights and priority update opportunities during the update process; adjustment coverage bias means giving key environmental factors a wider or denser search range in terms of the range or magnitude of parameter changes. For example, when temperature is identified as a key factor, a more detailed exploration will be conducted in the temperature dimension, while humidity may maintain a smaller adjustment range, thereby improving convergence speed and accuracy.

[0048] Next, optimal solutions are selected from the initial solution set based on the fitness evaluation results. This means selecting the solution with the smallest objective bias and best performance from the multiple solutions obtained in the stage evolution as the representative solution. The fitness evaluation results reflect the quality of each solution, and optimal solution selection refers to selecting a subset of solutions with the highest scores or smallest errors to guide the next optimization step. A second update adjustment bias is established using the optimal solution selection results, representing a new optimization trend based on common features or parameter distribution patterns among the optimal solutions. By extracting the features of optimal solutions, subsequent iterations can focus on searching these feature regions.

[0049] Subsequently, a random search function is configured, indicating that, based on the determined optimization bias, a certain proportion of random perturbation is introduced to prevent getting trapped in local optima. The random search function is a commonly used exploration method in optimization algorithms, ensuring the diversity of the search process by randomly sampling in the parameter space. Using the random search function for a random search of the initial solution set means that, based on the existing solution set, new candidate solutions are generated through random perturbation to test whether a better combination exists. Iterative updates are completed based on the random search results, the first update adjusting the bias, and the second update adjusting the bias. This means that three information sources are comprehensively considered: new solutions obtained through random exploration, the first bias guided by key factors, and the second bias formed by learning from optimal solutions, to dynamically update the overall solution set. This maintains the exploratory nature of the algorithm while continuously optimizing based on learning existing patterns, thus forming a multi-level, global search strategy.

[0050] Furthermore, collecting iteration round data and adaptation evaluation data from multiple iterations represents a systematic record of the operational status and performance of each iteration. Iteration round data includes information such as the number of iterations, the magnitude of parameter adjustments in each round, the range of search space changes, and the calculated convergence rate. Adaptation evaluation data from multiple iterations refers to the performance score or objective deviation value obtained after each round of optimization, used to measure the quality change trend of the solution set in each round. By simultaneously collecting both types of data, a time-series-based optimization behavior trajectory can be formed, thereby determining the evolution speed, directional stability, and performance bottlenecks of the stochastic search function at different stages. Evaluating the convergence phase based on iteration round data and adaptation evaluation data means analyzing the convergence characteristics of the stochastic search function to determine whether the optimization process has entered the convergence phase. Convergence phase evaluation identifies whether the algorithm has reached a stable state by calculating the performance improvement magnitude, objective function change rate, and parameter fluctuation range over several consecutive rounds. For example, if the fitness improvement magnitude is less than 0.01 for three consecutive rounds, the algorithm can be considered to have converged. After identifying convergence characteristics, stage constraints are generated to establish limitations for the current optimization stage, preventing the algorithm from getting bogged down in invalid calculations or over-adjustments.

[0051] Stochastic adjustment of the stochastic search function based on stage constraints means dynamically modifying the behavior of the stochastic search function during the convergence phase, allowing it to continue exploring potential better solutions while satisfying the stage constraints. Stochastic adjustment introduces controlled perturbations, i.e., adding small fluctuations to some constraint variables to prevent the stochastic search function from getting trapped in local optima. For example, in temperature regulation optimization, if the temperature tends to stabilize around 36.8 degrees Celsius, stochastic adjustment allows for slight searches within the range of 36.6 to 37.0 degrees Celsius to improve global optimization capabilities. Simultaneously, the bias weights of the first and second update adjustment biases are compensated, meaning that while performing stochastic adjustment, the weight ratios of the two types of update biases are reallocated to ensure a balance between the search direction and the learning trend. This compensation mechanism means that when the stochastic search introduces new perturbations, the parameter priority bias and the optimal solution learning bias also need to be recalibrated to prevent over-reliance on a single direction.

