Digital twinning-based insulin pump set health state monitoring method

Abstract

The invention belongs to the field of insulin pump set health state monitoring, and particularly relates to an insulin pump set health state monitoring method based on digital twinning, which adopts a detection system comprising a data set unit, a mechanism model set unit, a digital twinning model integration unit, a digital twinning model consistency evaluation unit and an insulin pump set fault monitoring unit. A high-fidelity blood glucose regulation and control digital twinborn model is constructed based on a virtual model mapped by an insulin pump set entity and a human body blood glucose mechanism and historical monitoring data through a hierarchical digital twinborn construction framework, and the health state of the insulin pump set is monitored through the combination of the high-fidelity digital twinborn model and real-time blood glucose monitoring data.

Description

Technical Field

[0001] The present invention belongs to the field of insulin pump group health status monitoring, and specifically relates to a digital twin-based insulin pump group health status monitoring method. Background Art

[0002] Diabetes is a disease caused by a series of metabolic disorders such as sugar, protein, fat, water and electrolytes due to absolute or relative lack of insulin secretion and reduced sensitivity of target tissue cells to insulin. As the most advanced insulin infusion device, insulin pump can continuously push insulin into the user's subcutaneous tissue according to the dosage required by the human body, maintain blood sugar stability throughout the day, and achieve the purpose of controlling diabetes. However, when the insulin pump group fails, it is likely to cause serious consequences such as ketoacidosis in patients. The insulin pump consists of a pump, a small syringe and an infusion tube connected to it. Common failures include mechanical jamming of the pump body, leakage and blockage of the catheter and needle. At present, insulin pumps mainly rely on cavity pressure signals and infusion control signals for health status monitoring, but due to the low insulin infusion rate and slow pressure change, it cannot achieve good results.

[0003] Digital twin is a technology that combines the digital model of a physical system with real-time data. It completes the mapping in the virtual space by integrating multi-disciplinary, multi-physical quantity, multi-scale, and multi-probability simulation processes. It monitors the physical world through the interaction and iteration between the virtual world and the physical world to reflect the entire life cycle of the corresponding physical equipment. In recent years, digital twin technology has been applied to more and more fields, including aerospace, manufacturing, medical care, cities, energy, etc. Especially in fault diagnosis, digital twin technology can monitor and analyze the health status of equipment in a dynamic environment in real time, which helps to improve the efficiency and accuracy of fault diagnosis. Summary of the invention

[0004] The purpose of the present invention is to provide a method for monitoring the health status of an insulin pump group based on digital twins. The method uses a hierarchical digital twin construction framework to construct a high-fidelity blood glucose regulation digital twin model based on a virtual model that maps the insulin pump group entity to the human blood glucose mechanism and historical monitoring data. The health status of the insulin pump group is monitored by combining the high-fidelity digital twin model with real-time blood glucose monitoring data.

[0005] A digital twin-based insulin pump group health status monitoring method adopts a detection system including a data set unit, a mechanism model set unit, a digital twin model integration unit, a digital twin model consistency evaluation unit, and an insulin pump group fault monitoring unit; The data set unit collects physiological data and stores it in a database table; The digital twin model integration unit is based on a unified interface integration mechanism model; using the human blood glucose monitoring data of the insulin pump group in a healthy state in the database, Bayesian estimation is used to obtain the dynamic parameters in the integration model, and a digital twin model of the insulin pump group in a healthy state is constructed; The digital twin model consistency evaluation unit performs consistency evaluation on the mechanism model, obtains a comprehensive consistency evaluation result of the digital twin model based on the hierarchical analysis method, adjusts the digital twin model based on the consistency evaluation result to meet the consistency evaluation requirements, and obtains a high-fidelity digital twin model; The insulin pump group fault monitoring unit is based on a high-fidelity digital twin model, monitors blood glucose data in real time, and uses an unscented Kalman filter algorithm to dynamically estimate fault parameters.

[0006] Furthermore, the data set unit collects physiological data and stores it in a database table.

[0007] Furthermore, the mechanism model set unit includes an insulin pump group geometry model, an insulin fluid dynamics model, an insulin infusion rule model, and a human blood glucose metabolism model.

