Intelligent Adjustment Method and System for Fluid Control of Microfluidic Chip

Abstract

The present invention relates to the field of fluid control technology, and discloses a fluid control intelligent adjustment method and system for a microfluidic chip. The method includes: collecting data of an adaptive fluid controller, a microfluidic channel pressure sensor array, a temperature sensor array and a flow sensor array based on the microfluidic chip to obtain real-time fluid parameter monitoring data; performing feature extraction to obtain a multi-dimensional fluid feature matrix; establishing a three-dimensional fluid model of the microfluidic channel and solving to obtain flow field distribution data; inputting the flow field distribution data into a Kriging surrogate model, establishing a mapping model between the channel width, depth, length and fluid control performance, and calculating to obtain an optimal geometric parameter combination; performing multi-level calculations on the fluid pressure uniformity, temperature distribution uniformity and flow rate stability to obtain a hierarchical fluid parameter control strategy. The present invention effectively solves the problem of multi-physical field coupling, reduces the computational complexity of the control system, and improves the execution efficiency of the control strategy.

Description

Technical Field

[0001] The present invention relates to the technical field of fluid control, and particularly to a method and system for intelligent adjustment of fluid control in a microfluidic chip. Background Art

[0002] Due to the complex fluid flow characteristics in the microfluidic channel, traditional fluid control methods are difficult to meet the requirements of high precision and high stability. Especially when facing dynamic changes under different working conditions, the control effect is poor.

[0003] Currently, fluid control mainly relies on fixed control strategies, lacking the ability of real-time monitoring and dynamic adjustment of fluid parameters. At the same time, there are complex coupling relationships among the pressure distribution, temperature field and flow rate change in the microfluidic channel, which makes it difficult for a single control method to ensure the accuracy and stability of fluid control. Summary of the Invention

[0004] The present invention provides a method and system for intelligent adjustment of fluid control in a microfluidic chip. The present invention effectively solves the problem of multi-physical field coupling, reduces the computational complexity of the control system, and improves the execution efficiency of the control strategy.

[0005] In a first aspect, the present invention provides a method for intelligent adjustment of fluid control in a microfluidic chip. The method for intelligent adjustment of fluid control in the microfluidic chip includes:

[0006] Collecting data of an adaptive fluid controller, a microfluidic channel pressure sensor array, a temperature sensor array and a flow sensor array based on the microfluidic chip to obtain real-time fluid parameter monitoring data;

[0007] Inputting the real-time fluid parameter monitoring data into a dynamic self-learning control network model for feature extraction to obtain a multi-dimensional fluid feature matrix;

[0008] Establishing a three-dimensional fluid model of the microfluidic channel within a set pressure range and temperature range, and solving to obtain flow field distribution data;

[0009] Inputting the flow field distribution data into a Kriging surrogate model, establishing a mapping model between the channel width, depth, length and fluid control performance, and calculating the optimal geometric parameter combination of the microfluidic channel through a non-dominated sorting genetic algorithm;

[0010] According to the multi-dimensional fluid feature matrix and the optimal geometric parameter combination, performing multi-level calculations on the fluid pressure uniformity, temperature distribution uniformity and flow rate stability to obtain a hierarchical fluid parameter control strategy.

[0011] Second aspect, the present invention provides a fluid control intelligent regulation system for a microfluidic chip, and the fluid control intelligent regulation system for the microfluidic chip includes:

[0012] A data acquisition module, configured to perform data acquisition on an adaptive fluid controller, a microfluidic channel pressure sensor array, a temperature sensor array, and a flow rate sensor array based on the microfluidic chip, so as to obtain real-time fluid parameter monitoring data;

[0013] A feature extraction module, configured to input the real-time fluid parameter monitoring data into a dynamic self-learning control network model for feature extraction, so as to obtain a multi-dimensional fluid feature matrix;

[0014] A solution module, configured to establish a three-dimensional fluid model of the microfluidic channel within a set pressure range and temperature range, and solve to obtain flow field distribution data;

[0015] An establishment module, configured to input the flow field distribution data into a Kriging surrogate model, establish a mapping model between the channel width, depth, length, and fluid control performance, and calculate an optimal geometric parameter combination of the microfluidic channel through a non-dominated sorting genetic algorithm;

[0016] A calculation module, configured to perform multi-level calculations on the fluid pressure uniformity, temperature distribution uniformity, and flow rate stability according to the multi-dimensional fluid feature matrix and the optimal geometric parameter combination, so as to obtain a hierarchical fluid parameter control strategy.

[0017] The third aspect of the present invention provides a computer device, including: a memory and at least one processor, wherein instructions are stored in the memory; the at least one processor calls the instructions in the memory, so that the computer device executes the above-mentioned fluid control intelligent regulation method for the microfluidic chip.

[0018] The fourth aspect of the present invention provides a computer-readable storage medium, wherein instructions are stored in the computer-readable storage medium, and when the instructions are run on a computer, the computer is made to execute the above-mentioned fluid control intelligent regulation method for the microfluidic chip.

[0019] In the technical solution provided by the present invention, the present invention adopts a combination of an adaptive fluid controller and a multi-type sensor array to achieve high-precision real-time monitoring of fluid parameters in a microfluidic channel, improving the accuracy of data acquisition; by constructing a dynamic self-learning control network model, it realizes the adaptive extraction of fluid characteristics and enhances the adaptability of the system to different working conditions; innovatively combines the Kriging surrogate model with the non-dominated sorting genetic algorithm to establish an accurate prediction model of fluid control performance and optimize control parameters; the hierarchical fluid parameter control strategy proposed by the present invention realizes the collaborative optimization of three control layers of pressure, temperature, and flow through multi-level calculations, improving the control accuracy; uses a hierarchical optimization method for inter-layer coordination, effectively solves the problem of multi-parameter coupling, and ensures the stability of the control system. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0021] Figure 1 It is a schematic diagram of the steps of the intelligent adjustment method for fluid control of a microfluidic chip in an embodiment of the present invention;

[0022] Figure 2 It is a schematic diagram of the structure of the intelligent adjustment system for fluid control of a microfluidic chip in an embodiment of the present invention;

[0023] Figure 3 It is a schematic block diagram of the structure of a computer device in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0024] The embodiments of the present invention provide an intelligent adjustment method and system for fluid control of a microfluidic chip. The terms "first", "second", "third", "fourth", etc. (if any) in the specification, claims, and above-mentioned drawings of the present invention are used to distinguish similar objects and do not necessarily describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments described here can be implemented in an order different from that shown or described here. In addition, the terms "comprising" or "having" and any variation thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0025] For ease of understanding, the specific process of the embodiment of the present invention will be described below. Please refer to Figure 1 , an embodiment of the intelligent adjustment method for fluid control of the microfluidic chip in the embodiment of the present invention includes:

[0026] Step S1: Based on the microfluidic chip, data collection is performed on the adaptive fluid controller, the microfluidic channel pressure sensor array, the temperature sensor array, and the flow sensor array to obtain real-time fluid parameter monitoring data;

[0027] It can be understood that the execution subject of the present invention can be the intelligent adjustment system for fluid control of the microfluidic chip, or a terminal or a server. Specifically, it is not limited here. In the embodiment of the present invention, the server is taken as the execution subject for illustration.

