Eye drop filtering equipment monitoring method and system based on Internet of Things, and storage medium

Abstract

The invention relates to the technical field of equipment monitoring and control, and discloses an eyedrop filtering equipment monitoring method and system based on the Internet of Things, and a storage medium. The method comprises the following steps: collecting drug molecule data through a multi-dimensional sensor, and carrying out dimension reduction processing to obtain an initial feature data set; extracting molecular structure features according to the feature data set to obtain risk assessment data; establishing a medication portrait and a risk coefficient based on the evaluation data, and obtaining a filtering parameter; monitoring pollution by adopting quantum dots and cleaning to obtain operation data; dividing risk levels according to the operation data to obtain an evaluation report; and updating a parameter prediction maintenance period according to the evaluation report to obtain a management and control instruction. According to the invention, intelligent monitoring of the eyedrop filtering equipment is realized, the performance attenuation trend of the equipment is accurately predicted, and the potential cross contamination risk is found in time.

Description

Technical Field

[0001] The present application relates to the field of equipment monitoring and control, and in particular to an eye drops filtering equipment monitoring method, system and storage medium based on the Internet of Things. Background Art

[0002] Currently, medical institutions widely use eye drop filtration equipment to filter compound eye drops. Existing filtration equipment typically employs a multi-stage filtration structure, including coarse filtration, fine filtration, and ultrafiltration layers, and is equipped with basic monitoring devices to record filtration parameters. These devices monitor filtration pressure using pressure sensors, test water quality using turbidity sensors, and feature simple cleaning devices to remove sediment from the filter membrane surface. Traditional monitoring methods rely primarily on regular manual inspections of equipment operating status, empirically determined maintenance opportunities, or fixed maintenance cycles.

[0003] However, existing technologies have the following shortcomings: Because compound eye drops contain multiple active ingredients, interactions between them can lead to unstable filtration performance. Traditional monitoring methods are unable to detect changes in drug components and their interactions in real time, nor can they accurately predict the performance degradation trends of filtration equipment. Furthermore, fixed-cycle maintenance methods often fail to adapt to changes in actual usage, potentially leading to unnecessary maintenance and missed critical maintenance opportunities. Furthermore, existing technologies lack in-depth analysis and utilization of equipment operating data, making it impossible to achieve intelligent predictive maintenance. Summary of the Invention

[0004] The present application provides an eye drops filtration equipment monitoring method, system and storage medium based on the Internet of Things, which are used to realize intelligent monitoring of eye drops filtration equipment, accurately predict the equipment performance degradation trend, and promptly detect potential cross-contamination risks.

[0005] In the first aspect, the present application provides an eye drops filtration equipment monitoring method based on the Internet of Things, and the eye drops filtration equipment monitoring method based on the Internet of Things includes: collecting drug molecular fingerprint maps and protein marker data through a multidimensional sensor, performing dimensionality reduction processing on the collected data, and obtaining an initial drug feature data set; constructing a drug interaction model based on the initial drug feature data set, performing feature extraction on the drug molecular structure, and obtaining drug interaction risk assessment data; based on the drug interaction risk assessment data, digitally storing medication information, establishing a medication portrait and evaluating the risk factor, and obtaining personalized filtering scheme parameters; dynamically adjusting the personalized filtering scheme parameters, using quantum dot tracing to monitor cross contamination, and obtaining equipment operation status data after pulse cleaning treatment; establishing a drug safety model based on the equipment operation status data, dividing the risk levels and integrating the monitoring data to obtain an equipment monitoring evaluation report; extracting optimization plans based on the equipment monitoring evaluation report, updating equipment parameters and predicting maintenance cycles, and obtaining intelligent management and control instructions.

[0006] In a second aspect, the present application provides an Internet of Things-based eye drops filtration equipment monitoring system, the Internet of Things-based eye drops filtration equipment monitoring system comprising:

[0007] The acquisition module is used to collect drug molecular fingerprints and protein marker data through multi-dimensional sensors, and perform dimensionality reduction processing on the collected data to obtain the initial drug feature data set;

[0008] A construction module is used to construct a drug interaction model based on the initial drug feature data set, perform feature extraction on the drug molecular structure, and obtain drug interaction risk assessment data;

[0009] A storage module is used to digitally store medication information based on the drug interaction risk assessment data, establish a medication profile and assess the risk factor, and obtain personalized filtering solution parameters;

[0010] An adjustment module is used to dynamically adjust the parameters of the personalized filtration solution, monitor cross contamination using quantum dot tracing, and obtain equipment operating status data through pulse cleaning treatment;

[0011] A classification module is used to establish a drug safety model based on the equipment operation status data, classify the risk levels and integrate the monitoring data to obtain an equipment monitoring evaluation report;

[0012] The prediction module is used to extract optimization solutions based on the equipment monitoring and evaluation report, update equipment parameters and predict maintenance cycles to obtain intelligent management and control instructions.

[0013] A third aspect of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned method for monitoring eye drops filtration equipment based on the Internet of Things.

[0014] The technical solution provided in this application uses multidimensional sensors to collect drug molecular fingerprints and protein marker data. Combined with wavelet transform denoising and principal component analysis dimensionality reduction, it can accurately capture changes in drug components, effectively reduce data noise interference, and improve the accuracy of data analysis. A drug interaction model is constructed based on the initial drug feature dataset, and feature extraction is performed using a graph neural network. This enables precise analysis of drug molecular structures and their interactions, providing reliable data support for subsequent risk assessment. During the digital storage of medication information, blockchain technology and deep learning algorithms are combined to create medication profiles, which not only ensures data security and traceability but also enables accurate assessment of medication risks. Quantum dot tracing technology is used to monitor cross contamination in real time, and combined with pulse cleaning technology, a complete pollution prevention and control mechanism is established, significantly reducing the risk of cross contamination. A drug safety model established based on equipment operating status data uses a multi-level assessment method to classify risk levels, enabling comprehensive monitoring of equipment operating status. By analyzing equipment monitoring and evaluation reports and formulating optimization plans, combined with reinforcement learning algorithms to dynamically adjust equipment parameters, continuous optimization of equipment performance is achieved. In particular, this solution integrates multiple deep learning algorithms and traditional data processing methods to establish an intelligent monitoring and decision-making framework. The application of graph neural networks in drug interaction analysis significantly improves the accuracy of molecular structure feature extraction; the application of deep learning algorithms in the construction of medication profiles enables precise assessment of medication risks; and the application of reinforcement learning algorithms in parameter optimization enables adaptive adjustment of device control strategies. This organic combination of algorithmic features not only enhances the system's intelligence but also enables refined management of the entire lifecycle of eye drop filtration equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0016] Figure 1 This is a schematic diagram of an embodiment of a method for monitoring eye drops filtration equipment based on the Internet of Things in an embodiment of the present application;

[0017] Figure 2This is a schematic diagram of the timing of collecting drug molecular fingerprints and protein marker data through a multi-dimensional sensor in an embodiment of the present application;

[0018] Figure 3 This is a flow chart of a method for performing time-series decomposition processing on drug interaction risk assessment data and then encoding the data according to medication time in an embodiment of the present application;

[0019] Figure 4 This is a schematic diagram of an embodiment of an eye drops filtration equipment monitoring system based on the Internet of Things in an embodiment of the present application. DETAILED DESCRIPTION

[0020] The present application embodiment provides a kind of eye drops filtration equipment monitoring method, system and storage medium based on the Internet of Things. The terms "first", "second", "third", "fourth" etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or precedence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described here can be implemented in a sequence other than that illustrated or described here. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0021] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In the embodiment of the present application, an embodiment of the method for monitoring eye drops filtration equipment based on the Internet of Things includes:

[0022] Step S101: collecting drug molecular fingerprints and protein marker data through a multidimensional sensor, performing dimensionality reduction processing on the collected data, and obtaining an initial drug feature data set;

[0023] Step S102: constructing a drug interaction model based on the initial drug feature data set, extracting features of the drug molecular structure, and obtaining drug interaction risk assessment data;

[0024] Step S103: Based on the drug interaction risk assessment data, the medication information is digitally stored, a medication profile is established, and the risk coefficient is assessed to obtain personalized filtering scheme parameters;

[0025] Step S104: Dynamically adjust the parameters based on the personalized filtration solution, use quantum dot tracing to monitor cross contamination, and obtain equipment operating status data through pulse cleaning.