[0008] Furthermore, the digital twin model integration unit integrates the above-mentioned mechanism model based on the FMI interface, and uses the ODBC application programming interface to access the MySQL database in the above-mentioned data set unit; uses the human blood glucose monitoring data of the insulin pump group in a healthy state to adopt Bayesian estimation to obtain the dynamic parameters in the integrated model, and constructs a digital twin model of the insulin pump group in a healthy state.

[0009] Furthermore, the digital twin model consistency evaluation unit; determines and measures the characteristic elements to be measured of the geometric model of the insulin pump group, and also measures the characteristic elements of the insulin pump entity, and calculates the consistency degree of the two; sets the physical consistency evaluation index Cd(PM) for the insulin fluid dynamics model and the human blood glucose metabolism model; through simulation comparison experiments, compares the insulin dose output by the insulin pump group entity and the insulin fluid dynamics model under the same conditions, n1 is the total number of simulation experiments, m1 is the number of simulation experiments in which the error between the entity and the virtual model output insulin dose is less than 10%; calculates the consistency evaluation index R of the insulin pump group fluid dynamics model. re11 =m1 / n1; In addition, the blood glucose values ​​of type Ⅰ diabetic patients wearing insulin pumps and the human blood glucose model established in this paper were tested under the same dose of insulin infusion. The error within the allowable range was considered to meet the consistency requirements; n2 is the total number of simulation experiments, m2 is the number of simulation experiments that the human blood glucose model meets the consistency requirements, and the consistency evaluation result of the human blood glucose model is R rel2 =m2 / n2; through the formula Cd(PM)=0.5×R rel1+0.5×R rel2 The physical consistency evaluation result is calculated; the actual data is input into the insulin infusion rule model to stimulate the insulin infusion rule model to make a corresponding response, and the response of the insulin infusion rule model is compared with the actual response; if the response error is within the allowable range, the rule model is valid, and the rule consistency result Cd(RM)=1; otherwise, the rule model is incorrect or inaccurate Cd(RM)=0; finally, the overall digital twin model consistency evaluation result is obtained through the hierarchical analysis method, and the digital twin model is iteratively adjusted based on the evaluation results to meet the consistency requirements to obtain a high-fidelity digital twin model.

[0010] Furthermore, the insulin pump group fault monitoring unit estimates the thickness h of the blockage layer of the blockage fault in the flexible hose of the insulin pump based on the high-fidelity digital twin model and the real-time blood glucose monitoring data using the unscented Kalman filter algorithm. z , through the threshold judgment method, real-time monitoring is performed to determine whether the insulin pump group is in a healthy state.

[0011] The advantages and positive effects of the present invention are: 1. Adopt a hierarchical digital twin construction method; 2. Use digital twin technology to establish a high-fidelity blood sugar control model and quickly identify insulin pump group failures based on monitoring data; 3. Use the unscented Kalman filter algorithm to achieve higher real-time performance and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a schematic diagram of the steps of the method of the present invention; Figure 2 It is a schematic diagram of the hierarchical digital twin framework. DETAILED DESCRIPTION

[0013] The digital twin-based insulin pump group health status monitoring method of the present invention is described in detail below in conjunction with the accompanying drawings and embodiments.

[0014] like Figure 1 As shown in the figure, a digital twin-based insulin pump group health status monitoring system includes five units: a data set unit, a mechanism model set unit, a digital twin model integration unit, a digital twin model consistency evaluation unit, and an insulin pump group fault monitoring unit. The five units are connected in sequence to monitor the fault status of the insulin pump group.

[0015] 1. Data set unit: collects and stores measurable physiological data of individual diabetic patients, including blood glucose monitoring data, insulin input data and body weight. The blood glucose monitoring data is obtained using a continuous data blood glucose monitoring system, that is, a real-time blood glucose value of a diabetic patient is obtained every 15 minutes. The insulin input data in the data unit is determined based on the amount of exogenous insulin infusion, and data processing is performed in the unit to obtain the average basal insulin infusion within a 15-minute time interval, and it is assumed that the basal insulin infusion rate is a constant within 15 minutes between two measurements. The weight in the data unit is obtained through physical examination of the patient, and a corresponding data table is established in the MySQL database, and the database is operated using Navicat software.

[0016] 2. Mechanism model set unit: In this unit, the geometric model of the insulin pump group, the insulin fluid dynamics model, and the insulin infusion rule model are constructed based on the insulin pump group entity. The human blood glucose metabolism mechanism is studied and the human blood glucose metabolism model is constructed.

