CN-116942952-B - Optimization method for insulin infusion rate in artificial pancreas

CN116942952BCN 116942952 BCN116942952 BCN 116942952BCN-116942952-B

Abstract

The invention relates to the technical field of artificial pancreas, in particular to an optimization method for insulin infusion rate in an artificial pancreas, which comprises the following steps of S1, establishing a blood sugar control mathematical model, calculating a general expression of the insulin infusion rate containing an equivalent part and an equivalent linear equation of an insulin concentration error, S2, improving a conventional linear insulin infusion rate method in a closed loop system of the artificial pancreas, and providing a nonlinear method capable of replacing the conventional linear insulin infusion rate, namely a non-smooth infusion rate, and S3, optimizing an insulin infusion rate improvement gain parameter in the nonlinear method in the step S2, so that the insulin concentration reaches a desired value faster, the required maximum infusion rate is smaller, and the consumed energy is smaller. The optimization method of the invention can obviously improve the control speed of the insulin concentration, and has smaller required maximum infusion rate and less energy consumption.

Inventors

JIANG BOYAN
XING YUEXIAN
XIANG SHOUKUI
LI CHENGLIN
DU WENHAO

Assignees

常州工学院
常州市第一人民医院

Dates

Publication Date: 20260512
Application Date: 20230720

Claims (3)

1. A system for performing an optimization method of insulin infusion rate in an artificial pancreas, the system performing an optimization method comprising the steps of: S1, establishing a blood sugar control mathematical model, and calculating an equivalent linear equation of an insulin infusion rate general expression containing an equivalent part and an insulin concentration error; the mathematical model of blood glucose control in S1 is as follows: wherein G (t) represents the blood glucose concentration in mg/dL in blood, Represents the basal value of blood sugar, the unit is mg/dL, The unit of the hypoglycemic effect of insulin is min -1 , The change in blood glucose concentration due to feeding is expressed in mg.dL -1 min -1 , Represents the insulin concentration in mU/L, Represents the basic value of insulin, the unit is mU/L, n represents the decay rate coefficient of insulin, the unit is min -1 , The insulin infusion rate, also referred to as the controller to be designed, is expressed in mU/min, Represents the distribution volume of insulin, in units of L, The rate coefficient representing the rate of glucose from the plasma space into the liver or into the periphery, in min -1 , The coefficient of the blood glucose reducing rate of insulin is expressed in min -1 , The rate coefficient representing the action of plasma insulin on insulin for reducing blood glucose is expressed as L.mU -1 ·min -2 ; the general expression of the insulin infusion rate in S1, including the equivalent part, is as follows: Wherein u 0 represents the equivalent insulin infusion rate and I d represents the calculated virtual insulin concentration in the inner annulus of the artificial pancreas using the back-stepping method; The equivalent linear equation for the insulin concentration error in S1 is expressed as: Wherein I e represents an insulin concentration error, the expression of which is I e =I-I d ; S2, improving a conventional linear insulin infusion rate method in an artificial pancreas closed loop system, and providing a nonlinear method capable of replacing the conventional linear insulin infusion rate, namely a nonlinear infusion rate; The conventional linear insulin infusion rate used in the closed loop system of the artificial pancreas in S2 is: Wherein the method comprises the steps of To control the gain parameters, the non-smooth infusion rate is improved with the explicit structure: Wherein the method comprises the steps of In order to improve the control gain parameter, Is a parameter of the power exponent, Functional representation ; And S3, optimizing the insulin infusion rate improvement gain parameter in the nonlinear method in the step S2 to enable the insulin concentration to reach a desired value faster, wherein the required maximum infusion rate is smaller and the consumed energy is smaller.
2. The system for performing the method for optimizing the rate of insulin infusion in an artificial pancreas according to claim 1, wherein the insulin infusion rate in step S3 improves the control gain parameter The method for optimizing is as follows: Improved control gain parameters The following constraints are satisfied: Wherein, the Representing variables Is used for the initial value of (a), The coefficient of precision is represented by a coefficient of precision, Indicating the expected error accuracy of the insulin concentration system, Is that the settling time of the non-smooth infusion rate is not greater than the minimum gain parameter of a conventional linear infusion rate, The physical meaning of (a) is that the maximum value of the non-smooth infusion rate is not greater than the maximum gain parameter of a conventional linear infusion rate.
3. The system for performing the method for optimizing the rate of insulin infusion in an artificial pancreas according to claim 1, wherein the insulin infusion rate in step S3 improves the control gain parameter The method for optimizing is as follows: Improved control gain parameters The following constraints are satisfied: Wherein, the Representing variables Is used for the initial value of (a), The coefficient of precision is represented by a coefficient of precision, Is that the settling time of the non-smooth infusion rate is not greater than the minimum gain parameter of a conventional linear infusion rate, The physical meaning of (a) is that a non-smooth infusion rate consumes no more energy than the maximum gain parameter of a conventional linear infusion rate.