[0052] Furthermore, controlling the storage environment based on the search and optimization results means that after completing the search and optimization phase of the optimization algorithm, the cell storage environment will be automatically adjusted according to the obtained optimal parameter combination. The search and optimization results refer to a set of optimal environmental control parameters, such as the optimal setpoints for temperature, humidity, oxygen, and carbon dioxide concentrations, obtained after multiple rounds of evolutionary calculations, fitness assessments, and parameter screening. Reading the monitoring response through monitoring module 1 indicates that after environmental adjustment is performed, monitoring module 1 will be activated to read the adjustment results in real time to determine the actual adjustment effect. The monitoring module is a feedback system composed of a sensor array and a biosignal acquisition unit, used to detect immediate changes in environmental variables and cell responses. The monitoring response is the feedback data received after the adjustment action occurs, used to evaluate the effectiveness of environmental control and the changing trend of cell state.

[0053] The monitoring response includes the environmental change monitoring response and the fibroblast biological response. The environmental change monitoring response reflects the effectiveness of external control conditions, such as changes in temperature, humidity, and gas composition. The biological response reflects the cell's intrinsic physiological response to external changes, such as metabolic rate, changes in cell membrane potential, and the intensity of active fluorescence. For example, when humidity increases, the cell fluorescence activity increases by 15%, indicating that environmental adjustment has a positive impact on cell state.

[0054] Furthermore, a first interactive correction term is established based on the environmental change monitoring response and the search-optimization control results. This term represents the deviation correction amount calculated at the environmental control level by comparing the difference between the target control value of environmental regulation and the actual monitored environmental response. The search-optimization control results are a set of ideal control commands derived from the optimization algorithm, such as a desired temperature of 36.8 degrees Celsius, humidity of 85%, and oxygen concentration of 19%. By performing difference analysis on the two sets of data, the degree of deviation in environmental control can be quantified, thereby establishing the first interactive correction term. Essentially, the first interactive correction term is a control response deviation correction term used to compensate for execution errors during environmental regulation. For example, if the target humidity is 85%, but the humidity detected by the sensor is only 82%, a 3% humidity deviation correction term will be generated, automatically increasing the humidification intensity in the next round of regulation to bring the environment closer to the target value.

[0055] Based on the adaptation analysis of biological responses and environmental change monitoring responses, a second interactive correction term is established to represent the actual physiological responses of fibroblasts under this environment. Biological responses refer to the intrinsic state changes of cells under environmental changes, such as metabolic rate, cell membrane potential changes, biofluorescence intensity, and survival rate, reflecting the cell's tolerance and adaptability to the environment. By comparing the biological response with the environmental change monitoring response, it can be determined whether the current environmental change has caused a positive adaptation or a negative stress response. For example, when the ambient temperature rises from 36.5 degrees Celsius to 37 degrees Celsius, the cell activity signal increases by 8%, indicating that this regulatory direction is suitable for the cell's physiological needs, and a positive feedback correction term is established accordingly. Conversely, if cell activity decreases, a negative correction term is generated to reduce the intensity of temperature regulation. The second interactive correction term is a feedback correction term for biological adaptation, used to correct environmental control strategies, making the control behavior more consistent with the cell's physiological characteristics.

[0056] An interactive correction term is established based on the first and second interactive correction terms, representing the fusion of two types of correction information from the environmental control layer and the biological response layer to form a comprehensive feedback correction model. The interactive correction term is a fusion parameter structure that considers both the control deviation of the physical environment and the adaptive response of cells to the environment, thereby achieving bidirectional feedback and synergistic correction during the regulation process. For example, when the temperature and humidity control deviation is 3%, but cell adaptation feedback shows that the environment actually promotes cell proliferation, the deviation correction weight will be reduced to maintain the current environmental state; conversely, if the control deviation is small but the cell physiological response is poor, the correction magnitude will be amplified to prioritize cell stability. The interactive correction term is ultimately input into the control optimization module for real-time adjustment of the control strategy, achieving refined and dynamic environmental management.

[0057] Furthermore, this application also includes: an intelligent early warning module, used to create an early warning strategy based on the adjustment granularity identifier and the search optimization results, the early warning strategy including an adjustment deviation early warning strategy and an adjustment gradient early warning strategy, and using the early warning strategy to monitor and verify the monitoring module during the cell storage environment adjustment process, and to execute early warning output.