[0017] 3. Digital twin model integration unit: build Figure 2 A digital twin framework is proposed, and a digital twin model integration method is designed based on the model layer and data layer in the framework structure.

[0018] 3.1. Encapsulate the insulin pump group geometry model, insulin fluid dynamics model, insulin infusion rule model, and human blood glucose metabolism model into a unified FMI standard interface, and connect them through the interactive input and output of each sub-model to achieve the integration of models in different fields.

[0019] 3.2. Use the ODBC interface to access the physiological monitoring data of patients in the healthy state of the insulin pump group in the MySQL database. Based on the transmitted data and the integrated model obtained in 3.1, Bayesian estimation is used to determine the dynamic parameters of the model to obtain the digital twin model of the insulin pump group in a healthy state.

[0020] 4. Digital twin model consistency evaluation unit. This unit conducts geometric consistency evaluation on the geometric model of the insulin pump group, physical consistency evaluation on the insulin fluid dynamics model and the human blood glucose metabolism model, and rule consistency evaluation on the insulin infusion rules, and then obtains comprehensive consistency evaluation results, and adjusts the digital twin model based on the comprehensive results.

[0021] 4.1. Consistency assessment of geometric model of insulin pump set: Step 1: Determine the characteristic factor f to be measured s In this embodiment, the three elements are the geometric dimensions of the insulin pump set, the elastic modulus of the insulin pump set, and the viscosity of insulin. s is the set of elements to be tested, expressed as: GF s =(f s1 ,fs2 ,f s3 ).

[0022] In actual measurement, the characteristic element to be measured f s Can be further increased, expressed as: GF s =(f s1 ,f s2 ,f s3 ,…,f sN ).

[0023] Step 2: Obtain the characteristic elements x of the insulin pump entity through measurement. The characteristic element set x(f s )=(x1,x2,x3,…,x N ), the model geometry parameter set is y(f s )=(y1,y2,y3,…,y N ), the degree of geometric consistency of the model is expressed as Cd(GM).

[0024]

[0025] In the formula, i is the element number.

[0026] 4.2 Physical consistency assessment of insulin fluid dynamics model and human blood glucose metabolism model: Under the same conditions, the insulin dose output by the insulin pump group entity and the insulin fluid dynamics model are compared. n1 is the total number of simulation experiments, m1 is the number of simulation experiments in which the error between the entity and the virtual model output insulin dose is less than 10%, and the consistency evaluation result of the insulin pump group fluid dynamics model is R rel1 , then: R rel1 =m1 / n1.

[0027] The blood glucose values ​​of patients with type 1 diabetes wearing insulin pumps and the human blood glucose model established in this paper were tested under the same dose of insulin infusion. The error within the allowable range of blood glucose values ​​after 15 minutes was considered to meet the consistency requirements. n2 is the total number of simulation experiments, m2 is the number of simulation experiments that the human blood glucose model meets the consistency requirements, and the consistency evaluation result of the human blood glucose model is R rel2 , then: R rel2 =m2 / n2.

[0028] The physical model consistency evaluation result Cd(PM) is: Cd(PM)=0.5×R rel1 +0.5×R rel2 .

[0029] 4.3. Rule consistency evaluation of insulin infusion rule model: Step 1: Input actual data into the insulin infusion rule model to stimulate the insulin infusion rule model to make corresponding responses.

[0030] Step 2: Compare the response of the insulin infusion rule model with the actual response. If the response error is within the allowable range, the rule model is valid and the rule consistency result Cd(RM) = 1. Otherwise, the rule model is incorrect or inaccurate Cd(RM) = 0.

[0031] 4.4. Use the analytic hierarchy process to evaluate and adjust the digital twin model’s comprehensive consistency: Step 1: Construct a judgment matrix A for the three evaluation indicators of model geometry consistency Cd(GM), physical model consistency evaluation result Cd(PM), and rule consistency result Cd(RM) to describe the importance of different indicators. Judgment matrix element a ij The scaling method is as follows: Table 1: Ratio scale table

[0032] .

[0033] Step 2: λ max Represents the maximum eigenvalue of matrix A. w represents the eigenvector, which is also the weight vector.

[0034] Aw=λ max w.