[0028] Specifically, according to the structural characteristics of the microfluidic channel, fine grid division is carried out. The grid division is based on the geometric shape of the microfluidic channel, and the channel area is divided into multiple small units. The nodes of these units are designed as the layout positions of the sensors, forming a monitoring point distribution network composed of a pressure sensor array. This network structure can cover all key areas of the channel, ensuring that the pressure sensor array can collect accurate pressure data in real time. At the same time, based on the physical characteristics of the microfluidic channel, key points are calibrated at its inlet and outlet positions, channel corners, and shunt nodes, and a temperature sensor array is arranged at these calibration point positions. The temperature sensor array can collect the real-time distribution data of the temperature in the microfluidic channel in matrix form, forming a temperature monitoring node matrix data set. At the same time, flow monitoring points are set at the inlet and outlet ends of the microfluidic channel, and a flow sensor array is arranged at these positions to ensure that the flow rate and flow volume changes of the fluid can be monitored and the fluid flow parameters can be obtained. To optimize the accuracy of the monitoring data, based on the reference fluid chamber of the adaptive fluid controller, the fluid flow area is intelligently partitioned. The intelligent partition takes the pressure characteristics as the core, and divides the entire fluid area into multiple reference pressure monitoring point arrays. These reference pressure monitoring points work in cooperation with the distribution network of the pressure sensor array. By comparing the actual data of the pressure monitoring points with the reference pressure data in real time, the real-time pressure difference is calculated, forming a pressure gradient distribution sequence, which reflects the pressure distribution characteristics in the microfluidic channel. The pressure gradient distribution sequence, the temperature monitoring node matrix data, and the fluid flow parameters provided by the flow sensor are comprehensively analyzed and fused. The data fusion is realized through a specific algorithm model, and the methods include weighted fusion, multi-source data calibration, and feature extraction of deep learning models, etc. Through the fusion process, the data deviation generated by a single sensor is eliminated, and the accuracy and comprehensiveness of the monitoring data are improved. The fusion process combines the dynamic characteristics of the microfluidic chip to perform synchronous correction on different data types on the time axis, and obtains a high-dimensional data set containing real-time fluid state information, that is, real-time fluid parameter monitoring data.

[0029] Step S2: Input the real-time fluid parameter monitoring data into the dynamic self-learning control network model for feature extraction to obtain a multi-dimensional fluid feature matrix;

[0030] Specifically, the pressure gradient distribution sequence collected in real time is input into the adaptive fluid control module in the dynamic self-learning control network model for processing. The core of the adaptive fluid control module lies in its included pressure feature extraction network and pressure parameter optimization network. The pressure feature extraction network consists of three convolutional layers and two fully connected layers. Through deep convolutional operations on the pressure gradient distribution sequence, it can extract complex pressure characteristics in the microfluidic channel, and these characteristics reflect the subtle changes in the fluid pressure distribution in the channel. The multi-level feature extraction process of the convolutional layer can effectively capture the local characteristics of the pressure gradient distribution, and at the same time, through the fully connected layer, global aggregation of features is achieved to generate a pressure control feature sequence. In the pressure parameter optimization network, the extracted pressure features are processed through two fully connected layers and one output layer, and the pressure characteristics are optimized and parameterized to ensure that the obtained pressure control feature sequence has good interpretability and efficiency. At the same time, the temperature monitoring node matrix data in the real-time fluid parameter monitoring data is input into the spatio-temporal feature extraction module in the dynamic self-learning control network model for processing. This module includes two parts: a time series feature network and a spatial feature network, and through three-dimensional convolutional operations, efficient extraction of temperature distribution features is achieved. The time series feature network adopts a long short-term memory network structure (LSTM), which can capture the dynamic change rules of temperature data in the time dimension and serialize these time characteristics to generate time series features of temperature changes. At the same time, the spatial feature network captures the local features of the temperature distribution in the spatial dimension through three-dimensional convolutional operations, and this spatial feature reflects the change rules of the temperature distribution in different regions of the microfluidic chip. Through the coordinated action of the time series feature network and the spatial feature network, the temperature monitoring node matrix data is efficiently deconstructed into a temperature distribution feature sequence, which not only contains multi-dimensional information of time and space, but also can reflect the comprehensive characteristics of heat conduction and temperature fluctuations in the chip. The fluid flow parameters of the real-time fluid parameter monitoring data are input into the parameter self-adjustment module in the dynamic self-learning control network model. This module consists of a flow prediction network and a parameter correction network, and through a dynamic weight adjustment mechanism, high-precision modeling of flow change characteristics is achieved. The role of the flow prediction network is to model and predict the fluid flow parameters, and use real-time monitoring data to generate prediction results on the future flow change trend. At the same time, the parameter correction network dynamically adjusts the prediction results to correct existing errors and generate a flow change feature sequence with high credibility, reflecting the flow velocity fluctuations and flow stability changes of the fluid under different working conditions. Through the dynamic self-learning control network model, feature fusion is performed on the pressure control feature sequence, the temperature distribution feature sequence, and the flow change feature sequence to generate a fusion feature matrix. This matrix is optimized through time series reconstruction operations, and in the reconstruction process, the continuity and dynamic change rules in the time dimension are utilized to integrate the fusion feature matrix into a multi-dimensional fluid feature matrix.

[0031] Step S3, establishing a three-dimensional fluid model of the microfluidic channel within a set pressure range and temperature range, and solving to obtain flow field distribution data;

[0032] Specifically, the pressure range and temperature range of the microfluidic channel are determined. These boundary conditions determine the physical constraints of the model and directly affect the flow characteristics of the fluid. On this basis, three-dimensional solid modeling is performed according to the geometric structure of the microfluidic channel. The model includes key areas of the channel, such as the feed end, the discharge end, the corners, and the diversion nodes. The modeling process uses CAD tools to ensure that the geometric shape can reflect the characteristics of the actual microfluidic chip. Hexahedral structured grid division is performed for the three-dimensional fluid model. Grid division needs to balance computational efficiency and accuracy. Conventional grids are used in general areas of the channel, while grid encryption methods are used in complex flow areas such as inlets and outlets, corners, and diversion nodes. By refining the grids of these key areas, the complexity of local fluid behavior can be effectively captured and the accuracy of the calculation can be improved. After the grid division is completed, the mass conservation equation, momentum conservation equation, and energy conservation equation are respectively applied to the computational domain grid, and the finite volume method is used to discretize these conservation equations. The finite volume method adapts to numerical calculations by converting continuous equations into discrete forms. Specifically, the second-order upwind scheme is used for spatial discretization to improve the stability and accuracy of the calculation. The control equations formed after discretization are the basis of fluid mechanics calculations and can describe the physical changes of the fluid in the channel. The discrete control equations are solved based on the SIMPLE algorithm. The SIMPLE algorithm is an iterative algorithm that can effectively deal with the pressure-velocity coupling problem in fluid calculations. By initially guessing the flow field value, the pressure and velocity fields are gradually corrected to obtain the initial flow field solution. In order to optimize the accuracy and efficiency of the calculation, the number of grids is adjusted based on the preliminary flow field solution, that is, the grid density is increased in the area where the fluid behavior changes drastically, and the number of grids is appropriately reduced in the area where the flow changes slowly to generate the optimal grid distribution. The initial value of the flow field is iteratively solved on the optimal grid distribution. The steady-state flow field solution is obtained by stepwise approximation. The steady-state flow field solution is the final fluid state of the system under given boundary conditions, reflecting the stable distribution of the pressure field, velocity field and temperature field. From this steady-state solution, detailed data of the fluid are extracted, including pressure distribution, temperature gradient and flow velocity distribution. These data together constitute the flow field distribution data, which describes the fluid behavior in the microfluidic channel.