[0026] Step S105: Establish a drug safety model based on the equipment operation status data, classify the risk levels, and integrate the monitoring data to obtain an equipment monitoring assessment report;

[0027] Step S106: Extract optimization solutions based on the equipment monitoring and evaluation report, update equipment parameters, predict maintenance cycles, and obtain intelligent management and control instructions.

[0028] It is understandable that the execution subject of this application can be an eye drops filtration equipment monitoring system based on the Internet of Things, or a terminal or a server, which is not limited here. The embodiment of this application is described by taking the server as the execution subject as an example.

[0029] Specifically, a multidimensional sensor performs real-time testing of eye drops. A micro-Raman spectrometer captures drug molecular fingerprints, which contain information about the characteristic peaks and vibrational modes of the drug molecules. An immunosensor array collects protein marker data, and an electrochemical sensing unit detects changes in protein content in the solution. The collected raw data is first subjected to wavelet transform denoising to eliminate noise caused by environmental interference and electrical signal fluctuations. Dimensionality reduction is then performed using principal component analysis, mapping the high-dimensional data into a lower-dimensional feature space to generate an initial drug feature dataset. Based on this initial drug feature dataset, the structural characteristics of the drug molecules are analyzed. Specifically, surface-enhanced Raman scattering (SERS) is used to obtain scattering spectra of the drug molecules, extracting molecular bonding information and spatial configuration features from the spectral data. This feature data is then input into a drug interaction analysis unit to calculate the forces and binding sites between different drug molecules and construct a drug interaction model. This model reflects the compatibility and potential mutual interference between the different drug components, and drug interaction risk assessment data is generated through calculations.

[0030] Based on drug interaction risk assessment data, a digital medication information storage system is established. The risk assessment data is time-series encoded to establish a data structure containing medication time, dosage, and interaction strength. By performing cluster analysis on this data, medication patterns and characteristics are extracted to form a medication profile. In combination with the risk factors in the medication profile, the weight coefficients of various parameters are calculated to generate a risk coefficient. Based on the risk coefficient, the adjustment range of the filtration parameters is determined to obtain personalized filtration scheme parameters. The personalized filtration scheme parameters are used to dynamically adjust the equipment to adjust key parameters such as flow rate, pressure, and temperature. During the filtration process, quantum dot tracing technology is used to monitor cross-contamination. Quantum dots are nanoscale fluorescent markers that determine the diffusion and migration of pollutants by detecting the intensity of their fluorescent signals. When cross-contamination is detected to exceed the threshold, a pulse cleaning program is initiated, using high-frequency oscillating water flow to flush the filter components and remove residual substances. The operating parameters and cleaning effects are recorded throughout the process to form equipment operating status data.

[0031] Segment equipment operating status data by time series to extract performance parameter variation characteristics. Analyze parameter fluctuation patterns and trends to identify different risk levels. Focus on monitoring data in high-risk intervals, recording abnormalities and handling measures. Integrate and analyze all monitoring data to generate an equipment monitoring and evaluation report. Finally, perform optimization analysis based on the equipment monitoring and evaluation report. Extract the variation trends of key performance indicators and compare them with historical data to identify parameters that require optimization. Conduct sensitivity analysis on equipment parameters to determine the optimal adjustment plan. Simultaneously, evaluate the operating status and life cycle of each component, predict maintenance time points, and develop a maintenance plan. Combine parameter optimization plans with maintenance plans to generate intelligent management and control instructions.

[0032] For example: A medical institution uses this eye drop filtration equipment to process compound eye drops. The equipment collects the molecular fingerprint maps of drug A and drug B, and detects changes in the concentration of protein markers in the solution. Through data analysis, it was found that the two drugs will interact at a specific pH value to form a precipitate. Based on this risk, the system automatically adjusts the filtration parameters, lowers the pH value, and increases the monitoring frequency. During operation, quantum dot tracing shows that contaminants have accumulated in certain areas of the filter component. The system immediately initiates directional pulse cleaning and effectively removes the contaminants. Through continuous monitoring and data analysis, the system discovered that the performance of a certain filter component began to decline, predicted its remaining service life, and arranged replacement and maintenance in advance. Throughout the process, the system continuously optimizes operating parameters to ensure filtration effect and equipment performance.

[0033] In the embodiment of the present application, drug molecular fingerprints and protein marker data are collected by multidimensional sensors, combined with wavelet transform denoising and principal component analysis dimensionality reduction processing, which can accurately obtain the changes in drug components, effectively reduce data noise interference, and improve the accuracy of data analysis. A drug interaction model is constructed based on the initial drug feature data set, and feature extraction is performed using a graph neural network to achieve accurate analysis of drug molecular structures and their interactions, providing reliable data support for subsequent risk assessment. In the digital storage process of medication information, a medication portrait is established by combining blockchain technology and deep learning algorithms, which not only ensures the security and traceability of data, but also achieves an accurate assessment of medication risks. Cross-contamination is monitored in real time through quantum dot tracing technology, and combined with pulse cleaning technology, a complete set of pollution prevention and control mechanisms is established, which greatly reduces the risk of cross-contamination. A drug safety model established based on equipment operating status data uses a multi-level assessment method to divide risk levels, achieving comprehensive monitoring of equipment operating status. By analyzing equipment monitoring and evaluation reports and formulating optimization plans, combined with reinforcement learning algorithms to dynamically adjust equipment parameters, continuous optimization of equipment performance is achieved. In particular, this solution integrates multiple deep learning algorithms and traditional data processing methods to establish an intelligent monitoring and decision-making framework. The application of graph neural networks in drug interaction analysis significantly improves the accuracy of molecular structure feature extraction; the application of deep learning algorithms in the construction of medication profiles enables precise assessment of medication risks; and the application of reinforcement learning algorithms in parameter optimization enables adaptive adjustment of device control strategies. This organic combination of algorithmic features not only enhances the system's intelligence but also enables refined management of the entire lifecycle of eye drop filtration equipment.