[0035] Step 3: Determine the degree of logical collapse of the matrix. CI represents the consistency index of matrix A. RI represents the average random consistency index, and its standard value is shown in Table 2. Then the consistency ratio CR is obtained. If CR<0.1, it is believed that the judgment matrix can meet the requirements. Otherwise, the judgment matrix should be adjusted.

[0036] .

[0037] Table 2: Average random consistency index RI standard value Matrix order 1 2 3 4 5 6 7 8 9 10 RI 0 0 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49 Step 4: Based on the weight vector w=(w1,w2,w3) obtained in steps 2 and 3, assuming that the weights corresponding to the geometric, physical, and rule consistency evaluation results Cd(GM), Cd(PM), and Cd(RM) are w1, w2, and w3 respectively, calculate the comprehensive digital twin model consistency evaluation result Cd: , If the consistency result is lower than 0.95, repeat step 3.

[0038] 5. Insulin pump group fault monitoring unit: Based on the high-fidelity digital twin model and real-time blood glucose monitoring data, the fault parameters are dynamically estimated through the unscented Kalman filter algorithm. The fault threshold is set to determine the health status.

[0039] The method of the present invention is used to monitor the health status of a certain type of insulin pump group: 1. Collect 7-day physiological monitoring data of individual diabetic patients and store them in the corresponding database and table in MySQL software. Use Navicat software to visualize the database.

[0040] 2. Determine the mechanism model set. Taking the Medtronic 722 insulin pump set as an example, the geometric model of the insulin pump set is constructed in the modeling software. Based on the bond graph theory, according to the connection mode and energy conversion process of the insulin pump set, the power bond graph of the hydraulic system of the insulin pump set is obtained. According to the relevant rules of bond graph modeling, the fluid dynamics state equation of the insulin pump set can be obtained: Let the system state variable be X= , input variable but:

[0041] in:

[0042]

[0043] Among them, Q1 is the outlet flow rate of the drug reservoir interface; P1 is the outlet pressure of the drug reservoir interface; Q2 is the outlet flow rate of the flexible hose; P2 is the outlet pressure of the flexible hose; Q3 is the outlet flow rate of the rigid needle; P3 is the outlet pressure of the rigid needle; Q d To block the insulin flow at the fault, P d S is the insulin pressure at the blockage fault; f Input flow for insulin pump; S e The back pressure provided to the subcutaneous space of the human body.

[0044] R1, R f1 , R f2 , R2 is the static friction fluid resistance of the drug reservoir interface, flexible pipeline, and rigid needle;

[0045] Where: μ is the kinematic viscosity of insulin (m2•s-1); L is the length of the pipe (m); d is the inner diameter of the pipe (m).

[0046] R d1 , R d2 The dynamic friction resistance of the flexible pipe

[0047] Where: L is the length of the flexible pipe segment (m); d is the inner diameter of the flexible pipe (m).

[0048] , It is the equivalent fluid resistance of the blockage fault state.

[0049] ,

[0050] Where: C d is the flow coefficient of insulin through the blocked tube; r is the radius of the flexible tube (m); h z is the thickness of the inner wall of the blocked pipe (m); ρ is the density of insulin (kg•m-3); P d is the pressure at the blocked pipe; L is the length of the blocked pipe (m).

[0051] I1, I2, and I3 are the pipeline fluid senses of the drug reservoir interface, flexible pipeline, and rigid needle, respectively.

[0052]

[0053] Where: A1 is the pipe cross-sectional area of ​​the drug reservoir interface.

[0054] C1, C2, and C3 are the liquid capacities of the drug reservoir interface, flexible tubing, and rigid needle, respectively.

[0055]

[0056] Where: B is the liquid bulk modulus of insulin (N / m 2 ); E is the bulk elastic modulus of the drug reservoir interface pipeline (N / m 2 ); δ is the wall thickness of the drug reservoir connector.