[0058] Specifically, an early warning strategy is created based on the regulation granularity identifier and the search optimization results. This indicates that during the control of the cell storage environment, the granularity identifier is needed to determine the level of fineness of control regulation. The regulation granularity identifier reflects the smallest controllable unit of the regulation parameter and is used to describe the sensitivity of the control. The search optimization results are the optimal control scheme obtained by calculating and comparing multiple combinations of control parameters, which can find the scheme that maximizes cell viability and stability among various possible regulation methods. Based on these two pieces of information, an early warning strategy can be automatically generated to identify potential risks in advance during the regulation process.

[0059] The early warning strategy includes an adjustment deviation early warning strategy and an adjustment gradient early warning strategy. The adjustment deviation early warning strategy is triggered when the actual monitored environmental parameters, such as temperature, humidity, or gas concentration, deviate from the target setpoint by more than a set threshold. The adjustment gradient early warning strategy is triggered when the adjustment rate or trend of change is too fast or too slow. These two types of early warnings not only detect static deviations but also dynamically sense changing trends, thus forming a dual-layer protection.

[0060] The monitoring module, which utilizes an early warning strategy to regulate the cell storage environment, was validated. In actual operation, monitoring module 1 collects environmental data in real time and compares it with standard values ​​and trend thresholds in the early warning strategy to determine if any abnormalities or potential risks exist. When a deviation exceeding the threshold or an abnormal gradient is detected, an early warning is triggered, i.e., a prompt signal or control command is activated, such as issuing a red warning on the display interface or automatically pausing the regulation process, to prevent adverse conditions from continuously affecting cell viability.

[0061] In summary, the fibroblast storage device provided in this application has the following technical effects: by achieving the technical goals of multimodal perception, dynamic prediction and intelligent adaptive regulation of the fibroblast storage environment, it can identify environmental deviations in real time, correct regulatory behavior in advance and continuously optimize storage conditions, thereby ensuring the long-term activity of fibroblasts and the operational stability of the storage system.

[0062] Example 2: Based on the same inventive concept as the fibroblast storage device in the foregoing examples, this application also provides a fibroblast storage method. Please refer to the appendix. Figure 2The process includes: performing fibroblast growth monitoring and establishing real-time cell state data; reading the storage tags of fibroblasts, dividing the cells into stages based on the storage tags and the real-time cell state data, establishing stage constraints, using the stage constraints and the real-time cell state data to predict the cell storage environment, and establishing a target prediction result, wherein the target prediction result is set with an adjustment granularity identifier; activating the sensor array, performing real-time acquisition of the fibroblast storage environment, and establishing a storage environment dataset, wherein the storage environment dataset includes temperature, humidity, and gas composition; using the storage environment dataset as the starting point of the environment, taking the target prediction result as the following target, performing environmental control search and optimization under the constraint of the adjustment granularity identifier, and performing intelligent adjustment of the storage environment based on the search and optimization results.

[0063] Furthermore, the fibroblast storage method further includes: after reading the stage constraints, constructing a stage-aware network based on the stage constraints, the stage-aware network including stage regulation channels mapped to the stage constraints; selecting and activating the corresponding stage regulation channel according to the current stage of the fibroblast, and synchronously inputting the environmental starting point, the target being followed, and the regulation granularity identifier into the activated stage regulation channel; and using the stage regulation channel to perform search optimization under a focusing strategy to establish search optimization results.

[0064] Furthermore, the fibroblast storage method further includes: configuring a focusing strategy using the stage regulation channel, wherein the focusing strategy is a sensitive environmental factor focusing strategy mapped to the stage; initializing and configuring an initial solution set based on the focusing strategy and regulation granularity identifier, and performing evolutionary simulation of the initial solution set in the stage regulation channel, wherein the evolutionary simulation is performed with the environmental starting point as the simulation starting point and the initial solution set as the evolutionary scheme, and the evolutionary process and the deviation between the evolutionary result and the target being followed are recorded; evaluating the fitness of the initial solution set based on the evolutionary process and the target deviation, and establishing fitness evaluation results; performing iterative updates of the initial solution set based on the fitness evaluation results and the focusing strategy to complete the search and optimization.