[0034] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0035] (1) Collecting drug molecular fingerprints through a micro-Raman spectroscopy sensor, performing signal preprocessing on the drug molecular fingerprints, and obtaining drug molecular characteristic values;

[0036] (2) collecting protein marker data through an immunosensor array, performing noise elimination on the protein marker data, and obtaining marker characteristic values;

[0037] (3) Fusing the drug molecule characteristic values ​​and the marker characteristic values, and denoising the fused data by wavelet transform to obtain denoised data;

[0038] (4) Performing principal component analysis and dimensionality reduction on the denoised data, and obtaining dimensionality-reduced data through feature extraction;

[0039] (5) Standardize the dimension-reduced data and obtain calibration data by controlling the flow rate sampling through a micropump;

[0040] (6) The calibration data is corrected for data drift and calibrated with data standard samples to obtain the initial drug characteristic data set.

[0041] Specifically, if Figure 2 Figure 1 shows a schematic diagram of the timing of drug molecular fingerprint and protein marker data acquisition using a multidimensional sensor in an embodiment of the present application. The device includes a Raman spectroscopy sensor, an immunosensor array, a data processing unit, a micropump, and a standard sample unit. The diagram illustrates the complete data processing flow from raw data acquisition to the generation of an initial drug signature dataset: First, the Raman spectroscopy sensor acquires the drug molecular fingerprint, and the data processing unit performs signal preprocessing. Simultaneously, the immunosensor array acquires protein marker data and performs noise elimination. The data processing unit fuses the two data sets, performs wavelet denoising, and performs principal component analysis dimensionality reduction. The micropump provides flow rate control data for sample standardization. Finally, drift correction is performed in conjunction with the standard sample data to generate the initial drug signature dataset. The timing diagram clearly illustrates the data transfer relationship and processing sequence between the various functional modules. Specifically, data acquisition is performed using a micro-Raman spectroscopy sensor. Based on the principle of surface-enhanced Raman scattering, the micro-Raman spectroscopy sensor performs spectral analysis of drug molecules on a silver nanoparticle-modified substrate. When laser light irradiates the sample surface, the drug molecules produce characteristic scattering spectra that contain information such as molecular structure, chemical bonds, and functional groups. The raw spectral data collected by the sensor undergoes signal preprocessing, including baseline correction, background subtraction, and peak identification, to obtain the characteristic values ​​of the drug molecules. Baseline correction removes the fluorescence background through polynomial fitting, background subtraction eliminates interference from solvent and impurity signals, and peak identification extracts the position and intensity information of the characteristic Raman peak. The immunosensor array consists of 8 independent electrochemical sensing units, each of which is modified with specific antibodies to identify different protein markers. When proteins bind to antibodies, it causes changes in the electrochemical properties of the electrode surface, and the protein concentration is detected by measuring changes in current, potential, or impedance. The collected electrochemical signals contain various random noises, which require noise elimination processing. Median filtering is used to remove pulse noise, and then Kalman filtering is used to smooth the signal. Finally, threshold filtering is used to remove baseline drift to obtain the characteristic values ​​of the markers.

[0042] When fusing drug molecule and marker eigenvalues, the different characteristics and dimensions of the two data types must be considered. The two data sets are time-aligned and normalized to make them comparable. A weighted fusion method is then used to assign weights based on the reliability and importance of the different data types to generate the fused data. The fused data is denoised using a wavelet transform. Appropriate wavelet basis functions and decomposition levels are selected, and high-frequency noise is thresholded to retain the primary signal features, resulting in denoised data. Principal component analysis is then performed on the denoised high-dimensional data for dimensionality reduction. The covariance matrix of the data is calculated, and then the eigenvalues ​​and eigenvectors are solved. The eigenvectors are sorted by eigenvalue size, and the top eigenvectors with the largest contribution are selected to form a transformation matrix. This transformation matrix is ​​used to project the original data into a low-dimensional feature space, achieving data dimensionality reduction. During feature extraction, the focus is on retaining the primary variation in the data and removing redundancy and noise.

[0043] The reduced-dimensional data is standardized, including mean centering and variance normalization. The sampling flow rate is controlled by a micropump to ensure the stability and consistency of data acquisition. The micropump utilizes a piezoelectric drive principle, which can precisely control flow rate fluctuations within a preset range. Based on the characteristics of different sample types, appropriate flow rate parameters are set and calibration data is collected. Drift correction is performed on the calibration data to eliminate signal drift caused by long-term sensor use. A drift compensation model is established by regularly collecting data from standard samples. Based on the deviation between the theoretical and measured values ​​of the standard samples, a correction coefficient is calculated, and the test data is calibrated to obtain an accurate initial drug feature dataset.

[0044] For example, when processing a compound eye drop, the micro-Raman spectrometer sensor collected the characteristic peak of the main ingredient, ciprofloxacin hydrochloride, at 1380 cm -1 and 1620cm -1 A clear Raman scattering signal is detected at ^. Signal preprocessing removes background fluorescence, and the intensity ratio of these two characteristic peaks is extracted as the characteristic value of the drug molecule. Simultaneously, the immunosensor array detects protein markers in the solution, and the current signal, after noise removal, displays the changing trend of protein content. These two sets of data are fused and dimensionally reduced to produce a feature dataset that characterizes the drug's composition and quality. During data processing, a micropump precisely controls the flow rate to ensure sampling accuracy. Regular calibration with standard solutions ensures data reliability.

[0045] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0046] (1) Analyze the initial drug feature data set through surface-enhanced Raman scattering structure to obtain the spatial structure data of drug molecules;

[0047] (2) Perform molecular force field calculations on the spatial structure data of drug molecules to obtain molecular force data and drug interaction models;

[0048] (3) Identify binding sites of drug interaction models using graphene molecular imprinting materials to obtain chemical bond binding data;

[0049] (4) Calculate the charge distribution of chemical bond data to generate a molecular interaction strength matrix;

[0050] (5) Perform threshold analysis on the molecular interaction strength matrix to screen out high-risk interaction combinations;

[0051] (6) Conduct hazard assessment on high-risk interaction combinations to obtain drug interaction risk assessment data.

[0052] Specifically, the initial drug feature dataset was analyzed using a surface-enhanced Raman scattering (SERS) structure. This structure consists of a metal substrate modified with silver nanoparticles. When laser light irradiates the sample surface, the local electromagnetic field is enhanced, amplifying the Raman scattering signal. The scattering spectrum is analyzed by wavelength, and the positions and intensities of characteristic peaks are extracted. Combined with theoretical calculations of molecular vibrational modes, spatial configuration parameters such as bond lengths and bond angles of the drug molecules are derived, forming the spatial structural data of the drug molecules.

[0053] The spatial structure data of drug molecules are calculated through molecular force field to obtain the intermolecular interaction force. The molecular force field calculation is based on the following formula:

[0054]

[0055] Where: α b is the bond stretching constant; r is the bond length; r eq is the equilibrium bond length; β c is the bond angle bending force constant; θ is the bond angle; θ eq is the equilibrium bond angle; γ d is the torsional potential energy coefficient; is the dihedral angle; δ is the phase angle; η e is the electrostatic interaction coefficient; q i ,q j is the atomic charge; ∈ is the dielectric constant; r ij is the interatomic distance; λ f is the van der Waals force coefficient; σ ij is the minimum distance between atoms.

[0056] The calculated molecular force data is analyzed to establish a drug interaction model. Graphene molecularly imprinted materials possess specific recognition sites that selectively bind to drug molecules. Binding site identification is performed on the drug interaction model, and chemical bonding data is obtained by analyzing binding energy and spatial matching.