[0057] According to the human blood glucose metabolism mechanism, this model takes into account the complete blood glucose regulation process, which is divided into three sub-processes: insulin is infused from subcutaneous tissue through an insulin pump and absorbed into plasma, the coupling between insulin and glucose in plasma and food, and the plasma glucose after coupling is transferred to the subcutaneous space and detected by a blood glucose meter. The complete glucose-insulin model is as follows:

[0058]

[0059]

[0060]

[0061]

[0062]

[0063] Where u(t) represents the exogenous insulin infusion rate at time t, the postprandial bolus insulin infusion is assumed to be a pulse input, x1(t) and x2(t) represent the insulin amounts in the first and second insulin absorption chambers, respectively, x(t) represents the insulin concentration in plasma, and t I represents the time (min) when the effective insulin concentration reaches the maximum value, M represents the metabolic clearance rate of effective insulin, and W represents the patient's weight. G(t) represents the plasma glucose concentration, S i Insulin sensitivity coefficient, S* i G is the insulin sensitivity factor, which represents a collection of factors related to insulin sensitivity. b Indicates the basal blood sugar value, K indicates the glucose self-regulation rate, KG b G represents the amount of endogenous glucose produced, and D(t) represents the sum of the interference of external factors such as carbohydrate intake on glucose concentration. I (t) represents the measurable glucose concentration in the subcutaneous interstitial space, τ represents the time for blood glucose to be transferred from plasma to the subcutaneous tissue space, and U(t) represents the lumped uncertainty.

[0064] 3. Digital twin model integration: integrating the insulin pump group geometry model, insulin fluid dynamics model, insulin infusion rule model, and human blood glucose metabolism model based on a unified interface.

[0065] The historical physiological monitoring data of patients wearing insulin pumps in the MySQL database were accessed through the ODBC interface, and the model parameters were determined using Bayesian estimation.

[0066] Based on the historical monitoring data of patients wearing insulin pumps in the data set, the model parameters are determined using Bayesian estimation. Glucose-insulin model parameters θ k =[x1,x2,G,G I ,t I ,S* i ,K,τ,U] joint posterior distribution P (θ k |D T ) can be obtained according to the following formula:

[0067] P ( θ k ) is the model parameter θ k The prior probability, L(θ k | D T ) is based on data D TThe calculated likelihood function, where D T is the data within the time window T, including continuous blood glucose monitoring data, insulin infusion data and eating data. Likelihood function L(θ k | D T ) is related to the model parameters and the observed data and can be expressed as:

[0068] m is the number of data sampling times in the time window T, D Ti Represents the i-th data in T, the model parameter set θ k The posterior probability can be expressed as (θ k | D T ), taking the mean of the posterior distribution of the parameter as the point estimate of the parameter, the parameter estimate of the model is It can be calculated by the joint posterior probability formula: .

[0069] 4. Consistency evaluation of digital twin models. According to the simulation, the geometric consistency evaluation result of the geometric model of the insulin pump group is Cd(GM)=1, the physical consistency evaluation result of the fluid dynamics model of the insulin pump group is Cd(PM)=1, and the consistency evaluation result of the infusion rule model of the insulin pump group is effective Cd(RM)=1.

[0070] The importance of geometry, physics, and rules is given in the following table: Table 3: Importance Assessment Table Geometric consistency Physical consistency Rule consistency Geometric consistency 1 1 / 5 1 Physical consistency 5 1 5 Rule consistency 1 1 / 5 1 According to the above table, the judgment matrix A is:

[0071] Calculate λ max =3, because CI=CR=0, the judgment matrix A meets the requirements, W=(1,5,1), the comprehensive evaluation results of the consistency of the digital twin model , determine whether the digital twin model meets the consistency requirements.

[0072] 5. Based on real-time blood glucose monitoring data and high-fidelity digital twin models, the thickness of the blockage layer h in the flexible hose of the insulin pump is calculated. z Set it as the fault parameter and use the unscented Kalman filter algorithm to reversely estimate h z The specific implementation steps of the unscented Kalman filter algorithm are as follows: Generate k Sigma points and corresponding weights , used to store the state transfer matrix and calculate the single-step prediction value of all sigma points :

[0073] Calculating state variables The one-step forecast value of is:

[0074] Calculate the covariance matrix ,in is the variance of the state equation.

[0075]

[0076] According to the single-step prediction results, the UT transformation is used again to generate a new Sigma point set. n is related to the state transfer matrix, and λ is the scale parameter:

[0077] Substitute the new Sigma point set into the observation equation H to obtain the output :

[0078] The system output value is obtained by weighted summation The mean and covariance of and , is the observation noise:

[0079] Calculate the Kalman gain matrix K(k): Update system status: Update the covariance matrix:

[0080] Based on the patient's real-time blood glucose data and the unscented Kalman filter algorithm, the fault parameters When it is greater than the safety value, it is determined that the flexible hose of the insulin pump group is blocked.

