CN-122005205-A - Volume quantitative control method of retinal intravenous injection system

CN122005205ACN 122005205 ACN122005205 ACN 122005205ACN-122005205-A

Abstract

The invention discloses a quantitative control method for the volume of a retinal intravenous injection system, which belongs to the technical field of ophthalmic robots and comprises the steps of firstly designing a smooth injection volume curve as the expected volume of the quantitative control method, designing an evolutionary observer aiming at disturbance characteristics of different stages of the injection system to estimate unknown disturbance and compensate the unknown disturbance into a controller, adopting an integral sliding mode surface and an equivalent control law designed by the evolutionary observer, and combining a switching control law designed by adopting a self-adaptive gain, wherein the obtained total control law can realize the reliable tracking of the expected volume. The invention can ensure the high-precision closed-loop control of the medicine volume in the retinal intravenous injection operation and has robust tracking performance.

Inventors

SUN MINGZHU
HU BO
ZHAO XIN
LIU RONGXIN
Li Zhouchurui

Assignees

南开大学

Dates

Publication Date: 20260512
Application Date: 20260408

Claims (9)

1. A quantitative control method for the volume of a retinal intravenous injection system, which is characterized by comprising the following steps: constructing an injection volume third-order dynamics model aiming at a retina intravenous injection system, and determining an input variable and an output variable of the injection system; designing a smooth injection volume curve as a desired output variable of a volumetric quantitative control method; designing an evolution observer aiming at unknown disturbance of an injection system, wherein the evolution observer comprises an initial observer and a steady-state observer, and the two parts are switched through preset time, and the steady-state observer particularly comprises an estimation mechanism of disturbance derivative; The control law is designed by adopting a fixed time integral sliding mode surface, comprises an equivalent control law designed based on the sliding mode surface and an evolution observer, and adopts a switching control law of a self-adaptive gain design, and the input variable in a volume third-order dynamics model constructed by the retinal intravenous injection system is controlled by combining the equivalent control law and the switching control law to obtain a total control law.
2. The method according to claim 1, wherein the construction of the injection volume third-order kinetic model is specifically as follows: The injection system adopts a motor-driven piston P 1 to push the silicone oil, the silicone oil pushes a syringe piston P 2 , and finally the medicine in the syringe is injected into the retinal vein cavity; The first order dynamics of the process of pushing the silicone oil flow by the piston P 1 are described as follows: ; Wherein the method comprises the steps of And Respectively represent the flow rate of the silicone oil and the first derivative thereof, Indicating the speed of the piston P 1 , Representing an unknown disturbance of the process, 、 And Positive parameters respectively; Speed of syringe piston P 2 Flow rate with silicone oil The relation of (2) is: ; Wherein the method comprises the steps of For the area of the syringe piston, Representing an unknown dynamics of the process; The pharmacokinetics driven by the syringe piston P 2 can be described as: ; Wherein the method comprises the steps of And Respectively representing the drug flow and the first derivative thereof, Indicating the speed of the piston P 2 , Representing an unknown disturbance of the process, 、 And Positive parameters respectively; based on the volume of drug injected The first order dynamics of which are described as follows: ; Wherein the method comprises the steps of Is the first derivative of the drug volume; a volumetric third-order kinetic model of the retinal intravenous injection system was obtained according to the four formulas above: ; Wherein the method comprises the steps of , , , As the third derivative of the volume of the drug, As the second derivative of the volume of the drug, As the first derivative of the volume of the drug, In order to be a lumped disturbance, Wherein Is that Is the first derivative of (a); selecting state variables , As control input, the third-order dynamics model is rewritten as: ; the volume of the medicament is controlled by designing a closed loop controller to control the speed at which the drive motor pushes the piston P 1 .
3. The method according to claim 2, characterized in that the injection volume curve as desired output variable is in particular as follows: Is provided with The time period firstly adopts a fixed flow injection strategy and is provided with The time period adopts a quadratic curve to make the track smoothly transition until the final set volume quantity, wherein, For a constant flow state stop time, For injection stop time; according to the principle of continuity of flow and volume, the following conditions are obtained: ; Wherein the method comprises the steps of Is that The desired volume of the drug at the moment in time, Is that The desired volume of the drug at the moment in time, Is the volume curve Is used as a first derivative of (a), Is that The first derivative of the expected drug volume at the moment is obtained by solving the above equation: ; the desired injection volume is: 。
4. a method according to claim 3, characterized in that the evolutionary observer is designed to estimate the lumped uncertainty of the system, in particular as follows: At a preset stage The following initial observers are used during the period to estimate the disturbance: ; Wherein the method comprises the steps of Is that An estimate of the output system uncertainty of the initial observer over a period of time, , , , Is a positive parameter, the parameter is a positive parameter, And For a parameter that is greater than 1, And For parameters greater than 0 and less than 1, Is an auxiliary variable; At a preset stage Estimating disturbance by adopting a steady-state observer in a time period, and setting unknown disturbance dynamics as Wherein For an unknown nonlinear mapping function, a finite dimensional linear system is utilized to approximate unknown disturbance dynamics based on a Koopman operator and an extended dynamic modal decomposition method, and a lifting function vector based on an unknown disturbance autoregressive vector and a state variable is firstly constructed: ; Wherein the method comprises the steps of In order to be an auto-regressive order, A state variable vector at the time t; Unknown disturbance dynamics The linear model in lifting function space is described as a discrete time form as follows: ; Wherein the method comprises the steps of To approximate a Koopman matrix, the vectors in the matrix Solving by linear bayesian regression: ; Wherein the method comprises the steps of , , Is that Is the first of (2) The number of elements to be added to the composition, In order to be able to measure the data quantity, Is a positive parameter, the parameter is a positive parameter, Is a unit matrix; Converting the discrete system model of unknown disturbance to a continuous system model to match the steady state observer design: ; Wherein the method comprises the steps of , In order to sample the time of the sample, Is a matrix of units which is a matrix of units, ; At a preset level The steady state observer over the period of time is designed to: ; Wherein the method comprises the steps of Is that An estimate of the output system uncertainty of the steady state observer over a period of time, Is a matrix Is used to determine the first row vector of (c), Is vector quantity Posterior standard deviation of (2); The update law is as follows: ; Wherein the method comprises the steps of Is a positive constant.
5. The method of claim 4, wherein the integrated sliding surface employed is expressed as: ; Wherein the method comprises the steps of , , , , In order to track the error in the tracking, Is a state vector Is the first of (2) The number of elements to be added to the composition, In order for the desired state variable to be a function of, , , And In order to set the threshold value in advance, As a function of the sign of the symbol, Is an integral variable.
6. The method of claim 5, wherein the equivalent control law is designed The method comprises the following steps: ; Wherein the method comprises the steps of Is the third derivative of the desired volume.
7. The method of claim 6, wherein the control law is switched The design is as follows: ; Wherein the method comprises the steps of In order to adaptively switch the gain, Is a positive parameter, the parameter is a positive parameter, Is a constant value, and is a function of the constant, , 。
8. The method of claim 7, wherein the adaptive switching gain The adaptive update law of (2) is: ; Wherein the method comprises the steps of Is a positive constant.
9. The method of claim 4, wherein data within a fixed time window is employed A solution to the approximate Koopman matrix is performed.