[0057] Chemical bonding data is used to calculate the charge distribution between molecules. The formula for calculating the charge density distribution is:

[0058]

[0059] Where: ρ(x, y, z) is the charge density at the spatial point (x, y, z); μ i is the charge contribution coefficient of the i-th atom; ψ i is the wave function of the i-th atom; ν j is the charge contribution coefficient of the j-th chemical bond; ξ j is the wave function of the jth chemical bond; ω k is the charge contribution coefficient of the kth molecular orbital; χ k is the wave function of the kth molecular orbital.

[0060] Based on the charge density distribution, a molecular interaction strength matrix was constructed. The matrix was subjected to eigenvalue decomposition, and threshold conditions were set to screen for high-risk interaction combinations. The hazard level of high-risk combinations was assessed using the following formula:

[0061]

[0062] Where: R is the hazard score; κ mn is the interaction weight between the mth and nth drugs; I mn is the interaction strength factor; S mn is the spatial matching coefficient; T mn is the time correlation factor; P and Q are the number of drug types.

[0063] For example, when processing compound eye drops containing antibiotic and anti-inflammatory ingredients, surface-enhanced Raman scattering structural analysis revealed overlapping characteristic peaks between the two ingredients at specific wavelengths, indicating the presence of an interaction. Molecular force field calculations revealed the presence of electrostatic attraction and van der Waals forces between them. Multiple binding sites were identified using graphene molecular imprinting materials, and binding energy calculations indicated that certain sites had strong interactions. Charge distribution calculations revealed the formation of a charge-dense region in the binding area, with the corresponding element value in the interaction strength matrix exceeding the preset threshold. Hazard assessment results indicate that this drug combination has a high risk of interaction and requires special attention during the filtration process.

[0064] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0065] (1) Decomposing the drug interaction risk assessment data in time series, encoding the data according to the medication time, and converting the encoded data through an encryption function to generate digital medication information;

[0066] (2) Perform multi-dimensional classification of digital medication information according to medication type and dosage, extract medication time intervals and medication duration periods, obtain medication regularity characteristics through data clustering, and generate medication frequency data;

[0067] (3) Perform spatiotemporal correlation analysis on medication frequency data, extract medication feature vectors based on drug action mechanisms, construct feature association matrices through feature mapping, and establish medication profiles;

[0068] (4) Encrypt the medication profile data in blocks, store them in a distributed manner through blockchain nodes, generate verification hash values, perform integrity verification on the stored data, and obtain securely stored data;

[0069] (5) Extracting temporal and spatial features from securely stored data, calculating feature weights based on the strength of drug interactions, and obtaining risk coefficients through multi-level evaluation;

[0070] (6) The filtering parameters are optimized with multi-objective constraints according to the risk factor, and the parameters are adjusted in combination with the filtering efficiency and safety threshold to obtain the personalized filtering scheme parameters.

[0071] Specifically, if Figure 3The figure shows a flow chart illustrating the process of performing time-series decomposition processing on drug interaction risk assessment data and then encoding the data according to medication time, in an embodiment of the present application. This diagram illustrates the complete conversion process from risk assessment data to digital medication information. First, the original risk assessment data is time-series decomposition processed, and temporal features are extracted by segmenting them into time windows. These features are then subjected to correlation analysis to generate a time-series code. A hierarchical coding structure is then used to encode the time information and risk information separately and then merge them. Finally, the merged encoded data is encrypted using the AES-256 encryption algorithm to generate digital medication information. During the monitoring of eye drop filtration equipment, the analysis and processing of drug interaction risk assessment data requires time-series decomposition. The time series is segmented into fixed time windows, and temporal features are extracted for each data segment, including medication time, duration, and interval. These temporal features are then associated with the corresponding risk assessment value to generate a time-series code. The time-series code uses a hierarchical coding structure, encoding the time information and risk information separately and then merging them. The encoded data is encrypted using the AES-256 encryption algorithm to generate digital medication information, ensuring data security. Processing digital medication information involves classification and analysis across multiple dimensions. A classification tree structure is established based on drug type, with each drug assigned to a corresponding category node based on its chemical composition and mechanism of action. A dosage classification system is also established based on dosage, and the specific dosage value for each medication is recorded. Medication time series are analyzed to extract statistical features of medication intervals, such as the mean interval and standard deviation, and calculate the duration of medication use. These features are clustered using the K-means clustering algorithm to identify different medication patterns and generate medication frequency data.

[0072] The spatiotemporal correlation analysis of medication frequency data involves multiple layers. Temporal correlation analysis primarily examines the periodicity and continuity of medication use patterns, extracting relevant features through time series analysis. Spatial correlation focuses on the associations between medication use, using association rule mining algorithms to discover drug combination patterns. Incorporating the drug's mechanism of action, a medication feature vector is constructed, encompassing drug attributes, usage patterns, and interaction characteristics. Feature mapping methods are used to transform these features into an association matrix, which describes the strength of relationships between different features, thereby creating a medication profile. Blockchain technology is used to securely store medication profile data. Data is divided into fixed-size blocks, each containing a timestamp, data content, and a hash value of the previous block. Blockchain blocks are hashed using the SHA-256 algorithm and signed using an asymmetric encryption algorithm. Blockchain nodes utilize a distributed storage architecture, with each node maintaining a complete copy of the data. Data consistency is ensured through a consensus mechanism, and data integrity is verified by verifying the hash value chain, resulting in secure data storage.

[0073] The temporal features extracted from the securely stored data include medication timing patterns and periodic features, while the spatial features include the strength of association between drugs and combination characteristics. The weight coefficients of each feature are calculated in combination with the previously obtained drug interaction strength values. A multi-level evaluation method is used to evaluate the risk at the level of a single feature, and then the comprehensive risk is evaluated at the level of the feature combination to obtain a risk coefficient that reflects the overall safety of medication. When optimizing the filtering parameters based on the risk coefficient, a constrained optimization problem with multiple objectives is established. The optimization objectives include filtration efficiency, drug residue, degree of cross-contamination, etc., and the constraints include the operating range of the equipment, safety thresholds, etc. The optimal parameter combination is solved through a multi-objective optimization algorithm, and the parameters are dynamically adjusted according to the real-time monitoring results to obtain personalized filtering solution parameters.

[0074] For example, when processing a batch of compound eye drops, a hospital's ophthalmology department analyzed drug interaction risk assessment data on an hourly basis, recording drug usage during different time periods. The data showed that antibiotics and steroids were frequently used during the same time period. A multidimensional analysis of this data revealed that the intervals between use of these two types of drugs were short, and that dosage changes followed specific patterns. A medication profile constructed based on these characteristics indicated a high risk of interaction. The system then adjusted filtration parameters, increasing the number of filtration layers and adjusting the flow rate to ensure effective separation of the two types of drugs. Through real-time monitoring and parameter optimization, the goal of safe filtration was achieved.

[0075] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0076] (1) Decomposing the parameters of the personalized filtration scheme according to flow rate, pressure, and temperature, limiting the dynamic range of the decomposed parameters, and obtaining the adjustment parameters of the filtration system;

[0077] (2) setting the quantum dot fluorescence labeling concentration according to the filter system adjustment parameters, collecting and calculating the intensity of the quantum dot fluorescence signal, and obtaining the pollution tracing data;

[0078] (3) Threshold judgment is performed on the pollution tracer data, the areas exceeding the threshold are marked and pollution indicators are extracted to obtain the cross-contamination assessment results;

[0079] (4) Divide the contaminated areas according to the cross-contamination assessment results, set the pulse frequency and intensity for each area, and obtain the cleaning control parameters;

[0080] (5) Inputting cleaning control parameters into the pulse cleaning system, monitoring the water quality changes during the cleaning process in real time, and obtaining cleaning effect data;

[0081] (6) Conduct a comprehensive multi-parameter analysis of the cleaning effect data, conduct a status assessment based on the residual pollutants and water quality indicators, and obtain the equipment operation status data.