Claims

1. A method for monitoring the health status of an insulin pump set based on digital twins, characterized in that: A detection system including a data set unit, a mechanism model set unit, a digital twin model integration unit, a digital twin model consistency evaluation unit, and an insulin pump set fault monitoring unit is used; The data set unit collects physiological data and stores it in a database table; The digital twin model integration unit is based on a unified interface integration mechanism model; using the human blood glucose monitoring data of the insulin pump group in a healthy state in the database, Bayesian estimation is used to obtain the dynamic parameters in the integration model, and a digital twin model of the insulin pump group in a healthy state is constructed; The digital twin model consistency evaluation unit performs consistency evaluation on the mechanism model, obtains a comprehensive consistency evaluation result of the digital twin model based on the hierarchical analysis method, adjusts the digital twin model based on the consistency evaluation result to meet the consistency evaluation requirements, and obtains a high-fidelity digital twin model; The insulin pump group fault monitoring unit is based on a high-fidelity digital twin model, monitors blood glucose data in real time, and uses an unscented Kalman filter algorithm to dynamically estimate fault parameters.

2. A method for monitoring the health status of an insulin pump set based on data twins according to claim 1, characterized in that: The data set unit collects physiological data and stores it in a database table.

3. A method for monitoring the health status of an insulin pump set based on data twins according to claim 1, characterized in that: The mechanism model set unit includes an insulin pump group geometry model, an insulin fluid dynamics model, an insulin infusion rule model, and a human blood glucose metabolism model.

4. The method for monitoring the health status of an insulin pump set based on data twins according to claim 1, characterized in that: The digital twin model integration unit integrates the above-mentioned mechanism model based on the FMI interface, and uses the ODBC application programming interface to access the MySQL database in the above-mentioned data set unit; uses the human blood glucose monitoring data of the insulin pump group in a healthy state to use Bayesian estimation to obtain the dynamic parameters in the integrated model, and constructs a digital twin model of the insulin pump group in a healthy state.

5. The method for monitoring the health status of an insulin pump set based on data twins according to claim 1, characterized in that: The digital twin model consistency evaluation unit; determines and measures the characteristic elements to be measured of the geometric model of the insulin pump group, and also measures the characteristic elements of the insulin pump entity, and calculates the consistency degree of the two; sets the physical consistency evaluation index Cd(PM) for the insulin fluid dynamics model and the human blood glucose metabolism model; through simulation comparison experiments, compares the insulin dose output by the insulin pump group entity and the insulin fluid dynamics model under the same conditions, n1 is the total number of simulation experiments, m1 is the number of simulation experiments in which the error of the insulin dose output by the entity and the virtual model is less than 10%; calculates the consistency evaluation index R of the insulin pump group fluid dynamics model. re11 =m1 / n1; In addition, the blood glucose values ​​of type Ⅰ diabetic patients wearing insulin pumps and the human blood glucose model established in this paper were tested under the same dose of insulin infusion. The error within the allowable range was considered to meet the consistency requirements; n2 is the total number of simulation experiments, m2 is the number of simulation experiments that the human blood glucose model meets the consistency requirements, and the consistency evaluation result of the human blood glucose model is R rel2 =m2 / n2; through the formula Cd(PM)=0.5×R rel1 +0.5×R rel2 A physical consistency evaluation result is calculated; actual data is input into the insulin infusion rule model to stimulate the insulin infusion rule model to make a corresponding response, and the response of the insulin infusion rule model is compared with the actual response; If the response error is within the allowable range, the rule model is valid and the rule consistency result Cd(RM)=1; otherwise, the rule model is incorrect or inaccurate Cd(RM)=0; finally, the hierarchical analysis method is used to obtain the consistency evaluation result of the overall digital twin model, and based on the evaluation results, the digital twin model is iteratively adjusted to meet the consistency requirements to obtain a high-fidelity digital twin model.

6. The method for monitoring the health status of an insulin pump set based on data twins according to claim 1, characterized in that: The insulin pump group fault monitoring unit estimates the thickness h of the blockage layer of the blockage fault in the flexible hose of the insulin pump based on the high-fidelity digital twin model and the real-time blood glucose monitoring data using the unscented Kalman filter algorithm. z , through the threshold judgment method, real-time monitoring is performed to determine whether the insulin pump group is in a healthy state.

Images