Administration control method for tumor cell inhibition

By combining model smooth super-distortion control and conventional super-distortion control algorithms in tumor cell administration control, the vibration problem of traditional first-order sliding mode controllers is solved, achieving more stable and accurate tumor cell inhibition administration, reducing health Cell killing enhances anti-interference ability.

WO2024239360A1PCT designated stage expired Publication Date: 2024-11-28SHENZHEN TECH UNIV
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
PCT/CN2023/097262
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-05-22
Filing Date
2023-05-30
Publication Date
2024-11-28

AI Technical Summary

Technical Problem

In the existing tumor cell dosing control methods, the first-order sliding mode controller causes vibration due to the switching function, which causes unstable drug delivery system and makes it difficult to accurately control the tumor cell model to avoid the side effects of killing healthy cells.

Method used

Using the control law based on the hybrid of model smoothing ultra-distortion control algorithm (MBSSTC) and conventional ultra-distortion control algorithm (STC), a drug delivery control model suitable for tumor cell inhibition is designed to replace the traditional first-order sliding mode controller to reduce jitter Vibrate, improve system stability and accuracy.

Benefits of technology

Through the mixed algorithm, the impact of vibration on the drug delivery system is reduced, the accuracy and stability of tumor suppression administration is improved, the killing of healthy cells is reduced, the anti-interference ability is enhanced, and the growth pattern control of tumor volume is ensured .

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN2023097262_28112024_PF_FP_ABST
    Figure CN2023097262_28112024_PF_FP_ABST
Patent Text Reader

Abstract

Disclosed in the present invention is an administration control method suitable for a tumor cell mathematical model. The method comprises: constructing a tumor cell mathematical model; determining the use of a specific administration time and / or dosage as a control objective, so as to ensure that the tumor cell mathematical model can control a growth rule of a tumor in terms of volume; on the basis of the control objective, designing a smooth super-twisting controller by using a model-based smooth super-twisting control method, and designing a control law for smooth super-twisting controller parameters and an uncertain parameter of an administration system, so as to obtain a control law based on a smooth super-twisting control algorithm; and designing a control law based on a combination of a conventional super-twisting control algorithm and the model-based smooth super-twisting control algorithm. The present invention is not limited by an interference term condition in conventional administration control systems, ensures convergence in the presence of interference, and can effectively suppress chattering by means of smooth super-twisting control, thereby controlling and improving the performance of an administration system, improving the robustness of an administration system for uncertainties and external interference, and ensuring the stability of the system.
Need to check novelty before this filing date? Find Prior Art

Description

Drug administration control method for tumor cell inhibition Technical Field

[0001] The present invention relates to the technical field of drug administration control for cell inhibition, and in particular to a drug administration system control method based on super-distortion control and smooth super-distortion hybrid algorithm. Background Art

[0002] Cancer has become the second leading cause of death after cardiovascular disease. Existing cancer treatments include radiotherapy, surgery, chemotherapy, and immunotherapy. These methods are used to eliminate or reduce cancer cells in the human body.

[0003] Chemotherapy is a common method for treating cancer. To study the effects of radiotherapy, researchers construct tumor cell models to simulate the growth of both tumor and healthy cells. They then adjust certain parameters of the tumor cell models to determine the effectiveness of radiotherapy. While drugs kill cancer cells, they also kill a certain number of healthy cells, causing side effects. Therefore, precisely controlling the delivery of drugs to tumor cell models to avoid killing healthy cells and minimize side effects remains a challenge in this research field.

[0004] Prior art uses a first-order sliding mode controller to control the drug delivery process. While this conventional first-order sliding mode controller features discontinuous switching, theoretically offering strong anti-interference capabilities, the switching function inevitably leads to chattering, causing instability in the drug delivery system.

[0005] Therefore, how to provide a stable method for drug administration control in tumor cell models is a technical problem to be solved.

[0006] Summary of the Invention

[0007] In order to solve the problem of inevitable chattering of the first-order sliding mode controller caused by the switching function in the prior art, the present invention proposes a drug delivery control method for a tumor cell model based on model smoothing super-distortion control.