[0082] Specifically, when monitoring eye drops filtration equipment, the parameters of the personalized filtration solution need to be decomposed into three dimensions according to their physical properties: flow rate parameters, pressure parameters, and temperature parameters. The parameter decomposition process follows the following formula:

[0083]

[0084] Where: P adj Adjust the parameters for the filtration system; W i is the weight coefficient of each parameter; V i is the current parameter value; V min , V max is the parameter range limit; F i (t) is the time response function; H i (x) is the spatial distribution function; i is the parameter type (1-flow rate, 2-pressure, 3-temperature).

[0085] Then, the concentration of quantum dot fluorescent markers is calculated based on the filter system adjustment parameters. The quantum dot fluorescence signal intensity is calculated using the following formula:

[0086]

[0087] Among them, I qd is the fluorescence signal intensity of quantum dots; U j is the quantum dot excitation intensity; D j is the diffusion distance; R j is the characteristic diffusion radius; Y j is the fluorescence yield; T j is the signal acquisition time; S j is the response time constant; Q j is the quantum dot concentration; j is the quantum dot type number.

[0088] The following evaluation formula is used for the comprehensive multi-parameter analysis of cleaning effect data:

[0089]

[0090] Where: E status Score the equipment operating status; B k is the pollutant residual coefficient; C k is the residual measurement value; L k is the weight of water quality index; Z k is the water quality parameter value; X k (t) is the time decay function; mis the weight of the operating parameters; G m is the performance index value; k is the pollutant type; m is the performance parameter number.

[0091] In actual application, after receiving the parameters of the personalized filtration solution, the filtration system performs parameter decomposition. The flow rate parameter is set within a range of 0.1-2.0 mL / s; the pressure parameter range is 0.05-0.5 MPa; and the temperature parameter range is 15-35°C. Based on these parameters, the filtration system adjustment parameters are calculated and the quantum dot labeling concentration is set accordingly. The quantum dots used are CdSe / ZnS core-shell quantum dots, and fluorescence signals are collected using a fluorescence microscope. When the fluorescence intensity in a certain area exceeds a threshold, it is marked as contaminated, and the contamination index for that area is calculated. Based on the results of the cross-contamination assessment, the filter is divided into different cleaning zones, and corresponding pulse cleaning parameters are set for each zone. Cleaning control parameters include pulse frequency (0.5-10 Hz) and intensity (0.1-1.0 MPa). During the cleaning process, water quality indicators such as turbidity, pH, and conductivity are monitored in real time, and cleaning performance data is recorded. Finally, the cleaning performance data is comprehensively analyzed, combined with the results of pollutant residual measurement and changes in water quality indicators, to evaluate the equipment's operating status.

[0092] For example, when a certain ophthalmology hospital used this filtration equipment to process compound eye drops, the equipment set the flow rate to 0.5 mL / s, the pressure to 0.2 MPa, and the temperature to 25°C according to the personalized filtration plan. A quantum dot tracer with a concentration of 10 nM was then injected, and fluorescence detection revealed a strong fluorescent signal in the middle area of ​​the filter, indicating cross-contamination. Based on this, the filter was divided into three cleaning areas. The middle area was cleaned with a pulse of 5 Hz frequency and 0.8 MPa intensity, while the two end areas were cleaned with a pulse of 2 Hz frequency and 0.3 MPa intensity. During the cleaning process, the turbidity was monitored to decrease from the initial 5 NTU to 0.5 NTU, and the pH value and conductivity gradually returned to the normal range. These data were combined to evaluate the equipment's operating status.

[0093] In a specific embodiment, the process of executing step S105 may specifically include the following steps:

[0094] (1) Segment the equipment operating status data according to the time series, extract the characteristics of operating parameter changes, and establish a drug safety model;

[0095] (2) Perform trend analysis on the parameter change curve in the drug safety model, perform statistics on the fluctuation range, and obtain the safety threshold range;

[0096] (3) Numerical classification of the safety threshold range is performed, and the risk probability is calculated in combination with the parameter fluctuation frequency to obtain the risk level classification standard;

[0097] (4) Classify the operating data according to the risk level classification standard, calculate the distribution characteristics of the data at each level, and obtain the risk distribution map;

[0098] (5) Extract key monitoring indicators based on the risk distribution map, perform time series correlation analysis on the monitoring data, and obtain monitoring data characteristics;

[0099] (6) Comprehensively evaluate the monitoring data characteristics and risk levels to generate operating status assessment results and obtain an equipment monitoring assessment report.

[0100] Specifically, an in-depth analysis of equipment operating status data is conducted. Equipment operating status data is segmented according to time series, using a fixed time window method to divide the data into several segments. Each segment contains operating parameters such as flow rate, pressure, temperature, pH value, and conductivity. Variation characteristics of these parameter data are extracted, including statistical features such as mean, variance, rate of change, and periodicity. The so-called drug safety model is essentially a multi-parameter data analysis framework, a data structure containing information such as parameter variation patterns, threshold ranges, and correlations. Trend analysis is performed on the variation curves of each parameter in the drug safety model, and the parameter variation trends are calculated using a sliding window method. Statistics are collected for the fluctuation range of each parameter, recording statistical quantities such as maximum value, minimum value, and standard deviation. By analyzing historical parameter fluctuation data and combining it with the operating characteristics of the equipment, the safe operating range of each parameter is determined. These safe threshold ranges constitute the boundary conditions for the normal operation of the equipment.

[0101] Safety threshold ranges are divided into multiple levels based on the degree of deviation. For example, the degree of deviation of parameter values ​​from the normal range is divided into three levels: mild, moderate, and severe. The probability of each level occurring is calculated by combining information on the frequency of parameter fluctuations. The frequency of parameter fluctuations is obtained by statistically analyzing abnormal events in historical data, which is then used to establish risk level classification criteria. The risk level classification criteria include information on two dimensions: the degree of parameter deviation and the frequency of occurrence. Real-time operating data is classified according to the risk level classification criteria, and the temporal and spatial distribution characteristics of data at different risk levels are statistically analyzed. Through data cluster analysis, high-risk areas and high-incidence time periods are identified, and this information is integrated into a risk distribution map. The risk distribution map intuitively displays the risk distribution during equipment operation.

[0102] Key monitoring indicators are extracted from the risk distribution map, including high-risk parameters and those prone to volatility. Time-series correlation analysis is performed on this monitoring data, using correlation analysis methods to identify correlations between parameters. By analyzing the sequence of parameter changes and their mutual influence, characteristic patterns of the monitoring data are identified. A comprehensive assessment of the monitoring data characteristics and risk levels is conducted, encompassing aspects such as parameter stability, risk level, and trend of change. Based on the assessment results, an equipment monitoring assessment report is generated, including fault warnings and maintenance recommendations.