[0065] Furthermore, the fibroblast storage method further includes: after locating key environmental factors according to the focusing strategy, setting a first update adjustment bias for parameters based on the location results, the first update adjustment bias including parameter priority bias and adjustment coverage bias; selecting optimal solutions within the initial solution set according to the fitness evaluation results, and establishing a second update adjustment bias using the optimal solution selection results; configuring a random search function, performing a random search of the initial solution set using the random search function, and completing iterative updates based on the random search results, the first update adjustment bias, and the second update adjustment bias.

[0066] Furthermore, the fibroblast storage method further includes: during the iterative update process, collecting iteration round data and adaptation evaluation data from multiple iterations; performing convergence phase evaluation based on the iteration round data and adaptation evaluation data; generating phase constraints; performing random adjustment of the random search function based on the phase constraints; and simultaneously compensating the bias weights of the first update adjustment bias and the second update adjustment bias to complete the iterative update.

[0067] Furthermore, the fibroblast storage method further includes: controlling the adjustment of the storage environment according to the search and optimization results, and reading the monitoring response through a monitoring module, wherein the monitoring response includes the environmental change monitoring response and the biological response of the fibroblasts; constructing an interactive correction term using the monitoring response, and performing control optimization management of the search and optimization results.

[0068] Furthermore, the fibroblast storage method further includes: establishing a first interactive correction term based on the environmental change monitoring response and the search optimization control result, wherein the first interactive correction term is a control response deviation correction term; performing biological response adaptation analysis based on the biological response and the environmental change monitoring response to establish a second interactive correction term, wherein the second interactive correction term is a biological adaptation feedback correction term; and establishing an interactive correction term based on the first interactive correction term and the second interactive correction term.

[0069] Furthermore, the fibroblast storage method further includes: inputting the stage constraints and the real-time cell state data into a multimodal prediction model, using the temporal analysis layer in the multimodal prediction model to perform environmental prediction of the growth stage, and establishing the target prediction result.

[0070] Furthermore, the fibroblast storage method further includes: creating an early warning strategy based on the adjustment granularity identifier and the search optimization results, the early warning strategy including an adjustment deviation early warning strategy and an adjustment gradient early warning strategy, using the early warning strategy to monitor and verify the cell storage environment adjustment process by a monitoring module, and executing early warning output.

[0071] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The fibroblast storage device and specific examples in the foregoing Embodiment 1 are also applicable to the fibroblast storage method in this embodiment. Through the foregoing detailed description of a fibroblast storage device, those skilled in the art can clearly understand the fibroblast storage device and method in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.

[0072] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0073] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. A fibroblast storage device, characterized in that, The device includes: The monitoring module is used to perform fibroblast growth monitoring and establish real-time cell status data; The storage prediction module is used to read the storage tags of fibroblasts, divide the cells into stages based on the storage tags and the real-time cell state data, establish stage constraints, predict the cell storage environment using the stage constraints and the real-time cell state data, and establish a target prediction result, wherein the target prediction result is set with an adjustment granularity identifier. An environmental monitoring module is used to activate the sensor array, perform real-time acquisition of the storage environment of fibroblasts, and establish a storage environment dataset, which includes temperature, humidity, and gas composition. The intelligent adjustment module is used to take the storage environment dataset as the starting point, take the target prediction result as the target to follow, perform environmental control search and optimization under the constraint of adjustment granularity identifier, and perform intelligent adjustment of the storage environment according to the search and optimization results.

2. The fibroblast storage device as described in claim 1, characterized in that, In the intelligent adjustment module, environmental control search and optimization are performed under the constraints of the adjustment granularity identifier, including: After reading the stage constraints, a stage-aware network is constructed based on the stage constraints. The stage-aware network includes stage adjustment channels mapped to the stage constraints. Select and activate the corresponding stage regulation channel according to the current stage of the fibroblast, and simultaneously input the environmental starting point, the target being followed, and the regulation granularity identifier into the activated stage regulation channel. The search optimization under the focusing strategy is performed using the aforementioned stage adjustment channel to establish the search optimization results.