[0033] Step S4, inputting the flow field distribution data into the Kriging proxy model, establishing a mapping model between channel width, depth, length and fluid control performance, and calculating the optimal geometric parameter combination of the microfluidic channel by a non-dominated sorting genetic algorithm;

[0034] Specifically, the flow field distribution data are used as sample points. These data contain the distribution information of the pressure field, temperature field, and flow field in the microfluidic channel, and can comprehensively reflect the dynamic characteristics of the fluid. On this basis, for the geometric parameter range of the microfluidic channel, including the width range, depth range, and length range, the Latin hypercube sampling method is used to generate the training sample space. Latin hypercube sampling can evenly distribute the sampling points in the high-dimensional space, ensuring coverage of the entire geometric parameter range, while avoiding the overlap and aggregation of sampling points, providing a diverse and efficient sample set for the subsequent training of the model. Based on the generated training sample space, a Kriging surrogate model is constructed. This model establishes the mapping relationship between geometric parameters and fluid performance through the interpolation method. The core of the Kriging surrogate model lies in using the Gaussian correlation function as the kernel function, setting the relevant parameter θ to 0.1, and setting the order of the regression function to 2, thereby enhancing the model's fitting ability for nonlinear characteristics. By calculating the correlation matrix of geometric parameters, the mutual influence relationship between different parameters can be captured. On this basis, the maximum likelihood estimation function is used to optimize the hyperparameters to determine the optimal interpolation weights of the model. The optimization result of the interpolation weights directly determines the prediction accuracy of the Kriging surrogate model. By substituting it into the model, the mapping model construction between the channel width, depth, and length and the fluid control performance index is completed. Using the established mapping model, the non-dominated sorting genetic algorithm is used for optimization and solution. The non-dominated sorting genetic algorithm is a multi-objective optimization method that can search for the optimal combination of geometric parameters while meeting multiple performance requirements. During the optimization process, an initial population is generated, the fitness of the individuals in the population is evaluated based on the evolutionary process, and the Pareto optimal solutions are calculated according to the multi-objective optimization criteria. To ensure the comprehensiveness and stability of the optimization process, the crowding distance is used to sort the individuals, and the individuals with a non-dominated rank of 1 are selected as the elite population to form the Pareto front. The solution set in the Pareto front represents the geometric parameter combinations that meet the multi-objective optimization conditions, and these combinations achieve the optimal trade-off among performance indicators such as pressure uniformity, temperature uniformity, and flow stability. To select the optimal solution from the Pareto front, the solution set is normalized. By mapping different performance indicators to a unified scale, it is convenient to compare the multi-objective performance. After normalization, candidate parameter combinations are obtained, and these combinations all meet the requirements of the microfluidic channel optimization design. To select the optimal parameter combination from them, sensitivity analysis is carried out. By analyzing the influence degree of each geometric parameter on the fluid control performance, the stability and reliability of the parameter combination are comprehensively evaluated. Based on the comprehensive evaluation value, the optimal geometric parameter combination is selected as the design scheme of the microfluidic channel.

[0035] Construct a regression function matrix based on the training sample space. To enhance the model's feature expression ability for sample points, a second-order polynomial basis function is used for expansion. The second-order polynomial basis function expansion can capture the non-linear features of the input sample points. By analytically solving the regression matrix, the coefficients of the regression model are obtained, reflecting the basic influence of the input geometric parameters on the response target. Substitute the coefficients of the regression model into the Gaussian correlation function, and through the calculation method of the kernel function, evaluate the spatial correlation between sample points. The Gaussian correlation function takes the distance between sample points as a variable. By setting the correlation parameter θ to 0.1, it ensures that the kernel function has an appropriate decay characteristic for the correlation of sample points. This decay characteristic can effectively characterize the variation law of the input geometric parameters in the spatial distribution and generate a correlation function matrix accordingly. The correlation function matrix is the core description of the spatial dependence between sample points. To simplify the processing process of high-dimensional data, eigenvalue decomposition is performed on this matrix. The dimensionality reduction maps the original high-dimensional data to a low-dimensional feature space by performing eigenvalue decomposition on the correlation function matrix, while retaining the main variation information of the data. The obtained low-dimensional feature space can significantly reduce the computational complexity while ensuring information integrity. Construct a maximum likelihood estimation function based on the low-dimensional feature space. The maximum likelihood estimation function aims to find the optimal values of the hyperparameters through optimization, and its goal is to minimize the prediction error of the Kriging model. For the convenience of numerical calculation and optimization, this function adopts a logarithmic transformation form, transforming the complex non-linear optimization problem into a form more suitable for numerical solution. By inputting the optimization objective function into the pattern search algorithm, the optimal hyperparameters are efficiently found. The pattern search algorithm is a global optimization method that does not require gradient information and can find the global optimal solution in complex multi-modal optimization problems, thus ensuring the reliability and stability of the obtained hyperparameters. After obtaining the optimal hyperparameters, verify the accuracy of the model. Through the cross-validation method, the model is trained and tested multiple times to evaluate the performance of the Kriging surrogate model on different sample sets. Cross-validation can effectively avoid the problem of model overfitting and generate model accuracy evaluation indicators such as mean squared error (MSE), coefficient of determination ( ), etc. These indicators provide a quantitative basis for model correction. On this basis, the Kriging surrogate model is corrected according to the accuracy evaluation indicators to ensure that its prediction ability meets the requirements of practical applications. Calculate the optimal interpolation weights of the Kriging model by the generalized least squares method. The generalized least squares method combines the coefficients of the regression model, the correlation function matrix, and the optimized hyperparameters to improve the accuracy of the interpolation results while ensuring the stability of the model. The finally obtained interpolation weights are the core of the Kriging surrogate model, which can perform high-precision response prediction on new input samples, thus providing an accurate performance mapping for the geometric parameter optimization of microfluidic chips.

[0036] Step S5: According to the multi-dimensional fluid characteristic matrix and the optimal geometric parameter combination, perform multi-level calculations on the fluid pressure uniformity, temperature distribution uniformity, and flow rate stability to obtain a hierarchical fluid parameter control strategy.

[0037] Specifically, by performing correlation analysis on the pressure control characteristic sequence, temperature distribution characteristic sequence, and flow rate change characteristic sequence in the multi-dimensional fluid characteristic matrix with the optimal geometric parameter combination respectively, the quantitative relationships between these fluid characteristics and geometric parameters are revealed. The correlation analysis is based on statistical and machine learning methods, mapping the relationship between the characteristic matrix and geometric parameters into a multi-dimensional space to form a characteristic-parameter mapping relationship, reflecting the contribution degree and sensitivity of geometric parameters to fluid control performance. Input the characteristic-parameter mapping relationship into the control strategy generation module to construct an initial hierarchical control framework including a pressure control layer, a temperature control layer, and a flow rate control layer. Decompose the complex fluid control problem into sub-problems that can be independently optimized. By handling pressure, temperature, and flow rate control in a hierarchical manner, the complexity of optimization is significantly simplified, while the pertinence and efficiency of each layer of control are improved. In the pressure control layer, perform dynamic programming on the pressure data based on the initial hierarchical control framework. Dynamic programming decomposes the overall optimization problem into sub-problems at multiple stages and solves them through a recursive method, making the decision at each stage optimal, thereby ensuring that the overall pressure uniformity reaches the best level. Thus, generate pressure uniformity control parameters as the optimized output of the pressure layer. In the temperature control layer, optimize the temperature data based on the initial hierarchical control framework using model predictive control (MPC). MPC is based on the real-time prediction of temperature distribution. By solving the rolling optimization problem, calculate the control parameters that can maximize the temperature distribution uniformity. Its core lies in real-time predicting the temperature change trend and dynamically adjusting the control strategy according to the prediction results to effectively address the temperature fluctuation problem in the microfluidic system and generate temperature distribution uniformity control parameters. At the same time, in the flow rate control layer, perform robust control calculations on the flow rate data based on the initial hierarchical control framework. The goal of robust control is to handle the uncertainty of flow rate changes and ensure the flow rate stability of the fluid under different operating conditions. Through the robust control method, the control strategy can be optimized to minimize the flow rate fluctuation in the presence of disturbances and modeling errors, and generate flow rate stability control parameters. Input the pressure uniformity control parameters, temperature distribution uniformity control parameters, and flow rate stability control parameters into the hierarchical coordinator. The role of the hierarchical coordinator is to achieve coordination and consistency among layers through hierarchical optimization. Hierarchical optimization is based on the dependency relationship between layers, gradually transmitting the optimization results from the top-level decision downwards, and at the same time transmitting the feedback information from the lower layer to the upper layer to ensure the coordination and optimality of the overall system performance. During the process of hierarchical optimization, the output parameters of each control layer are dynamically adjusted to meet the global optimization goal, and finally generate a hierarchical fluid parameter control strategy.