[0103] For example, when processing a batch of compound eye drops containing antibiotics and anti-inflammatory drugs, equipment operating status data showed periodic fluctuations in filtration pressure over a certain period of time. Time series analysis revealed a clear correlation between pressure fluctuations and changes in drug concentration. The pressure data was segmented into hourly time windows, and features such as the pressure change rate and fluctuation amplitude were extracted. Analysis revealed that pressure fluctuations exceeding safety thresholds primarily occurred during periods of high drug concentration. Risk levels were categorized into three levels based on the amplitude and frequency of pressure fluctuations. Data analysis revealed that filtration effectiveness decreased significantly when the pressure fluctuation frequency exceeded three times per hour and the amplitude exceeded 30%. These features were recorded in a risk distribution map and used as key monitoring indicators. Based on these data features, a monitoring and assessment report was generated, which included recommendations for adjusting filtration parameters and increasing maintenance frequency.

[0104] In a specific embodiment, the process of executing step S106 may specifically include the following steps:

[0105] (1) Perform data segmentation processing on the equipment monitoring and evaluation report, extract the equipment performance parameter change trend from it, and obtain the performance degradation curve;

[0106] (2) Compare and analyze the performance degradation curve with historical operation data, calculate the parameter deviation, and obtain the optimization solution;

[0107] (3) Conduct sensitivity analysis on the equipment parameters in the optimization scheme, screen the parameters based on the operational stability requirements, and obtain a parameter update scheme;

[0108] (4) Correlate the parameter update plan with the equipment operation time, evaluate the life of key components, and obtain component status data;

[0109] (5) Calculate the failure probability distribution based on component status data, predict the maintenance time point, and obtain the maintenance cycle parameters;

[0110] (6) Integrate and calculate the maintenance cycle parameters and parameter update scheme, optimize and adjust the control strategy, and obtain intelligent management and control instructions.

[0111] Specifically, during the monitoring process of the eye drops filtration equipment, the data in the equipment monitoring and evaluation report is segmented in time series. The data is segmented according to a fixed time window (for example, every 24 hours), and the equipment performance parameters are extracted from each segment of data, including flow rate stability, filtration accuracy, cleaning efficiency, etc. These parameters are analyzed in time series, and the sliding average method is used to eliminate short-term fluctuations and extract long-term change trends. The changes in parameters over time are plotted into a curve to reflect the attenuation of equipment performance with time of use, and a performance attenuation curve is obtained. The newly generated performance attenuation curve is compared with the data in the equipment historical operation database. The historical data contains records of performance changes of the equipment under different working conditions. By calculating the deviation between the current performance parameters and the historical data for the same period, the parameters with faster performance degradation are found. The causes of the parameters with large deviations are analyzed, and targeted optimization plans are formulated in combination with factors such as the working environment and usage intensity. The optimization plan includes the parameter items that need to be adjusted, the adjustment direction and the adjustment range.

[0112] Conduct a sensitivity analysis on each device parameter in the optimization plan to evaluate the impact of each parameter change on the overall performance of the equipment. Using the single-factor analysis method, change each parameter one by one and record the changes in equipment performance. Set parameter adjustment constraints based on the equipment's operational stability requirements. Based on the sensitivity analysis results and constraints, screen out the most critical parameters and form a parameter update plan. The parameter update plan specifies in detail the adjustment values ​​and adjustment sequence for each parameter. Correlate the parameter update plan with the actual operating time of the equipment. Count the cumulative operating time of each key component, including the filter membrane, pump body, sensor, etc. Evaluate the degree of performance degradation based on the changes in the component's operating parameters. By comparing the parameter values ​​under normal operating conditions, calculate the performance deviation, evaluate the current status of each component, and obtain component status data. Component status data includes information such as the degree of performance degradation and remaining service life.

[0113] Calculate the failure probability distribution based on component status data. Establish a failure prediction model by analyzing the component's workload, operating environment, and performance degradation trends. Calculate the probability of failure at different time points based on the component's historical failure records and current status parameters. Determine the optimal maintenance time points by combining the failure probability distribution and maintenance costs. Organize these time points and corresponding maintenance items into maintenance cycle parameters. Integrate the maintenance cycle parameters with the parameter update plan. Assign parameter adjustment tasks to each maintenance cycle based on the maintenance time points. Optimize and adjust the control strategy based on the equipment's work schedule and maintenance resource constraints. Integrate the adjusted parameter setting values, maintenance plan, and other information into intelligent management and control instructions.

[0114] For example, a hospital's ophthalmology department's filtration equipment processes compound eye drops. After the equipment had been operating for a period of time, the monitoring system extracted data on filtration pressure changes from evaluation reports. Time series analysis revealed that the rate of increase in filtration pressure was gradually accelerating, while the pressure recovery effect after cleaning was gradually decreasing. Comparing this trend with historical data revealed that the pressure rise rate exceeded the normal range. A sensitivity analysis determined that flow rate and cleaning intensity were the key parameters influencing pressure changes. Combined with an analysis of the filter membrane's service life, it was found that the filter membrane was nearing its design lifespan and its performance had begun to decline significantly. Based on this data, it was predicted that the probability of the filter membrane clogging would increase significantly within the next two weeks. Consequently, control instructions were issued, including adjusting the flow rate, increasing the cleaning frequency, and prematurely replacing the filter membrane. This example demonstrates the complete logical chain from data analysis to generating control instructions.

[0115] In a specific embodiment, the process of performing correlation analysis between the parameter update plan and the equipment operation time and evaluating the life of key components may specifically include the following steps:

[0116] (1) Classify the operating parameters in the parameter update scheme according to functional modules, extract the parameter change rules of different modules, and obtain the module operation characteristics;

[0117] (2) Decompose the module operation characteristics in time series, calculate the parameter change rate based on the equipment operation time, and obtain the operation trend data;

[0118] (3) Analyze the fluctuation cycle of the operation trend data, calculate the fluctuation amplitude, and obtain the stability evaluation value;

[0119] (4) Calculate component wear based on the stability assessment value, quantitatively analyze the wear rate, and obtain life prediction parameters;

[0120] (5) Correlate the life prediction parameters with the operating load, predict the trend of component performance degradation, and obtain the performance degradation curve;

[0121] (6) Fit and analyze the performance degradation curve, conduct status evaluation based on component operating parameters, and obtain component status data.

[0122] Specifically, the operating parameters in the parameter update scheme are classified into functional modules. The parameters are divided into filtration modules (including parameters such as flow rate, pressure, pore size, etc.), cleaning modules (including parameters such as pulse intensity, frequency, duration, etc.) and monitoring modules (including parameters such as water quality, turbidity, pH value, etc.). Time series analysis is performed on the parameters of each module to extract the trend characteristics, periodic characteristics and mutation characteristics of parameter changes to obtain module operation characteristics. The module operation characteristics are decomposed in time series, and the wavelet analysis method is used to decompose the data into trend items and fluctuation items of different scales. Combined with the actual operating time of the equipment, the rate of change of the parameters is calculated, including the growth rate, attenuation rate and fluctuation rate of the parameter value. These change rate data are arranged in chronological order to form operation trend data.

[0123] Perform a periodic analysis of operating trend data, using Fourier transforms to identify the primary periodic components within the data. The amplitude and frequency of fluctuations across different periods are statistically analyzed, and the parameter's fluctuation intensity index is calculated. Based on the fluctuation intensity index and the parameter's standard range, the stability of the equipment's operation is assessed, resulting in a stability evaluation value.