[0008] The method for controlling drug administration in a tumor cell model proposed by the present invention comprises:

[0009] 1) Construct a mathematical model of tumor cells;

[0010] 2) determining the control target to adopt a specific administration time and / or dosage to ensure that the tumor cell model can control the growth pattern of tumor volume;

[0011] 3) Based on the determined control objective, a conventional super-twisted control algorithm is used to design a first drug delivery controller model suitable for tumor cell inhibition, and a control law based on the conventional super-twisted control algorithm is designed;

[0012] 4) Based on the modeled based smooth super-twisting control algorithm (MBSSTC) and the conventional super-twisting control algorithm, a second drug delivery controller model suitable for tumor cell inhibition is designed, and a control law based on a hybrid of the conventional super-twisting control algorithm (STC algorithm) and the modeled smooth super-twisting control (MBSSTC algorithm) algorithm is designed.

[0013] Furthermore, the tumor cell mathematical model at least includes the logarithmic death model dx / dt=-rxln(x)-δxu(t), the Norton-Simon model dx / dt=-ln(x)[rx-δu(t)], and the Emax model

[0014] Where x is the tumor volume, r is the growth rate of the tumor volume, δ is the coefficient related to the drug usage, λ is the model constant, and u(t) is the drug concentration controlled at time t.

[0015] Furthermore, without considering external interference, the first drug administration control model suitable for tumor cell inhibition is:

[0016] Taking external interference into account, the second drug administration control model suitable for tumor cell inhibition is:

[0017] Where x(t) represents the tumor volume, u(t) represents the control input, d(t) represents the external interference, and f(x) represents the tumor cell model, where f(x) is expressed as -rxln(x) in the logarithmic death model, g(x) is expressed as δx in the logarithmic death model, f(x) is expressed as -rxln(x) in the Norton-Simon model, g(x) is expressed as δln(x) in the Norton-Simon model, f(x) is expressed as -rxln(x) in the Emax model, and g(x) is expressed as

[0018] Furthermore, the control law based on the conventional super-twist control algorithm is designed as follows:

[0019] Where k1 and k2 are adjustable gain parameters, and e represents the finite time convergence function.

[0020] Furthermore, the control law based on the hybrid of the conventional super-distortion control algorithm and the model smoothing super-distortion control algorithm is designed as follows:

[0021] Where k1, k2 represent the super-twisted controller parameters, k1>0, k2>0, ε represents the adjustable parameter, and e represents the finite time convergence function.

[0022] Furthermore, the tracking error of the tumor volume is defined as: e(t) = x(t) - x d (t); where x d is the expected value of tumor growth, with a maximum value of 1;

[0023] The first-order derivatives of the tracking error and drug delivery control model are: Substituting the control law based on the hybrid of conventional super-distortion control algorithm and model smoothing super-distortion control algorithm into the formula corresponding to the tracking error, the system dynamics formula can be obtained:

[0024] Where η1=e, e represents the finite time convergence function, which converges to the optimal solution within a finite time, and η2 is the calculus expression of η1 with the administration time.

[0025] Furthermore, the method further comprises the steps of:

[0026] 5) A quadratic-like Lyapunov function is used to perform stability analysis on the second drug delivery controller model when subjected to constant value disturbances and variable value disturbances to ensure the asymptotic stability of the system.

[0027] Furthermore, the stability analysis includes the following steps:

[0028] Based on the first definition and the first to fourth lemmas, we prove that the states η1 and η2 of the system will approach the following regions:

[0029] Where k1 and k2 are positive constants, and φ∈(0,1) is a constant that must be small enough;

[0030] The first definition is: given a vector x = [x1, x2, x3, ..., x n ]∈R n , and constant τ i >0,i=1,2,…,n.If p∈R, and definition Then the function R n →R, where τ=[τ1,τ2,…,τ n ] is the weight of the state. R is a positive definite matrix function;

[0031] The first lemma is: positive definite function R n →R and function Φ(x):Rn →RFor the same expansion weight, R has the same uniformity, and there is a constant and if is positive definite, then Where k≥0;

[0032] The second lemma is: Assume 0<a≤1, then y∈R,|[x] a -[y] a |≤2 1-a |xy| a ;

[0033] The third lemma is: Let m>0,n>0, then for y∈R, we can get:

[0034] Where h(x,y) is a positive function;

[0035] The fourth lemma is: Let b≥1, then for y∈R, then the following holds: |[x] b -[y] b |≤b(2 b-2 +2)(|xy| b +|xy||y| b-1 ; |[x] b -[y] b |≥2 1-b |xy| b ;