[0038] In the embodiments of the present invention, the present invention adopts a combination of an adaptive fluid controller and a multi-type sensor array to achieve high-precision real-time monitoring of fluid parameters in a microfluidic channel, improving the accuracy of data acquisition; by constructing a dynamic self-learning control network model, it realizes the adaptive extraction of fluid characteristics and enhances the adaptability of the system to different working conditions; innovatively combines the Kriging surrogate model with the non-dominated sorting genetic algorithm to establish an accurate prediction model of fluid control performance and optimize control parameters; the hierarchical fluid parameter control strategy proposed by the present invention realizes the collaborative optimization of the pressure, temperature, and flow control layers through multi-level calculations, improving the control accuracy; uses a hierarchical optimization method for inter-layer coordination, effectively solving the multi-parameter coupling problem and ensuring the stability of the control system.

[0039] In a specific embodiment, the process of executing step S1 may specifically include the following steps:

[0040] Perform mesh division on the microfluidic channel. At the same time, arrange the pressure sensor array at the grid nodes to obtain a pressure monitoring point distribution network; calibrate key points at the inlet, outlet, corners, and split nodes of the microfluidic channel, and arrange the temperature sensor array at the calibrated point positions to collect the temperature monitoring node matrix data; set flow monitoring points at the feed end and discharge end of the microfluidic channel, and arrange the flow sensor array at the monitoring points to collect the fluid flow parameters.

[0041] Intelligently partition the reference fluid chamber of the adaptive fluid controller to obtain a reference pressure monitoring point array, and perform real-time difference calculation on the pressure data of the pressure monitoring point distribution network and the pressure data of the reference pressure monitoring point array to obtain a pressure gradient distribution sequence.

[0042] Based on the microfluidic chip, perform data fusion on the pressure gradient distribution sequence, temperature monitoring node matrix data, and fluid flow parameters to obtain real-time fluid parameter monitoring data.

[0043] Specifically, perform mesh division on the microfluidic channel. Divide the internal geometric structure of the channel into small units to more finely describe the fluid behavior. During this process, arrange the pressure sensor array on each grid node, which are key positions after mesh division. Assume that the geometric area of the channel is divided into grid cells, and the central position of each cell is defined as the grid node position. For each node , the pressure sensor records the local pressure value , a three-dimensional pressure monitoring point distribution network is formed. This network can capture the global pressure distribution within the channel and reflect the local pressure gradient changes through the differences between nodes. After completing the mesh generation and pressure sensor layout, key positions of the microfluidic channel are calibrated, including the inlet, outlet, corners, and splitting nodes of the channel. These positions are areas where significant changes in fluid behavior occur, such as pressure mutations, changes in flow velocity direction, or changes in flow rate distribution. After calibrating these key points ( ), temperature sensors are arranged at the calibrated point positions to collect temperature data . These temperature data can reflect the heat conduction characteristics and temperature change laws of the fluid within the key areas, forming a temperature monitoring node matrix data set. At the same time, flow monitoring points are set at the inlet and outlet ends of the microfluidic channel. Assuming there are monitoring points at the inlet end and monitoring points at the outlet end, the flow data of each monitoring point are denoted as and . By arranging a flow sensor array, real-time flow parameters are recorded, thereby accurately grasping the flow distribution and stability of the channel. For a single monitoring point, its flow data is obtained through integration:

[0044] ;

[0045] wherein, is the flow velocity vector, is the normal vector of the monitoring point cross-section, is the cross-sectional area of the monitoring point. Similarly, the flow data at the outlet end is calculated in the same way. Based on the reference fluid chamber of the adaptive fluid controller, the channel is intelligently partitioned. The reference area is divided according to the fluid control requirements to generate a reference pressure monitoring point array. Assuming the reference fluid chamber is divided into sub-regions, the central position of each sub-region is defined as the reference pressure monitoring point and its reference pressure value is . By calculating the difference between the data of the pressure monitoring point distribution network and the reference pressure value, a pressure gradient distribution sequence is obtained:

[0046] ;

[0047] wherein, represents the pressure gradient at the th grid node. The gradient sequence reflects the direction and intensity of the pressure change within the channel.

[0048] In a specific embodiment, the process of executing step S2 may specifically include the following steps:

[0049] The pressure gradient distribution sequence of the real-time fluid parameter monitoring data is input into the adaptive fluid control module in the dynamic self-learning control network model for processing to obtain a pressure control feature sequence. The adaptive fluid control module includes a pressure feature extraction network and a pressure parameter optimization network; the pressure feature extraction network includes three convolutional layers and two fully connected layers, and the pressure parameter optimization network includes two fully connected layers and one output layer;

[0050] The temperature monitoring node matrix data of the real-time fluid parameter monitoring data is input into the spatio-temporal feature extraction module in the dynamic self-learning control network model for three-dimensional convolution operation to obtain a temperature distribution feature sequence. The spatio-temporal feature extraction module includes a temporal feature network and a spatial feature network. The temporal feature network adopts a long short-term memory network structure, and the spatial feature network adopts a three-dimensional convolution structure;

[0051] The fluid flow parameters of the real-time fluid parameter monitoring data are input into the parameter self-regulation module in the dynamic self-learning control network model for dynamic weight adjustment to obtain a flow rate change feature sequence. The parameter self-regulation module includes a flow rate prediction network and a parameter correction network;

[0052] Feature fusion is performed on the pressure control feature sequence, the temperature distribution feature sequence, and the flow rate change feature sequence to obtain a fusion feature matrix, and temporal reconstruction is performed on the fusion feature matrix to obtain a multi-dimensional fluid feature matrix.

[0053] Specifically, the pressure gradient distribution sequence in the real-time fluid parameter monitoring data is input into the adaptive fluid control module in the dynamic self-learning control network model for processing. This module includes a pressure feature extraction network and a pressure parameter optimization network, and its function is to extract key features from the input pressure gradient distribution sequence and optimize these features to generate a pressure control feature sequence. In the pressure feature extraction network, the data is processed through three convolutional layers and two fully connected layers. Assume that the input pressure gradient distribution sequence is , where represents the grid node index, represents the time dimension. The convolution operation extracts local spatial features through the weight convolution kernel , and the formula is:

[0054] ;

[0055] Among them, is the output of the th convolutional layer, is the activation function, is the bias term, is the local index of the convolution kernel. After three layers of convolution, the feature data is input into two fully connected layers for global feature aggregation. The output of the fully connected layer is expressed as:

[0056] ;

[0057] Among them, is the weight matrix of the fully connected layer, is the feature input of the previous layer, is the bias term. The pressure feature extraction network outputs a high-dimensional pressure feature vector . The pressure feature vector is input into the pressure parameter optimization network, which includes two fully connected layers and one output layer. Through weight optimization and non-linear mapping, a pressure control feature sequence is obtained, and this sequence can reflect the key change characteristics of the pressure in the microfluidic channel. At the same time, the temperature monitoring node matrix data in the real-time fluid parameter monitoring data is input into the spatio-temporal feature extraction module of the dynamic self-learning control network model. This module includes a temporal feature network and a spatial feature network, which process the dynamic changes in the time dimension and the spatial distribution characteristics respectively. In the spatial feature network, a three-dimensional convolution structure is used to extract the spatial distribution features of the temperature. The convolution calculation is similar to that of the pressure feature extraction network, but in the three-dimensional convolution, each voxel of the input data is convolved with the three-dimensional convolution kernel to generate a spatial feature map. In the temporal feature network, a long short-term memory network (LSTM) structure is used to model the features in the time dimension. Assuming that the temperature matrix of the time series is , the calculation formula of the LSTM is:

[0058] ;

[0059] Among them, is the hidden state at time , is the weight matrix of the LSTM, is the bias term. Through the processing of this network, a temperature distribution feature sequence is generated, which can capture the dynamic characteristics of the temperature change. The fluid flow parameters are input into the parameter self-regulation module, which includes a flow prediction network and a parameter correction network. The flow prediction network establishes a prediction model based on historical flow data to generate an estimated value of the future flow change:

[0060] ;

[0061] Among them, is the weight matrix of the prediction network, is the bias term. The parameter correction network dynamically adjusts the weights of the prediction value to generate a flow change feature sequence to reflect the stability of the fluid flow. The pressure control feature sequence, the temperature distribution feature sequence, and the flow change feature sequence are feature fused to generate a fused feature matrix 。Feature fusion is achieved through weighted summation:

[0062] ;

[0063] Among them, are the weight coefficients of the pressure, temperature, and flow rate features respectively. The fused feature matrix generates a multi-dimensional fluid feature matrix through a time series reconstruction operation , fusing the features of different time steps into a unified high-dimensional space to provide a comprehensive description of the fluid state.