[0124] Component wear is calculated using the following formula:

[0125]

[0126] Where: W total is the total wear amount; A i is the wear reference coefficient; t i is the running time; M i is the material characteristic coefficient; N i is the load factor; P i is the working pressure; V i is the flow rate; K i is the durability coefficient; J i is the wear index; R i (t) is the time correction function; i is the component number.

[0127] The performance degradation trend is predicted using the following formula:

[0128]

[0129] Where: D(t) is the performance attenuation function; Q j is the performance weight coefficient; L j is the cumulative load; U j is the load threshold; S j is the stress coefficient; H j is the working hours; X j is the design life; Y j is the decay index; Z j (t) is the environmental impact function; j is the performance index number.

[0130] Performance degradation curves were fitted and analyzed, using polynomial regression to approximate performance trends. A comprehensive assessment of the component's current status was conducted based on its operating parameters, including operating load, environmental conditions, and maintenance records, generating component status data. For example, in an eye drop filtration device at an ophthalmology hospital, the pressure parameters of the filter module exhibited periodic fluctuations. Time series analysis revealed two primary cycles: a short 4-hour fluctuation (corresponding to the solution replacement cycle) and a long 24-hour fluctuation (corresponding to daily usage patterns). Based on the filter membrane's usage history, the calculated pressure rise rate gradually increased, indicating that the filter membrane was slowly clogging. Substituting this data into a wear calculation formula, the accumulated wear was calculated, taking into account factors such as the membrane's material properties, operating pressure, and flow rate. Using the performance degradation prediction formula and the equipment's usage load, the remaining service life of the filter membrane was predicted under the current usage intensity. The resulting component status data indicated that the filter membrane's performance had entered a rapid degradation phase, necessitating prompt replacement and maintenance.

[0131] The above describes the eye drops filtration device monitoring method based on the Internet of Things in the embodiment of the present application. The following describes the eye drops filtration device monitoring system based on the Internet of Things in the embodiment of the present application. Figure 4 In the embodiment of the present application, an embodiment of the eye drops filtration equipment monitoring system based on the Internet of Things includes:

[0132] The acquisition module 201 is used to collect drug molecular fingerprints and protein marker data through a multi-dimensional sensor, and perform dimensionality reduction processing on the collected data to obtain an initial drug feature data set;

[0133] A construction module 202 is configured to construct a drug interaction model based on the initial drug feature data set, extract features of the drug molecular structure, and obtain drug interaction risk assessment data;

[0134] The storage module 203 is used to digitally store medication information based on the drug interaction risk assessment data, establish a medication profile and assess the risk factor, and obtain personalized filtering scheme parameters;

[0135] An adjustment module 204 is configured to dynamically adjust parameters based on the personalized filtration solution, monitor cross contamination using quantum dot tracing, and obtain device operating status data through pulse cleaning.

[0136] A classification module 205 is used to establish a drug safety model based on the equipment operation status data, classify the risk levels and integrate the monitoring data to obtain an equipment monitoring assessment report;

[0137] The prediction module 206 is used to extract optimization solutions based on the equipment monitoring and evaluation report, update equipment parameters, predict maintenance cycles, and obtain intelligent management and control instructions.

[0138] Through the collaborative efforts of these components, multi-dimensional sensors collect drug molecular fingerprints and protein marker data. Combined with wavelet transform denoising and principal component analysis dimensionality reduction, this system accurately captures changes in drug composition, effectively reduces data noise, and improves data analysis accuracy. A drug interaction model is constructed based on the initial drug feature dataset, and feature extraction using a graph neural network enables precise analysis of drug molecular structures and their interactions, providing reliable data support for subsequent risk assessment. During the digital storage of medication information, blockchain technology and deep learning algorithms are combined to create medication profiles, ensuring data security and traceability while also enabling accurate assessment of medication risks. Quantum dot tracing technology provides real-time monitoring of cross-contamination, combined with pulsed cleaning technology, to establish a comprehensive contamination prevention and control mechanism, significantly reducing cross-contamination risks. A drug safety model, built based on equipment operating status data, employs a multi-level assessment approach to categorize risk levels, enabling comprehensive monitoring of equipment operating status. By analyzing equipment monitoring and evaluation reports and developing optimization plans, coupled with reinforcement learning algorithms, equipment parameters are dynamically adjusted to continuously optimize equipment performance. In particular, this solution integrates multiple deep learning algorithms and traditional data processing methods to establish an intelligent monitoring and decision-making framework. The application of graph neural networks in drug interaction analysis significantly improves the accuracy of molecular structure feature extraction; the application of deep learning algorithms in the construction of medication profiles enables precise assessment of medication risks; and the application of reinforcement learning algorithms in parameter optimization enables adaptive adjustment of device control strategies. This organic combination of algorithmic features not only enhances the system's intelligence but also enables refined management of the entire lifecycle of eye drop filtration equipment.

[0139] The present application also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions, and when the instructions are executed on a computer, the computer executes the steps of the eye drops filtration equipment monitoring method based on the Internet of Things.

[0140] Those skilled in the art will 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 aforementioned method embodiments and will not be repeated here.

[0141] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for monitoring eye drops filtration equipment based on the Internet of Things, characterized in that: The eye drops filtration equipment monitoring method based on the Internet of Things includes: The drug molecular fingerprint and protein marker data are collected through multi-dimensional sensors, and the collected data are processed by dimensionality reduction to obtain the initial drug feature data set; Constructing a drug interaction model based on the initial drug feature data set, extracting features of the drug molecular structure, and obtaining drug interaction risk assessment data; Based on the drug interaction risk assessment data, medication information is digitally stored, a medication profile is established, and risk factors are assessed to obtain personalized filtering solution parameters; Dynamic adjustment is performed based on the parameters of the personalized filtration solution, cross contamination is monitored using quantum dot tracing, and equipment operating status data is obtained through pulse cleaning treatment; Establishing a drug safety model based on the equipment operating status data, classifying risk levels and integrating monitoring data to obtain an equipment monitoring assessment report; An optimization plan is extracted based on the equipment monitoring and evaluation report, equipment parameters are updated and maintenance cycles are predicted, and intelligent management and control instructions are obtained.

2. The method for monitoring eye drops filtration equipment based on the Internet of Things according to claim 1, characterized in that: The method collects drug molecular fingerprints and protein marker data through a multidimensional sensor, performs dimensionality reduction processing on the collected data, and obtains an initial drug feature data set, including: The drug molecule fingerprint spectrum is collected by a micro Raman spectroscopy sensor, and the drug molecule fingerprint spectrum is preprocessed to obtain the drug molecule characteristic value; Collecting protein marker data through an immunosensor array, performing noise elimination on the protein marker data, and obtaining marker characteristic values; Fusing the drug molecule characteristic values ​​and the marker characteristic values, and denoising the fused data by wavelet transform to obtain denoised data; Performing principal component analysis and dimensionality reduction processing on the denoised data, and obtaining dimensionality-reduced data by feature extraction; Performing data standardization on the dimension-reduced data, and controlling flow rate sampling by a micropump to obtain calibration data; The calibration data is subjected to data drift correction and calibrated with a data standard sample to obtain the initial drug characteristic data set.