[0036] Need to prove

[0037] Region Q1 is constructed as:

[0038] because Available

[0039] From the first definition, we get and The 1-γ degree is homogeneous and is related to the expansion weight (1, 1+ε);

[0040] By using the first lemma, we can conclude the inequality:

[0041] Substituting the conclusion of the inequality into get:

[0042] Based on the facts get:

[0043] For any initial state (x1(0),x2(0))∈Q2, the region ensures that there exists a time instance T such that (x1,x2)∈Q2,

[0044] Q2 is defined as:

[0045] Considering the first definition, and H(x1,x2) is homogeneous to order 2+ε with respect to the expansion weight (1, 1+ε);

[0046] Using the first lemma, we can conclude that:

[0047] Restating the conclusion drawn using the first lemma, we obtain:

[0048] Based on the above formula, we get (x1,x2)∈Q2, that is

[0049] Alternatively, there exists a finite time instance T such that V(x1,x2)≤K,

[0050] When t≥T, we can get:

[0051] because a1=k1, from the above formula we can get:

[0052] On the other hand, when t ≥ T, we get:

[0053] Considering x1=η1, we can get:

[0054] Based on the above formula, it can be determined that in a finite time, the system states η1 and η2 will be driven and converge to the region Ω.

[0055] The present invention is directed to a tumor cell model in a drug delivery system, and the above technical solution proposed has the following beneficial effects.

[0056] 1. The drug delivery control scheme based on the hybrid algorithm of SCT and MBSTC replaces the discontinuous switching function used in the traditional SCT control algorithm with a continuous function, reducing the impact of chattering on the stability and accuracy of the tumor cell drug delivery system.

[0057] 2. By designing appropriate controller parameters, the error between the predicted tumor growth volume and the actual tumor growth volume can be reduced, thereby ensuring the effectiveness of the tumor suppression drug delivery system.

[0058] 3. The proposed control scheme for the tumor suppression drug delivery system based on the hybrid algorithm of SCT and MBSTC can be regarded as a new method to reclassify between SCT control and PI controller. By selecting appropriate parameters, the necessary anti-interference ability and tumor volume performance can be obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The present invention is described in detail below with reference to the embodiments and accompanying drawings, in which:

[0060] FIG1 is a main flow chart of the present invention;

[0061] Figure 2(a) shows the tumor volume response process of the STC controller.

[0062] Figure 2(b) shows the tumor volume response process of the MBSTC controller. DETAILED DESCRIPTION

[0063] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0064] Thus, a feature indicated in this specification is intended to illustrate one of the features of one embodiment of the present invention, rather than to imply that every embodiment of the present invention must have the described feature. In addition, it should be noted that this specification describes many features. Although certain features can be combined together to illustrate possible system designs, these features can also be used in other, not explicitly described, combinations. Thus, unless otherwise noted, the described combinations are not intended to be limiting.

[0065] The drug administration control method for the tumor cell model proposed in the present invention mainly includes the following steps.

[0066] Step 1: construct a mathematical model of tumor cells;

[0067] Step 2: Determine the control target by using a specific administration time and / or dosage to ensure that the tumor cell model can control the growth pattern of tumor volume;

[0068] Step 3: Based on the determined control objective, a conventional super-twisting control (STC) algorithm is used to design a first drug administration control model suitable for tumor cell inhibition, and a control law u(t) based on the conventional super-twisting control algorithm is designed;

[0069] Step 4: Design a second drug delivery controller model suitable for tumor cell inhibition based on the model smoothing super-distortion control algorithm and the conventional super-distortion control algorithm, and design a control law u(t) based on a hybrid of the STC algorithm and the MBSSTC algorithm.

[0070] In a preferred embodiment, step 5 may be further included, using a quadratic-like Lyapunov function to perform stability analysis on the above-mentioned tumor cell inhibition drug control system (including the above-mentioned drug control model and its control law) when it is subjected to constant value disturbances and variable value disturbances, to ensure the asymptotic stability of the system.

[0071] 1. Mathematical Model of Tumor Cells

[0072] In the above step 1, existing tumor cell models can be used. These tumor cell models can simulate the growth and death processes of tumor cells.