[0064] In a specific embodiment, the process of executing step S3 may specifically include the following steps:

[0065] Determine the pressure range and temperature range of the microfluidic channel, and perform three-dimensional solid modeling on the microfluidic channel to obtain a three-dimensional fluid model including a feed end, a discharge end, a corner, and a splitting node;

[0066] Perform hexahedral structured grid division on the three-dimensional fluid model, and encrypt the grids at the inlet, outlet, corners, and splitting nodes to obtain a computational domain grid;

[0067] Discretize the mass conservation equation, momentum conservation equation, and energy conservation equation on the computational domain grid, and perform spatial discretization using a second-order upwind scheme to obtain a discrete control equation set;

[0068] Solve the discrete control equation set based on the SIMPLE algorithm to obtain the initial flow field value, and encrypt and coarsen the number of grids based on the initial flow field value to obtain the optimal grid distribution;

[0069] Iteratively solve the initial flow field value on the optimal grid distribution to obtain the steady-state flow field solution, and extract the fluid pressure field, temperature field, and velocity field data based on the steady-state flow field solution to obtain the flow field distribution data.

[0070] Specifically, determine the pressure range and temperature range of the microfluidic channel. These boundary conditions determine the physical properties of the fluid and the constraints in the solution process. Set the pressure range as , and the temperature range as , where and respectively represent the lowest and highest pressures in the channel, and represent the lowest and highest temperatures respectively. Three-dimensional solid modeling is carried out according to the actual geometric structure of the microfluidic channel. The geometric model includes key structures such as the inlet end, outlet end, corners and splitting nodes of the channel to accurately describe the fluid flow path and dynamic characteristics. The three-dimensional geometric model is divided into hexahedral structured grids. The continuous computational domain is discretized into multiple finite elements so that numerical calculations can be carried out on these elements. Let the number of elements after grid division be , where represent the distribution numbers of the grids in , , directions respectively. To improve the calculation accuracy, the key areas of the geometric structure (such as inlets and outlets, corners and splitting nodes) are encrypted with grids. The size of the encrypted grids is smaller, so as to capture the changing characteristics of local flow more precisely. For example, at the inlet end, the grid size is set to 1 / 5 of the overall grid size to ensure the accuracy of the inlet and outlet flows. The mass conservation equation, momentum conservation equation and energy conservation equation are discretized onto the computational domain grids. For any grid element , the mass conservation equation is expressed as:

[0071] ;

[0072] where, is the fluid density, is the velocity vector. The momentum conservation equation is expressed as:

[0073] ;

[0074] where, is the pressure, is the dynamic viscosity. The energy conservation equation is:

[0075] ;

[0076] where, is the specific internal energy, is the heat flux, is the viscous dissipation term. By discretizing the above equations using the finite volume method, the spatial discretization adopts the second-order upwind scheme to improve the calculation accuracy and stability. After discretization, the discrete form of the mass conservation equation in the th grid element is:

[0077] ;

[0078] where, is the grid face normal area vector, and They are the density and velocity values on the grid surface respectively. Based on the above discrete equations, the SIMPLE algorithm is used for solving. The SIMPLE algorithm solves the coupling problem of non-linear equations by stepwise correcting the pressure and velocity fields. Assume an initial flow field , calculate the velocity correction term and the pressure correction term , and then update the pressure and velocity fields:

[0079] ;

[0080] wherein, and are relaxation factors, represents the number of iterations. Through multiple iterations, the initial value of the flow field is finally obtained. Based on the initial value of the flow field, the number of grids is refined and coarsened to form an optimal grid distribution. The principle of grid optimization is to increase the grid density in the regions where the fluid behavior changes violently (such as high-gradient regions), while reducing the grid density in the regions with gentle changes. For example, in the region where the fluid velocity gradient is higher than a certain threshold , the grids are refined to half of the original grid size, so as to improve the calculation accuracy. On the optimal grid distribution, using the initial flow field as the starting point, iterative solution is carried out until a steady-state flow field solution is obtained. The steady-state flow field solution includes the pressure field , temperature field and velocity field .

[0081] In a specific embodiment, the process of executing step S4 may specifically include the following steps:

[0082] Taking the flow field distribution data as sample points, performing Latin hypercube sampling on the width range, depth range and length range of the microfluidic channel to obtain a training sample space;

[0083] Based on the training sample space, a Kriging surrogate model is established. Taking the Gaussian correlation function as the kernel function, setting the relevant parameter θ to 0.1, the regression function order to 2, obtaining the correlation matrix of geometric parameters, and constructing the maximum likelihood estimation function based on the correlation matrix to optimize the hyperparameters and obtain the optimal interpolation weights of the Kriging model;

[0084] Substituting the optimal interpolation weights into the Kriging surrogate model to establish a mapping model between the channel width, depth, length and pressure uniformity, temperature uniformity, flow rate stability;

[0085] The non-dominated sorting genetic algorithm is optimized and solved based on the mapping model to obtain an evolutionary population. The Pareto optimal solution is calculated according to the evolutionary population. The crowding distance is used for individual ranking, and the individuals with a non-dominated rank of 1 are selected to form an elite population, obtaining the Pareto front.

[0086] The solution set in the Pareto front is normalized according to pressure uniformity, temperature uniformity, and flow stability to obtain candidate parameter combinations.

[0087] Sensitivity analysis is performed on the candidate parameter combinations, and the parameter combination with the optimal comprehensive evaluation value is selected as the optimal geometric parameter combination of the microfluidic channel.

[0088] Specifically, the flow field distribution data is used as sample points, combined with the geometric parameter range of the microfluidic channel, including the width range , depth range , and length range . The Latin hypercube sampling method is used to generate a training sample space. Latin hypercube sampling is an efficient multi-dimensional space sampling method. By uniformly dividing the parameter range and randomly arranging the sample point positions in each dimension, it ensures the uniform distribution of samples in the geometric parameter space. Assuming the sampling number is , for each sample point , the width, depth, and length are generated according to the following formulas respectively:

[0089] ;

[0090] ;

[0091] ;

[0092] where is the random permutation index in Latin hypercube sampling. In the generated training sample space, using the flow field distribution data as the response variable, a Kriging surrogate model is established. The Kriging model uses the Gaussian correlation function as the kernel function, and its formula is:

[0093] ;

[0094] where and are any two sets of geometric parameters, is the correlation parameter, is the normalization scale factor of the parameter range. The regression function adopts a second-order polynomial form:

[0095] ;

[0096] where is the regression coefficient, representing the linear and quadratic non - linear effects of geometric parameters on the target response. Combining the Gaussian correlation function and the regression function, the correlation matrix of sample points is calculated , whose elements are . The optimization objective function is constructed by maximum likelihood estimation:

[0097] ;

[0098] where is the determinant of the correlation matrix, is the response variable vector, is the number of samples. By optimizing to maximize the objective function, the optimal hyperparameters and interpolation weights are obtained. Substituting the optimal interpolation weights into the Kriging model, a mapping model between geometric parameters and pressure uniformity, temperature uniformity, and flow stability is established:

[0099] ;

[0100] where is the predicted value, is the correlation function vector, is the regression matrix, is the regression coefficient. Based on the mapping model, the non - dominated sorting genetic algorithm (NSGA - II) is used for optimization and solution. After initializing the population, the fitness of each individual is calculated according to the mapping model. An evolutionary population is generated through selection, crossover, and mutation operations. The set of Pareto - optimal solutions in the population is calculated, with the goal of maximizing pressure uniformity, temperature uniformity, and flow stability. During the screening process of Pareto - optimal solutions, the crowding distance is used for sorting, and individuals with a non - dominated rank of 1 are preferentially selected to form an elite population, obtaining the Pareto front. The solution set in the Pareto front is normalized according to the three performance indicators, and the normalization formula is:

[0101] ;

[0102] where is the normalized value, and are the minimum and maximum values of the corresponding performance indicators respectively. After normalization, the candidate parameter combinations are obtained. Sensitivity analysis is performed on the candidate parameter combinations, and the sensitivity index of each geometric parameter to the performance indicators is calculated. The formula is:

[0103] ;

[0104] where represents the variance of the predicted value when a certain parameter is fixed, It represents the variance of the overall predicted value. Based on the comprehensive evaluation of the sensitivity analysis results, the parameter combination with the optimal comprehensive evaluation value is selected as the optimal geometric parameter combination of the microfluidic channel.