3. The fibroblast storage device as described in claim 2, characterized in that, In the intelligent adjustment module, the search and optimization under the focusing strategy is performed using the stage adjustment channel to establish the search and optimization results, including: The focus strategy is configured using the stage adjustment channel, and the focus strategy is a sensitive environmental factor focus strategy mapped to the stage; The initial solution set is configured based on the focusing strategy and the adjustment granularity identifier, and the evolution simulation of the initial solution set is performed in the stage adjustment channel. The evolution simulation is performed with the environmental starting point as the simulation starting point and the initial solution set as the evolution scheme. The evolution process and the deviation between the evolution results and the target being followed are recorded. The fitness of the initial solution set is evaluated based on the evolution process and the target deviation, and the fitness evaluation results are established. Based on the fitness evaluation results and the focusing strategy, the initial solution set is iteratively updated to complete the search optimization.

4. The fibroblast storage device as described in claim 3, characterized in that, In the intelligent adjustment module, the initial solution set is iteratively updated based on the fitness evaluation results and the focusing strategy, including: After locating key environmental factors according to the focusing strategy, a first update adjustment bias is set based on the location results. The first update adjustment bias includes a parameter priority bias and an adjustment coverage bias. Based on the fitness evaluation results, a best solution is selected from the initial solution set, and a second update adjustment bias is established using the best solution selection results. Configure a random search function, and after performing a random search on the initial solution set using the random search function, complete the iterative update based on the random search results, the first update adjustment bias, and the second update adjustment bias.

5. A fibroblast storage device as described in claim 4, characterized in that, During the iterative update process, iteration round data and adaptation evaluation data from multiple iterations are collected. Convergence phase evaluation is performed based on the iteration round data and adaptation evaluation data to generate phase constraints. Based on the phase constraints, the random search function is randomly adjusted, and the bias weights of the first update adjustment bias and the second update adjustment bias are compensated simultaneously to complete the iterative update.

6. The fibroblast storage device as described in claim 1, characterized in that, The intelligent adjustment module intelligently adjusts the storage environment based on the search optimization results, including: The storage environment is adjusted according to the search and optimization results, and the monitoring response is read through the monitoring module. The monitoring response includes the environmental change monitoring response and the biological response of fibroblasts. The monitoring response is used to construct interactive correction items, and the control and optimization management of the search results is performed.

7. A fibroblast storage device as described in claim 6, characterized in that, The interactive correction item is constructed using the monitoring response, including: A first interactive correction term is established based on the environmental change monitoring response and the search optimization control results of the monitoring response. The first interactive correction term is a control response deviation correction term. Based on the biological response and the environmental change monitoring response, an adaptation analysis of the biological response is performed to establish a second interactive correction term, which is a feedback correction term for biological adaptation. An interaction correction item is created based on the first interaction correction item and the second interaction correction item.

8. A fibroblast storage device as described in claim 1, characterized in that, In the storage prediction module, the cell storage environment is predicted using the stage constraints and the real-time cell state data to establish a target prediction result, including: The stage constraints and the real-time cell state data are input into the multimodal prediction model. The time-series analysis layer in the multimodal prediction model is used to perform environmental prediction of the growth stage and establish the target prediction result.

9. A fibroblast storage device as described in claim 1, characterized in that, The device further includes: The intelligent early warning module is used to create early warning strategies based on the adjustment granularity identifier and the search optimization results. The early warning strategies include adjustment deviation early warning strategies and adjustment gradient early warning strategies. The monitoring module uses the early warning strategies to monitor and verify the cell storage environment adjustment process and execute early warnings.

10. A method for storing fibroblasts, characterized in that, Performed by any one of the fibroblast storage devices according to claims 1 to 9, comprising: Perform fibroblast growth monitoring and establish real-time cell status data; Read the storage tag of fibroblasts, divide the cells into stages according to the storage tag and the real-time cell state data, establish stage constraints, use the stage constraints and the real-time cell state data to predict the cell storage environment, and establish a target prediction result, wherein the target prediction result is set with an adjustment granularity identifier. The sensor array is activated to perform real-time acquisition of the storage environment of fibroblasts and establish a storage environment dataset, which includes temperature, humidity, and gas composition. Using the aforementioned storage environment dataset as the starting point, the target prediction result is used as the target to be followed. Under the constraint of adjusting the granularity identifier, the search and optimization of the environment control is performed, and the storage environment is intelligently adjusted according to the search and optimization results.

Images