[0073] In a specific embodiment of the present invention, the provided tumor cell mathematical model at least includes a logarithmic death model, a Norton-Simon model, and an Emax model.

[0074] The logarithmic death model reduces cancer cells in a logarithmic manner.

[0075] When the Norton-Simon model was first used, the regression rate (tumor regression rate) was proportional to the immediate growth rate of the untreated tumor. Faster-growing tumors had higher regression rates than slower-growing tumors. This situation reinforces the idea that faster-growing cancer cells, which require less drug administration at an earlier stage of the tumor, can be killed.

[0076] The Emax model is the relationship between drug dosage and drug effect.

[0077] In this embodiment, the mathematical models of three types of tumor cells can be described by the following three formulas.

[0078] Logarithmic death model: dx / dt=-rxln(x)-δxu(t) (1)

[0079] Norton-Simon model: dx / dt=-ln(x)[rx-δu(t)] (2)

[0080] Emax model:

[0081] Where x is the tumor volume, r is the growth rate of the tumor volume, δ represents the coefficient related to the amount of drug used, and λ represents the model constant, which can also be called the function constant.

[0082] u(t) represents the drug concentration controlled at time t. When the concentration of the drug delivered at time t is 0, u(t) = 0 means that the cell does not use the drug, and when the concentration of the drug delivered at time t is greater than 0, u(t)>0 means the drug concentration used. In addition, when the initial time t = 0, given the tumor volume x(0) = x0, where 0 <x0<1。

[0083] In step 2 above, the growth trajectory of the tumor volume is designed to be x d (t). When a drug is injected into a tumor cell model, the tumor volume should decrease or reach a safe level. For example, a specific administration time, a specific dosage, or a specific dosage at a specific administration time. Therefore, it is necessary to ensure that the constructed cell model can control the growth trajectory of the tumor volume x d (t).

[0084] 2. Tumor Cell Inhibitory Drug Control Model

[0085] In the above step 3, based on the determined control target, the design of a tumor cell inhibition drug administration control model based on the STC algorithm can be specifically obtained by the following steps.

[0086] Without considering external interference, design the first drug delivery control model suitable for tumor cell models:

[0087] Taking external interference into consideration, a second drug administration control model suitable for tumor cell models was designed:

[0088] Where x(t) represents the tumor volume, u(t) represents the control input, d(t) represents the external interference, and f(x) represents the tumor cell model, which can be expressed by formulas (1)-(3). The specific expression is as follows.

[0089] f(x) is expressed as -rxln(x) in the logarithmic death model, and g(x) is expressed as δx in the logarithmic death model;

[0090] f(x) is expressed as −rxln(x) in the Norton-Simon model, and g(x) is expressed as δln(x) in the Norton-Simon model;

[0091] f(x) is expressed as -rxln(x) in the Emax model, and g(x) is expressed as

[0092] The tracking error of the designed tumor volume is: e(t) = x(t) - x d (t); (6)

[0093] where x dis the expected value of tumor growth, with a maximum value of 1;

[0094] The first-order derivatives of formulas (6) and (5) are:

[0095] Design the control law u based on the conventional STC algorithm:

[0096] Where k1 and k2 are adjustable gain parameters, and e represents the finite time convergence function.

[0097] In the control law formula (8), if k1 is chosen to be large enough and k2>D, the control law can guarantee finite time convergence. The sign function, sign(e), extracts the sign of a function. In mathematics and computer science, it determines the sign (positive or negative) of a number. When e > 0, sign(e) = 1; when e = 0, sign(e) = 0; and when e < 0, sign(e) = -1.

[0098] In step 4 above, in order to avoid unnecessary chattering and improve the stability of the cell delivery system, the discontinuous term k2 in the STC algorithm is removed, and a control law u based on the hybrid algorithm of SCT and MBSTC is designed:

[0099] Where k1, k2 represent the super-twisted controller parameters, k1>0, k2>0, ε represents the adjustable parameter, and e represents the finite time convergence function.

[0100] Further, substituting formula (9) into formula (7), we can obtain:

[0101] Where η1=e, e represents the finite-time convergence function, which converges to the optimal solution within a finite time. η2 is the calculus expression of η1 with the administration time, that is, the calculus expression of the system state with the administration time.