[0105] In a specific embodiment, the process of performing the step of establishing a Kriging surrogate model based on the training sample space, using the Gaussian correlation function as the kernel function, setting the relevant parameter θ to 0.1, and the regression function order to 2, obtaining the correlation matrix of geometric parameters, and constructing the maximum likelihood estimation function based on the correlation matrix to optimize the hyperparameters and obtain the optimal interpolation weights of the Kriging model can specifically include the following steps:

[0106] Construct a regression function matrix based on the training sample space, and expand it using the second-order polynomial basis function to obtain the regression model coefficients;

[0107] Substitute the regression model coefficients into the Gaussian correlation function, calculate the spatial correlation between sample points to obtain the correlation function matrix, and perform eigen-dimension reduction decomposition on the correlation function matrix to obtain the dimension-reduced feature space;

[0108] Construct a maximum likelihood estimation function based on the dimension-reduced feature space, use the logarithmic transformation form to obtain the optimization objective function, and input the optimization objective function into the pattern search algorithm to obtain the optimal hyperparameters;

[0109] Perform cross-validation on the optimal hyperparameters to obtain the model accuracy evaluation index, and correct the Kriging surrogate model according to the model accuracy evaluation index. Through the calculation of the generalized least squares method, obtain the optimal interpolation weights of the Kriging model.

[0110] Specifically, construct a regression function matrix for the data in the training sample space. Assume that the training sample space contains sample points, and each sample point is a -dimensional geometric parameter combination, and the target response value is . The regression function is expanded using the second-order polynomial basis function to represent the basic relationship between sample points, and the specific form is:

[0111] ;

[0112] Among them, is the constant term, and are the regression coefficients of the linear term and the quadratic term respectively, representing the linear and non-linear effects of the sample points in each dimension. Through the input and the corresponding output of all sample points, construct the regression matrix and the response vector :

[0113] ;

[0114] ;

[0115] Regression model coefficient vector Calculated by the least squares method:

[0116] ;

[0117] Substitute the calculated regression coefficients into the Gaussian correlation function to construct the spatial correlation matrix between sample points. The form of the Gaussian correlation function is:

[0118] ;

[0119] where represents the correlation between sample points and and is the correlation parameter of the th dimension, which is used to control the decay rate of the correlation function. By calculating the correlation for all sample point pairs, the correlation function matrix is constructed:

[0120] ;

[0121] Perform eigen-dimension reduction decomposition on the correlation function matrix to obtain the dimension-reduced feature space. By calculating the eigenvalues and eigenvectors and of

[0122] ;

[0123] where is the number of features after dimension reduction. Select the eigenvectors corresponding to the main eigenvalues to construct the dimension-reduced feature space, thereby reducing the computational complexity. Based on the dimension-reduced feature space, construct the maximum likelihood estimation function and adopt the logarithmic transformation form to transform the optimization objective function into:

[0124] ;

[0125] where represents the determinant of the correlation matrix, is the weighted sum of squares of the prediction errors. Optimize through the pattern search algorithm to obtain the optimal hyperparameter Cross-validate the optimal hyperparameters, evaluate the prediction accuracy of the model by dividing the training set and the test set, and calculate evaluation metrics such as mean squared error (MSE) and coefficient of determination ( ):

[0126] ;

[0127] ;

[0128] wherein, is the model prediction value, is the true value, is the mean of the true values. Modify the Kriging model according to the evaluation metrics, and calculate the optimal interpolation weight by the generalized least squares method:

[0129] ;

[0130] Finally, the constructed Kriging model accurately predicts the target response value.

[0131] In a specific embodiment, the process of executing step S5 may specifically include the following steps:

[0132] Perform correlation analysis on the pressure control feature sequence, temperature distribution feature sequence, and flow rate change feature sequence in the multi-dimensional fluid feature matrix with the optimal geometric parameter combination respectively to obtain the feature-parameter mapping relationship;

[0133] Input the feature-parameter mapping relationship into the control strategy generation module to construct an initial hierarchical control framework for the pressure control layer, temperature control layer, and flow rate control layer;

[0134] Based on the initial hierarchical control framework, perform dynamic programming on the pressure control layer data to obtain the pressure uniformity control parameters;

[0135] Based on the initial hierarchical control framework, perform model predictive control on the temperature control layer data to obtain the temperature distribution uniformity control parameters;

[0136] Based on the initial hierarchical control framework, perform robust control calculation on the flow rate control layer data to obtain the flow rate stability control parameters;

[0137] Input the pressure uniformity control parameters, temperature distribution uniformity control parameters, and flow rate stability control parameters into the hierarchical coordinator, and perform hierarchical optimization for inter-layer coordination to obtain the hierarchical fluid parameter control strategy.

[0138] Specifically, associate the pressure control feature sequence, temperature distribution feature sequence, and flow rate change feature sequence with the optimal geometric parameter combination respectively to construct the feature-parameter mapping relationship. Assume that the multi-dimensional fluid feature matrix is , where represents the pressure control feature sequence, represents the temperature distribution feature sequence, represents the flow rate change feature sequence, is the number of samples. The optimal geometric parameter combination is denoted as , where respectively represent the width, depth, and length of the channel. Through correlation analysis, using a multiple linear regression model or a non - linear kernel method, a mapping relationship is established between each feature sequence and the geometric parameter combination. For example:

[0139] ;

[0140] ;

[0141] ;

[0142] where, is the constant term, etc. are the regression coefficients, is the error term. The regression coefficients are optimized by the least - squares method to construct a complete feature - parameter mapping relationship. These mapping relationships are input into the control strategy generation module to construct an initial hierarchical control framework for the pressure control layer, temperature control layer, and flow rate control layer. Each control layer optimizes specific performance objectives to ensure pressure uniformity, temperature distribution uniformity, and flow rate stability. Taking the pressure control layer as an example, its initial inputs are the feature - parameter mapping relationship and the pressure control feature sequence . The data of the pressure control layer is optimized by the dynamic programming method. Dynamic programming decomposes the global optimization problem into a series of stage - based sub - problems, and each stage ensures the overall optimality through recursive solution. Assume the pressure uniformity objective function is:

[0143] ;

[0144] where, is the target pressure value. The optimization update formula of dynamic programming is:

[0145] ;

[0146] where, is the loss function at stage , is the control variable. The final output is the pressure uniformity control parameter . Similarly, in the temperature control layer, the model predictive control method is used to optimize the temperature distribution feature sequence. Model predictive control performs rolling optimization on temperature data based on a prediction model, and its objective function is:

[0147] ;

[0148] Among them, is the target temperature value, is the weight factor of the control increment. By optimizing the control input within the rolling window, the control parameters for temperature distribution uniformity are generated. In the flow control layer, a robust control method is adopted to optimize the flow change characteristic sequence to handle the uncertainties and disturbances in the system. The flow stability objective function is:

[0149] ;

[0150] Among them, is the target flow value, is the robust control weight coefficient, is the stage flow change. By solving the robust control problem, the flow stability control parameter is output. The optimization results of the three control layers, namely the pressure uniformity control parameter , the temperature distribution uniformity control parameter , and the flow stability control parameter , are input to the hierarchical coordinator for hierarchical optimization. The objective of hierarchical optimization is to coordinate the control strategies of each layer to make the global performance objective reach the optimal. The objective function of hierarchical optimization is:

[0151] ;

[0152] Among them, are the weight coefficients of each layer. Through hierarchical optimization, the control parameters of each layer are adjusted to make the optimization results of pressure, temperature, and flow coordinated globally, and the hierarchical fluid parameter control strategy is obtained.