Description

Optimization method for insulin infusion rate in artificial pancreas Technical Field The invention relates to the technical field of control of artificial pancreas, in particular to an optimization method for insulin infusion rate in the artificial pancreas. Background Diabetes causes vascular and neurogenic lesions, and various complications appear, which seriously threatens the health and life of patients, and the total number of diabetics in China reaches more than 1 hundred million people and is in continuous rapid growth. Common diabetes is mainly classified into type 1 diabetes and type 2 diabetes, wherein the type 1 diabetes patients have early onset, large blood sugar fluctuation and more serious symptoms, and the quality of life of the patients is greatly influenced. In order to prevent various complications caused by hyperglycemia from harming the human body, type 1 diabetics need to rely absolutely on insulin for glycemic control. Insulin therapy is generally classified into intravenous insulin injection and subcutaneous insulin injection using an insulin pen or an insulin pump. According to the experience of doctors or patients, although the injection of insulin once or several times a day can help the blood sugar control of patients, the optimal insulin dosage of each patient is greatly changed on different days and at different times of the same day, so that the aim of stabilizing the blood sugar is difficult to achieve by using the conventional insulin infusion mode for patients with type 1 diabetes with poor liver function. Thanks to the development of sensor technology and internet of things technology, sustainable glucose monitors are currently used in blood glucose detection for diabetes. The insulin subcutaneous pump and the sustainable glucose monitor are combined to form the wearable artificial pancreas, so that insulin can be automatically secreted according to the blood sugar level of a human body, and the insulin subcutaneous pump can replace beta cells in the pancreas. Insulin pump auto-injection of too much insulin may result in hypoglycemia and injected too little insulin may result in hyperglycemia and subsequent induction of diabetic ketoacidosis. In conclusion, the aim of safely and stably controlling blood sugar can be achieved only by injecting proper insulin according to the individual blood sugar condition. Therefore, a key technology of the artificial pancreas is to accurately control the blood sugar of the patient by using an effective closed-loop control algorithm. Proportional Integral Derivative (PID) algorithms were the earliest and most widely used control algorithms in artificial pancreas, after which many intelligent control methods based on PID algorithms were used in tandem in artificial pancreas. For example, a Chinese patent application with publication number of CN104958077A discloses an intelligent control closed-loop artificial pancreas system, wherein a fuzzy self-adaptive proportional-calculus control algorithm adopts a proportional-calculus algorithm to simulate the physiological transmission process of human beta cells secreting insulin as a basic model, and adopts a fuzzy logic algorithm to perform rolling optimization on each parameter in the basic model established by the proportional-calculus algorithm to perform intelligent control on insulin infusion equipment. For example, the method disclosed in the chinese patent application CN106860955A and based on fuzzy adaptive proportional-calculus control of insulin pump closed-loop infusion takes the blood glucose concentration of diabetics as the control object, the real-time blood glucose measured value as the input of the PID controller, the insulin pump injection quantity as the output of the PID controller, and uses fuzzy logic reasoning to simulate the decision process of human to continuously roll and optimize each parameter in the PID prediction model according to the blood glucose data monitored in real time, so that the control method of the PID controller can accurately calculate the injection time and injection quantity of insulin, provide blood glucose control close to the normal range for patients, and realize the optimal closed-loop control of the fuzzy adaptive PID control algorithm on the infusion of insulin pump. The control methods proposed in the above patents all use PID control algorithm to a certain extent, and from the aspect of system stability analysis, the algorithm can ensure that the system tends to be stable at an exponential convergence rate, but if the convergence rate of insulin concentration is to be further improved, the infusion rate is often only improved. However, too high an infusion rate of insulin may result in an infusion rate higher than the absorption rate of insulin by subcutaneous tissue, the insulin temporarily accumulating subcutaneously and forming a hard mass, which cannot be absorbed normally, resulting in controlled saturation and controlled residual ph