[0102] Figure 2(a) shows the tumor volume response process using the STC controller. Figure 2(b) shows the tumor volume response process using the SCT-based and MBSTC-based controllers.

[0103] 3. Stability Analysis of Tumor Cell Inhibitory Drug Delivery Control System

[0104] In the above step 5, a quadratic-like Lyapunov function is used to perform stability analysis on the above tumor cell inhibition drug control system when it is subjected to constant value disturbances and variable value disturbances to ensure the asymptotic stability of the system.

[0105] First definition: Given a vector x = [x1, x2, x3, ..., x n ]∈R n , and constant τ i >0,i=1,2,…,n.If p∈R, and definition Then the function r n →R: called p-order homogeneous, where τ=[τ1,τ2,…,τ n ] is the weight of the state. R is a matrix function, which is positive definite. n →R, the following are the instructions.

[0106] The symbol “→” means that the quantity on the left side of the symbol is replaced by the quantity on the right side of the symbol.

[0107] First Lemma: Positive Definite Function and function Φ(x):R n →RFor the same expansion weight, R has the same uniformity. There is a constant and if It is positive, there is Where k≥0.

[0108] Second Lemma: Assume 0<a≤1, then y∈R, |[x] n -[y] a |≤2 1-a |xy| a

[0109] The third lemma: Let m>0,n>0, then for y∈R, we can get:

[0110] Where h(x,y) is a positive function.

[0111] Fourth Lemma: Let b≥1, then for y∈R, the following holds |[x] b -[y] b |≤b(2 b-2 +2)(|xy| b +|xy||y| b-1 (a) |[x] b -[y] b |≥2 1-b |xy| b (b)

[0112] The states η1 and η2 of the system will be close to the following regions:

[0113] where k1 and k2 are positive constants, and φ∈(0,1) is a constant that must be sufficiently small.

[0114] Obviously, the system state η1 is affected by For example, the tumor volume error e is bounded. By properly adjusting the parameters k1 and ε, the tumor volume error e can be reduced to an arbitrarily small value, thereby ensuring the accuracy of the cancer chemotherapy process.

[0115] Definition: x1=η1 and The system dynamics formula (10) can be restated as:

[0116] Where a1=k1, a2=k2 / k1.

[0117] In order to simplify formula (12), we define:

[0118] Then, the reconstruction of the system dynamics (12):

[0119] Dependence on Lemma 2 and The following conclusions were drawn from these facts.

[0120] Applying Lemma 4(a) we obtain:

[0121] in

[0122] In order to ensure the stability of the system (14), the Lyapunov function is chosen as follows: V(x1, x2) = W1(x1, x2) + W2(x2) (17)

[0123] in At the same time, along the system (14), V(x1, x2) is differentiated with respect to time as follows:

[0124] use and choose a2≥2

[0125] By using Lemma 3 and (15), we can obtain:

[0126] where γ1 is a positive constant.

[0127] Substituting equation (20) into (19), we get

[0128] Similarly, the second term can be rewritten

[0129] The following can be simply calculated:

[0130] According to Lemma 3, we can get:

[0131] Where γ2 is a positive constant.

[0132] Similarly, we can get:

[0133] Where γ3 is a positive constant.

[0134] In addition, With the help of Lemma 4(b), it is clear that

[0135] Substituting (24), (25) and (26) into (22) yields:

[0136] Through (21) and (27), (18) can be written as:

[0137] Additionally, the constant a1 can be chosen as follows:

[0138] Substituting (29) into (28) we obtain:

[0139] in

[0140] Based on assumption 1, we can get:

[0141] The region Ω is then shown to be reachable in finite time by using the next two steps.

[0142] Step 1: In this step, we will prove And the region Q1 is constructed as:

[0143] because Available

[0144] From Definition 1, we have and The 1-γ degree is homogeneous and is related to the expansion weight (1, 1+ε).

[0145] By using Lemma 1, we can draw the following conclusions:

[0146] Substituting (32) into (31) we obtain:

[0147] Based on the facts Formula (33) can be rewritten as

[0148] However, when There will be no guarantee, that is, that the state can escape from region Q1.