[0153] Among them, the pressure uniformity control parameter, temperature distribution uniformity control parameter, and flow stability control parameter are input into the hierarchical coordinator, and hierarchical optimization is used for inter-layer coordination to obtain a hierarchical fluid parameter control strategy, including: performing a hierarchical coupling degree analysis on the pressure uniformity control parameter, temperature distribution uniformity control parameter, and flow stability control parameter to obtain a parameter coupling relationship matrix; inputting the parameter coupling relationship matrix into the hierarchical coordinator, and using the analytic hierarchy process to calculate the weight coefficients of each control layer to obtain an inter-layer weight distribution scheme; constructing a hierarchical optimization objective function based on the inter-layer weight distribution scheme, setting the priority of the pressure layer as 1, the priority of the temperature layer as 2, and the priority of the flow layer as 3 to obtain a hierarchical optimization sequence; performing a first-order optimization calculation on the pressure layer in the hierarchical optimization sequence, setting the pressure uniformity constraint condition as ±0.5 MPa to obtain the optimal control value of the pressure layer; substituting the optimal control value of the pressure layer into the temperature layer for a second-order optimization calculation, setting the temperature uniformity constraint condition as ±1 °C to obtain the optimal control value of the temperature layer; performing a third-order optimization calculation on the flow layer based on the optimal control value of the pressure layer and the optimal control value of the temperature layer, setting the flow stability constraint condition as ±0.1 ml / s to obtain the optimal control value of the flow layer; performing a coordination verification on the optimal control values of the pressure layer, the optimal control value of the temperature layer, and the optimal control value of the flow layer, using the cross-validation method to calculate the mutual influence degree of the control parameters of each layer to obtain a coordinated control sequence; organizing the coordinated control sequence in time series according to the control period, setting the pressure control period as 0.001 s, the temperature control period as 0.01 s, and the flow control period as 0.1 s to obtain a hierarchical fluid parameter control strategy.

[0154] The above describes the fluid control intelligent adjustment method of the microfluidic chip in the embodiment of the present invention. Next, the fluid control intelligent adjustment system of the microfluidic chip in the embodiment of the present invention will be described. Please refer to Figure 2 , an embodiment of the fluid control intelligent adjustment system of the microfluidic chip in the embodiment of the present invention includes:

[0155] A data acquisition module, configured to collect real-time fluid parameter monitoring data based on the microfluidic chip for an adaptive fluid controller, a microfluidic channel pressure sensor array, a temperature sensor array, and a flow sensor array.

[0156] A feature extraction module, configured to input the real-time fluid parameter monitoring data into a dynamic self-learning control network model for feature extraction to obtain a multi-dimensional fluid feature matrix.

[0157] A solution module, configured to establish a three-dimensional fluid model of the microfluidic channel within a set pressure range and temperature range, and solve to obtain flow field distribution data.

[0158] A building module is used to input the flow field distribution data into the Kriging surrogate model, establish a mapping model between the channel width, depth, length and fluid control performance, and calculate the optimal geometric parameter combination of the microfluidic channel through the non-dominated sorting genetic algorithm;

[0159] A calculation module is used to perform multi-level calculations on the fluid pressure uniformity, temperature distribution uniformity and flow rate stability according to the multi-dimensional fluid characteristic matrix and the optimal geometric parameter combination, and obtain a hierarchical fluid parameter control strategy.

[0160] Through the collaborative cooperation of the above-mentioned various components, the present invention adopts the combination of an adaptive fluid controller and a multi-type sensor array to achieve high-precision real-time monitoring of the fluid parameters in the microfluidic channel, improving the accuracy of data acquisition; by constructing a dynamic self-learning control network model, it realizes the adaptive extraction of fluid characteristics and enhances the adaptability of the system to different working conditions; innovatively combines the Kriging surrogate model with the non-dominated sorting genetic algorithm to establish an accurate prediction model of fluid control performance and optimize the control parameters; the hierarchical fluid parameter control strategy proposed by the present invention realizes the collaborative optimization of the three control layers of pressure, temperature and flow rate through multi-level calculations, improving the control accuracy; using a hierarchical optimization method for inter-layer coordination effectively solves the multi-parameter coupling problem and ensures the stability of the control system.

[0161] Referring to Figure 3 , an embodiment of the present invention also provides a computer device, which can be a server, and its internal structure can be as Figure 3 shown. The computer device includes a processor, a memory, a display screen, an input device, a network interface and a database connected through a system bus. Among them, the processor of the computer design is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is realized.

[0162] Those skilled in the art can understand that Figure 3 the structure shown in

[0163] An embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0164] Those of ordinary skill in the art can understand that all or part of the processes in the above-described method embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the above-described method embodiments. Among them, any reference to memory, storage, database, or other media provided by the present invention and used in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM, etc.

[0165] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, systems, and units can refer to the corresponding processes in the foregoing method embodiments and will not be described herein again.

[0166] When the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs.

[0167] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of various embodiments of the present invention.

Claims

1. A fluid control intelligent adjustment method for a microfluidic chip, characterized in that, The method includes: Collecting real-time fluid parameter monitoring data based on a microfluidic chip for an adaptive fluid controller, a microfluidic channel pressure sensor array, a temperature sensor array, and a flow sensor array; Inputting the real-time fluid parameter monitoring data into a dynamic self-learning control network model for feature extraction to obtain a multi-dimensional fluid feature matrix; Establishing a three-dimensional fluid model of the microfluidic channel within a set pressure range and temperature range, and solving to obtain flow field distribution data; Inputting the flow field distribution data into a Kriging surrogate model, establishing a mapping model between the channel width, depth, length, and fluid control performance, and calculating the optimal geometric parameter combination of the microfluidic channel through a non-dominated sorting genetic algorithm; specifically including: using the flow field distribution data as sample points, performing Latin hypercube sampling on the width range, depth range, and length range of the microfluidic channel to obtain a training sample space; establishing a Kriging surrogate model based on the training sample space, using a Gaussian correlation function as the kernel function, setting the correlation parameter θ to 0.1, and the regression function order to 2 to obtain a correlation matrix of geometric parameters, and constructing a maximum likelihood estimation function based on the correlation matrix to optimize the hyperparameters to obtain the optimal interpolation weight of the Kriging model; substituting the optimal interpolation weight into the Kriging surrogate model to establish a mapping model between the channel width, depth, length, and pressure uniformity, temperature uniformity, and flow stability; performing non-dominated sorting genetic algorithm optimization on the mapping model to obtain an evolutionary population, calculating the Pareto optimal solution based on the evolutionary population, performing individual ranking using crowding distance, selecting individuals with a non-dominated rank of 1 to form an elite population to obtain the Pareto front; normalizing the solution set in the Pareto front according to pressure uniformity, temperature uniformity, and flow stability to obtain candidate parameter combinations; performing sensitivity analysis on the candidate parameter combinations, and selecting the parameter combination with the optimal comprehensive evaluation value as the optimal geometric parameter combination of the microfluidic channel; According to the multi-dimensional fluid feature matrix and the optimal geometric parameter combination, performing multi-level calculations on fluid pressure uniformity, temperature distribution uniformity, and flow stability to obtain a hierarchical fluid parameter control strategy.