Description

Volume quantitative control method of retinal intravenous injection system Technical Field The invention belongs to the technical field of ophthalmic robots, and particularly relates to a volume quantitative control method of a retinal intravenous injection system. Background The retinal intravenous injection operation can achieve accurate administration of blood vessels, and particularly, thrombolytic drugs are injected into occluded retinal veins to restore local blood circulation and reduce the risk of blindness. However, in the existing retinal intravenous injection operation, a viscous fluid control unit of a vitrectomy machine is often used to drive drug injection, and the process uses a fixed pressure, so that the risk of damaging local blood vessels occurs due to the instantaneous high-flow impact of the initial process. Considering the side effects of the drug, excessive amounts of drug may also contribute to the bleeding risk of local tissues, while insufficient amounts of drug may impair the intended therapeutic effect. Disclosure of Invention Aiming at the technical problems existing in the prior art, the invention provides a quantitative control method for the volume of a retinal intravenous injection system, which ensures accurate tracking of the expected volume in the retinal intravenous injection process, ensures drug administration effect and reduces operation risks. The invention realizes the aim through the following technical scheme: A method for quantitatively controlling the volume of a retinal intravenous injection system, comprising the following steps: constructing an injection volume third-order dynamics model aiming at a retina intravenous injection system, and determining an input variable and an output variable of the injection system; designing a smooth injection volume curve as a desired output variable of a volumetric quantitative control method; An evolutionary observer is designed for unknown disturbance of an injection system, and comprises an initial observer and a steady-state observer, and the observer is switched through preset time. Wherein the steady state observer comprises in particular an estimation mechanism of the disturbance derivative. The observer not only realizes the safety of the initial stage of injection, but also ensures the accuracy of the steady-state stage of injection. The control law is designed by adopting a fixed time integral sliding mode surface, comprises an equivalent control law designed based on the sliding mode surface and an evolution observer, and adopts a switching control law of a self-adaptive gain design, and the input variable in a volume third-order dynamics model constructed by the retinal intravenous injection system is controlled by combining the equivalent control law and the switching control law to obtain a total control law. The injection volume third-order kinetic model is constructed as follows: The injection system adopts a motor-driven piston P 1 to push the silicone oil, the silicone oil pushes a syringe piston P 2, and finally the medicine in the syringe is injected into the retinal vein cavity; The first order dynamics of the process of pushing the silicone oil flow by the piston P 1 are described as follows: ; Wherein the method comprises the steps of AndRespectively represent the flow rate of the silicone oil and the first derivative thereof,Indicating the speed of the piston P 1,Representing an unknown disturbance of the process,、AndPositive parameters respectively; Speed of syringe piston P 2Flow rate with silicone oilThe relation of (2) is: ; Wherein the method comprises the steps of For the area of the syringe piston P 2,Representing an unknown dynamics of the process; The pharmacokinetics driven by the syringe piston P 2 can be described as: ; Wherein the method comprises the steps of AndRespectively representing the drug flow and the first derivative thereof,Indicating the speed of the piston P 2,Representing an unknown disturbance of the process,、AndPositive parameters respectively; based on the volume of drug injected The first order dynamics of which are described as follows: ; Wherein the method comprises the steps of Is the first derivative of the drug volume; a volumetric third-order kinetic model of the retinal intravenous injection system was obtained according to the four formulas above: ; Wherein the method comprises the steps of ,,,As the third derivative of the volume of the drug,As the second derivative of the volume of the drug,As the first derivative of the volume of the drug,In order to be a lumped disturbance,WhereinIs thatIs the first derivative of (a); selecting state variables ,As control input, the third-order dynamics model is rewritten as: ; the volume of the medicament is controlled by designing a closed loop controller to control the speed at which the drive motor pushes the piston P 1. Further, the injection volume curve as the desired output variable is specifically as follows: Is pr