[0149] Therefore, a larger region Q2 should be determined. For any initial state (x1(0), x2(0))∈Q2, this region ensures that there exists a time instance T such that (x1, x2)∈Q2, Q2 is defined as:

[0150] Step 2: Based on the above analysis, the premise is that it can be proved that Q1 is a subset of Q2. Derived

[0151] Considering Definition 1, and H(x1,x2) is homogeneous to 2+ε with respect to the expansion weight (1, 1+ε). Using Lemma 1, we can conclude that:

[0152] The above formula can be restated as:

[0153] Therefore, from the above formula, we can conclude that (x1,x2)∈Q2, that is

[0154] In other words, there exists a finite time instance T such that V(x1,x2)≤K,

[0155] When t≥T, we can get:

[0156] because a1=k1, from the above formula we can get:

[0157] On the other hand, when t ≥ T, we get:

[0158] Considering x1=η1, it is easy to derive:

[0159] Therefore, it can be simply inferred from (37) and (38) that the system states η1 and η2 will be driven and converge to the region Ω in finite time.

[0160] The fractional power parameter ε can be obtained from Select Except for 0.

[0161] when When ε = 0, the control scheme for the cancer chemotherapy model based on the smoothing super-warp algorithm is reduced to a standard first-order sliding mode control scheme. When ε = 0, the control scheme for the cancer chemotherapy model based on the smoothing super-warp algorithm is converted to PI control. In view of this, the control scheme for the tumor cell model based on the smoothing super-warp algorithm designed in this invention can be regarded as a non-smooth controller between traditional first-order sliding mode control and PI controller.

[0162] Furthermore, due to the influence of sampling time on the digital application of the controller, the anti-disturbance performance of the first-order sliding mode control may not be better than that of the non-smooth controller in practical situations. Due to the careful selection of the parameter ε, the proposed cancer chemotherapy controller based on the smoothed super-warp algorithm not only has smaller drug usage oscillations than the standard first-order sliding mode control, but also shows stronger anti-disturbance performance.

[0163] Researchers studied the application of a hypertorsion sliding mode control algorithm to a nonlinear model of a brain tumor drug delivery system to stabilize the closed-loop system, suppress uncertainty and disturbances, reduce chattering, and optimize drug dosage. Simulation results showed that adjusting the concentration and rate of drug delivery effectively improves drug delivery effectiveness in the presence of system uncertainty and disturbances.

[0164] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A drug administration control method suitable for tumor cell inhibition, characterized in that: The following steps are involved: 1) Construct a mathematical model of tumor cells; 2) determining the control target to adopt a specific administration time and / or administration amount to ensure that the tumor cell model can control the growth pattern of tumor volume; 3) Based on the determined control target, a conventional super-twisted control algorithm is used to design a first drug delivery controller model suitable for tumor cell inhibition, and a control law based on the conventional super-twisted control algorithm is designed; 4) Based on the model smoothing super-twist control algorithm and the conventional super-twist control algorithm, a second drug delivery controller model suitable for tumor cell inhibition is designed, and a control law based on a mixture of the conventional super-twist control algorithm and the model smoothing super-twist control algorithm is designed.

2. The method for controlling drug administration for tumor cell inhibition according to claim 1, characterized in that: The tumor cell mathematical model at least includes the logarithmic death model dx / dt=-rxln(x)-δxu(t), the Norton-Simon model dx / dt=-ln(x)[rx-δu(t)], the Emax model Where x is the tumor volume, r is the growth rate of the tumor volume, δ is the coefficient related to the amount of drug used, and λ is The model constant, u(t) represents the controlled drug concentration at time t.

3. The method for controlling drug administration for tumor cell inhibition according to claim 1, characterized in that: Without considering external interference, the first drug administration control model suitable for tumor cell inhibition is: Taking external interference into account, the second drug administration control model suitable for tumor cell inhibition is: Where x(t) represents the tumor volume, u(t) represents the control input, d(t) represents the external interference, and f(x) represents the tumor cell model, where f(x) is expressed as -rxln(x) in the logarithmic death model, g(x) is expressed as δx in the logarithmic death model, f(x) is expressed as -rxln(x) in the Norton-Simon model, g(x) is expressed as δln(x) in the Norton-Simon model, f(x) is expressed as -rxln(x) in the Emax model, and g(x) is expressed as 4. The method for controlling drug administration for tumor cell inhibition according to claim 1, characterized in that: The control law based on the conventional super-twist control algorithm is designed as follows: Where k1 and k2 are adjustable gain parameters and e represents the finite time convergence function.