3. The method for monitoring eye drops filtration equipment based on the Internet of Things according to claim 1, characterized in that: The drug interaction model is constructed based on the initial drug feature data set, and feature extraction is performed on the drug molecular structure to obtain drug interaction risk assessment data, including: Analyzing the initial drug characteristic data set through surface enhanced Raman scattering structure to obtain drug molecular spatial structure data; Performing molecular force field calculation on the spatial structure data of the drug molecules to obtain molecular force data and a drug interaction model; identifying binding sites of the drug interaction model using graphene molecular imprinting materials to obtain chemical bond binding data; Performing charge distribution calculation on the chemical bonding data to generate a molecular interaction strength matrix; Performing threshold analysis on the molecular interaction strength matrix to screen out high-risk interaction combinations; The high-risk effect combination is subjected to a hazard assessment to obtain the drug interaction risk assessment data.

4. The method for monitoring eye drops filtration equipment based on the Internet of Things according to claim 1, characterized in that: According to the drug interaction risk assessment data, the medication information is digitally stored, a medication profile is established and the risk coefficient is assessed to obtain personalized filtering scheme parameters, including: Decomposing the drug interaction risk assessment data in time series, encoding the data according to medication time, and converting the encoded data through an encryption function to generate digital medication information; Perform multi-dimensional classification on the digital medication information according to medication type and medication dosage, extract medication time interval and medication duration, obtain medication regularity characteristics through data clustering, and generate medication frequency data; Performing spatiotemporal correlation analysis on the medication frequency data, extracting medication feature vectors based on the drug action mechanism, constructing a feature association matrix through feature mapping, and establishing a medication profile; The medication portrait data is encrypted in blocks, distributedly stored through blockchain nodes and a verification hash value is generated, and the stored data is integrity checked to obtain securely stored data; Extracting temporal and spatial features from the securely stored data, calculating feature weights based on the strength of drug interactions, and obtaining risk coefficients through multi-level evaluation; The filtering parameters are optimized with multi-objective constraints according to the risk coefficient, and the parameters are adjusted in combination with the filtering efficiency and the safety threshold to obtain the personalized filtering scheme parameters.

5. The method for monitoring eye drops filtration equipment based on the Internet of Things according to claim 1, characterized in that: The parameters of the personalized filtration scheme are dynamically adjusted, cross contamination is monitored using quantum dot tracing, and the equipment operation status data is obtained through pulse cleaning, including: Decomposing the personalized filtration scheme parameters according to flow rate, pressure, and temperature, and limiting the dynamic range of the decomposed parameters to obtain the filtration system adjustment parameters; Setting the quantum dot fluorescence labeling concentration according to the filter system adjustment parameters, collecting and calculating the intensity of the quantum dot fluorescence signal, and obtaining pollution tracing data; Performing threshold judgment on the pollution tracer data, marking areas exceeding the threshold and extracting pollution indicators to obtain cross-contamination assessment results; Divide the contaminated areas according to the cross-contamination assessment results, set the pulse frequency and intensity for each area, and obtain cleaning control parameters; The cleaning control parameters are input into the pulse cleaning system to monitor the water quality changes during the cleaning process in real time to obtain cleaning effect data; A multi-parameter comprehensive analysis is performed on the cleaning effect data, and a status assessment is performed in combination with the pollutant residue and water quality indicators to obtain the equipment operation status data.

6. The method for monitoring eye drops filtration equipment based on the Internet of Things according to claim 1, characterized in that: The drug safety model is established based on the equipment operation status data, the risk levels are divided and the monitoring data is integrated to obtain an equipment monitoring assessment report, including: Segmenting the equipment operating status data according to time series, extracting operating parameter change characteristics, and establishing a drug safety model; Performing trend analysis on the parameter change curve in the drug safety model, performing statistics on the fluctuation range, and obtaining a safety threshold range; Numerical grading is performed on the safety threshold range, and the risk probability is calculated in combination with the parameter fluctuation frequency to obtain a risk level classification standard; Classify the operating data according to the risk level classification standard, calculate the distribution characteristics of the data at each level, and obtain a risk distribution map; Extract key monitoring indicators based on the risk distribution map, perform time series correlation analysis on the monitoring data, and obtain monitoring data features; The monitoring data characteristics and risk levels are comprehensively evaluated to generate an operation status evaluation result and obtain the equipment monitoring evaluation report.

7. The method for monitoring eye drops filtration equipment based on the Internet of Things according to claim 1, characterized in that: Extracting an optimization plan based on the equipment monitoring and evaluation report, updating equipment parameters and predicting maintenance cycles, and obtaining intelligent management and control instructions include: Performing data segmentation processing on the equipment monitoring and evaluation report, extracting the equipment performance parameter change trend therefrom, and obtaining a performance attenuation curve; Comparing and analyzing the performance decay curve with historical operating data, calculating parameter deviations, and obtaining an optimization solution; Perform sensitivity analysis on the equipment parameters in the optimization scheme, screen the parameters in combination with the operational stability requirements, and obtain a parameter update scheme; Correlation analysis is performed between the parameter update scheme and the equipment operation time, and the life of key components is evaluated to obtain component status data; Calculate the failure probability distribution based on the component status data, predict the maintenance time point, and obtain the maintenance cycle parameters; The maintenance cycle parameters and the parameter update plan are integrated and calculated, and the control strategy is optimized and adjusted to obtain the intelligent management and control instructions.

8. The method for monitoring eye drops filtration equipment based on Internet of Things according to claim 7, characterized in that: The parameter update scheme is correlated with the equipment operation time and analyzed to evaluate the life of key components to obtain component status data, including: The operating parameters in the parameter update scheme are classified according to functional modules, and the parameter change patterns of different modules are extracted to obtain module operation characteristics; Decompose the module operation characteristics in time series, calculate the parameter change rate based on the equipment operation time, and obtain the operation trend data; Performing a fluctuation cycle analysis on the operation trend data, performing statistics on the fluctuation amplitude, and obtaining a stability evaluation value; Calculate component wear based on the stability evaluation value, perform quantitative analysis on the wear rate, and obtain life prediction parameters; Correlating the life prediction parameters with the operating load, performing trend prediction on component performance degradation, and obtaining a performance degradation curve; The performance degradation curve is fitted and analyzed, and a status assessment is performed in combination with component operating parameters to obtain component status data.

9. An eye drops filtration equipment monitoring system based on the Internet of Things, used to implement the eye drops filtration equipment monitoring method based on the Internet of Things as described in any one of claims 1 to 8, characterized in that: The eye drops filtration equipment monitoring system based on the Internet of Things includes: The acquisition module is used to collect drug molecular fingerprints and protein marker data through multi-dimensional sensors, and perform dimensionality reduction processing on the collected data to obtain the initial drug feature data set; A construction module is used to construct a drug interaction model based on the initial drug feature data set, perform feature extraction on the drug molecular structure, and obtain drug interaction risk assessment data; A storage module is used to digitally store medication information based on the drug interaction risk assessment data, establish a medication profile and assess the risk factor, and obtain personalized filtering solution parameters; An adjustment module is used to dynamically adjust the parameters of the personalized filtration solution, monitor cross contamination using quantum dot tracing, and obtain equipment operating status data through pulse cleaning treatment; A classification module is used to establish a drug safety model based on the equipment operation status data, classify the risk levels and integrate the monitoring data to obtain an equipment monitoring evaluation report; The prediction module is used to extract optimization solutions based on the equipment monitoring and evaluation report, update equipment parameters and predict maintenance cycles to obtain intelligent management and control instructions.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the method for monitoring eye drops filtration equipment based on the Internet of Things as described in any one of claims 1 to 8 is implemented.

Images