2. The intelligent adjustment method for fluid control of the microfluidic chip according to claim 1, characterized in that, The collecting real-time fluid parameter monitoring data based on a microfluidic chip for an adaptive fluid controller, a microfluidic channel pressure sensor array, a temperature sensor array, and a flow sensor array includes: Performing grid division on the microfluidic channel, and at the same time, arranging the pressure sensor array at grid nodes to obtain a pressure monitoring point distribution network; calibrating key points at the inlet, outlet, corners, and shunt nodes of the microfluidic channel, and arranging the temperature sensor array at the calibrated point positions to collect temperature monitoring node matrix data; setting flow monitoring points at the feed end and discharge end of the microfluidic channel, and arranging the flow sensor array at the monitoring points to collect fluid flow parameters; Intelligently partition the reference fluid chamber based on an adaptive fluid controller to obtain an array of reference pressure monitoring points, and perform real-time difference calculation on the pressure data of the pressure monitoring point distribution network and the pressure data of the reference pressure monitoring point array to obtain a pressure gradient distribution sequence; Based on the microfluidic chip, perform data fusion on the pressure gradient distribution sequence, the temperature monitoring node matrix data, and the fluid flow parameters to obtain real-time fluid parameter monitoring data.

3. The fluid control intelligent adjustment method of the microfluidic chip according to claim 2, characterized in that, Input the real-time fluid parameter monitoring data into a dynamic self-learning control network model for feature extraction to obtain a multi-dimensional fluid feature matrix, including: Input the pressure gradient distribution sequence of the real-time fluid parameter monitoring data into the adaptive fluid control module in the dynamic self-learning control network model for processing to obtain a pressure control feature sequence. The adaptive fluid control module includes a pressure feature extraction network and a pressure parameter optimization network; the pressure feature extraction network includes three convolutional layers and two fully connected layers, and the pressure parameter optimization network includes two fully connected layers and one output layer; Input the temperature monitoring node matrix data of the real-time fluid parameter monitoring data into the spatio-temporal feature extraction module in the dynamic self-learning control network model for three-dimensional convolution operation to obtain a temperature distribution feature sequence. The spatio-temporal feature extraction module includes a time series feature network and a spatial feature network. The time series feature network adopts a long short-term memory network structure, and the spatial feature network adopts a three-dimensional convolution structure; Input the fluid flow parameters of the real-time fluid parameter monitoring data into the parameter self-adjustment module in the dynamic self-learning control network model for dynamic weight adjustment to obtain a flow rate change feature sequence. The parameter self-adjustment module includes a flow rate prediction network and a parameter correction network; Perform feature fusion on the pressure control feature sequence, the temperature distribution feature sequence, and the flow rate change feature sequence to obtain a fusion feature matrix, and perform time series reconstruction on the fusion feature matrix to obtain a multi-dimensional fluid feature matrix.

4. The intelligent adjustment method for fluid control of the microfluidic chip according to claim 3, characterized in that, Establish a three-dimensional fluid model of the microfluidic channel within a set pressure range and temperature range, and solve to obtain flow field distribution data, including: Determine the pressure range and temperature range of the microfluidic channel, and perform three-dimensional solid modeling on the microfluidic channel to obtain a three-dimensional fluid model including a feed end, a discharge end, a corner, and a splitting node; Perform hexahedral structured grid division on the three-dimensional fluid model, and encrypt the grids at the inlet, outlet, corner, and splitting node to obtain a computational domain grid; Discretize the mass conservation equation, the momentum conservation equation, and the energy conservation equation on the computational domain grid using the finite volume method, and perform spatial discretization using the second-order upwind scheme to obtain a discrete control equation set; Solve the discrete control equation set based on the SIMPLE algorithm to obtain an initial value of the flow field, and encrypt and coarsen the number of grids based on the initial value of the flow field to obtain an optimal grid distribution; Iteratively solve the initial value of the flow field on the optimal grid distribution to obtain a steady-state flow field solution, and extract fluid pressure field, temperature field, and velocity field data based on the steady-state flow field solution to obtain flow field distribution data.

5. The intelligent adjustment method for fluid control of the microfluidic chip according to claim 1, characterized in that Based on the above training sample space, a Kriging surrogate model is established. The Gaussian correlation function is used as the kernel function, the relevant parameter θ is set to 0.1, and the order of the regression function is 2. The correlation matrix of geometric parameters is obtained, and the maximum likelihood estimation function is constructed based on the correlation matrix to optimize the hyperparameters, and the optimal interpolation weights of the Kriging model are obtained, including: Construct a regression function matrix based on the training sample space, and expand it using a second-order polynomial basis function to obtain the regression model coefficients; Substitute the regression model coefficients into the Gaussian correlation function, calculate the spatial correlation between sample points to obtain the correlation function matrix, and perform eigen-dimension reduction decomposition on the correlation function matrix to obtain the dimension-reduced feature space; Construct a maximum likelihood estimation function based on the dimension-reduced feature space, adopt a logarithmic transformation form to obtain the optimization objective function, and input the optimization objective function into the pattern search algorithm to obtain the optimal hyperparameters; Perform cross-validation on the optimal hyperparameters to obtain the model accuracy evaluation index, and correct the Kriging surrogate model according to the model accuracy evaluation index. Through the calculation of the generalized least squares method, the optimal interpolation weights of the Kriging model are obtained.

6. The fluid control intelligent adjustment method of the microfluidic chip according to claim 5, characterized in that, According to the multi-dimensional fluid feature matrix and the optimal geometric parameter combination, perform multi-level calculations on the fluid pressure uniformity, temperature distribution uniformity, and flow stability to obtain a hierarchical fluid parameter control strategy, including: Perform correlation analysis on the pressure control feature sequence, temperature distribution feature sequence, and flow change feature sequence in the multi-dimensional fluid feature matrix respectively with the optimal geometric parameter combination to obtain the feature-parameter mapping relationship; Input the feature-parameter mapping relationship into the control strategy generation module to construct the initial hierarchical control framework of the pressure control layer, temperature control layer, and flow control layer; Perform dynamic programming on the data of the pressure control layer based on the initial hierarchical control framework to obtain the pressure uniformity control parameters; Perform model predictive control on the data of the temperature control layer based on the initial hierarchical control framework to obtain the temperature distribution uniformity control parameters; Perform robust control calculation on the data of the flow control layer based on the initial hierarchical control framework to obtain the flow stability control parameters; Input the pressure uniformity control parameters, the temperature distribution uniformity control parameters, and the flow stability control parameters into the hierarchical coordinator, and perform hierarchical coordination using hierarchical optimization to obtain the hierarchical fluid parameter control strategy.

7. A fluid control intelligent regulation system for a microfluidic chip, characterized in that, For implementing the intelligent adjustment method for fluid control of the microfluidic chip according to any one of claims 1-6, the intelligent adjustment system for fluid control of the microfluidic chip includes: A data acquisition module for collecting real-time fluid parameter monitoring data by collecting data from an adaptive fluid controller, a microfluidic channel pressure sensor array, a temperature sensor array, and a flow sensor array based on the microfluidic chip; A feature extraction module for inputting the real-time fluid parameter monitoring data into a dynamic self-learning control network model for feature extraction to obtain a multi-dimensional fluid feature matrix; A solution module, configured to establish a three-dimensional fluid model of a microfluidic channel within a set pressure range and temperature range, and solve to obtain flow field distribution data; An establishment module, configured to input the flow field distribution data into a Kriging surrogate model, establish a mapping model between the channel width, depth, length and fluid control performance, and calculate the optimal geometric parameter combination of the microfluidic channel through a non-dominated sorting genetic algorithm; A calculation module, configured to perform multi-level calculations on fluid pressure uniformity, temperature distribution uniformity and flow rate stability according to the multi-dimensional fluid characteristic matrix and the optimal geometric parameter combination, and obtain a hierarchical fluid parameter control strategy.

8. A computer device, characterized in that, It includes a memory and a processor, and the memory stores a computer program that can be run on the processor. It is characterized in that when the processor executes the computer program, it implements the fluid control intelligent adjustment method of the microfluidic chip according to any one of claims 1 to 6.

9. A computer-readable storage medium, on which a computer program is stored, and when the computer program is run by a processor, the processor is caused to execute the fluid control intelligent adjustment method of the microfluidic chip according to any one of claims 1 to 6.

Images