5. The method for controlling drug administration for tumor cell inhibition according to claim 1, characterized in that: The control law based on the hybrid of the conventional super-distortion control algorithm and the model smoothing super-distortion control algorithm is designed as follows: Where k1, k2 represent the super-twist controller parameters, k1>0, k2>0, ε represents the adjustable parameter, and e represents the finite time convergence function.

6. The method for controlling drug administration for tumor cell inhibition according to claim 5, characterized in that: The tracking error of the tumor volume is defined as: e(t) = x(t) - x d (t); where x d is the expected value of tumor growth, with a maximum value of 1; The first-order derivatives of the tracking error and drug delivery control model are: Substituting the control law based on the mixture of conventional super-distortion control algorithm and model smoothing super-distortion control algorithm into the formula corresponding to the tracking error, the system dynamics formula can be obtained: where η1=e, e represents the finite time convergence function, which converges to the optimal solution within a finite time. η2 is the calculus expression of η1 with the administration time.

7. The method for controlling drug administration for tumor cell inhibition according to any one of claims 1 to 6, characterized in that: Also includes the steps: 5) A quadratic-like Lyapunov function is used to perform stability analysis on the second drug delivery controller model when it is subjected to constant value disturbances and variable value disturbances to ensure the asymptotic stability of the system.

8. The method for controlling drug administration for tumor cell inhibition according to claim 7, characterized in that: The stability analysis consists of the following steps: Based on the first definition and the first to fourth lemmas, we prove that the states η1 and η2 of the system will approach the following regions: Where k1 and k2 are positive constants, and φ∈(0,1) is a constant that must be small enough; The first definition is: given a vector x = [x1, x2, x3, ..., x n ]∈R n , and the constant τ i >0,i=1,2,…,n.If p∈R, and definition Then the function where τ=[τ1,τ2,…,τ n ] is the weight of the state. R is a positive definite matrix function; The first lemma is: positive definite function Sum function Φ(x):R n →RFor the same expansion weight, R has the same uniformity, and there is a constant and if is positive, then Where k ≥ 0; The second lemma is: Assume that \(0 < a\leqslant1\), then y∈R,|[x] a -[and] a |≤2 1-a |xy| a ; The third lemma is: Let m>0,n>0, then for y∈R, we can get: Where h(x,y) is a positive function; The fourth lemma is: Let b ≥ 1, then for y∈R, then the following holds: |[x] b -[and] b |≤b(2 b-2 +2)(|xy| b +|xy||y| b-1 ; |[x] b -[and] b |≥2 1-b |xy| b ; prove And the region Q1 is constructed as: because Available From the first definition, we get and The 1-γ degree is homogeneous and is related to the expansion weight (1, 1+ε); By using the first lemma, we can conclude the following inequality: Substituting the conclusion of the inequality into get: Based on the facts get: For any initial state (x1(0),x2(0))∈Q2, the region ensures that there exists a time instance T such that (x1,x2)∈Q2, Q2 is defined as: Considering the first definition, and H(x1,x2) are homogeneous 2+ε with respect to the expansion weight (1, 1+ε); using the first lemma, we can conclude that: Restating the conclusion drawn using the first lemma, we obtain: Therefore, (x1,x2)∈Q2, Alternatively, there exists a finite time instance T such that V(x1,x2)≤K, When t≥T, we can get: because a1=k1, so we get: On the other hand, when t ≥ T, we get: Considering x1=η1, we can get: It can be determined that in a finite time the system states η1 and η2 will be driven and converge to the region Ω.

Citation Information

Patent Citations

  • Isogeometric simulation system for studying inhibition of chemotherapeutic drugs on tumor growth

    CN113468827A

  • Anti-cancer drug response prediction method and device based on tumor cell line self-supervised learning

    CN114842983A

  • Cancer chemotherapy intelligent monitoring and optimized drug delivery system based on multi-stage Gaussian pseudo-spectrum method

    CN115312153A

  • Computational model, method, and system for kinetically-tailoring multi-drug chemotherapy for individuals

    US20030144798A1

  • Method for estimating or predicting the anti-tumor activity of a compound and for estimating or predicting the tumor growth in mammals

    US20050143927A1