A practical finite-time command filter backstepping control method for MPCVD reactors

Abstract

A practical finite-time command filter backstepping control method for an MPCVD reactor relates to the technical field of MPCVD reactor process control. Aiming at the problems of uncertainty and slow convergence in the MPCVD reactor process control system in the prior art, the present application proposes a practical finite-time command filter to estimate unknown nonlinearity and proposes an equivalent augmented matrix of the MPCVD reactor to solve the problem of unknown control direction function; realizes the design of feedback controller u i , so that the system output can track the given target signal within a small error range, thereby solving the problems of uncertainty and slow convergence in the MPCVD reactor process control system in the existing technology.

Description

Technical Field

[0001] The present invention relates to the technical field of MPCVD reactor process control, in particular to a practical finite-time command filtering backstepping control method for an MPCVD reactor. Background Art

[0002] Microwave plasma chemical vapor deposition (MPCVD) technology is commonly used to produce materials such as single-crystal diamond and diamond-like carbon films. Currently, the most widely used control method for MPCVD reactor systems is still based on proportional-integral-differential (PID). MPCVD reactor systems are highly nonlinear and strongly coupled, and PID control still has certain limitations when dealing with such systems. Current MPCVD reactor process control systems also suffer from uncertainty and slow convergence. Summary of the Invention

[0003] The purpose of the present invention is to propose a practical finite-time command filtering backstepping control method for an MPCVD reactor in view of the problems of uncertainty and slow convergence in the process control system of the MPCVD reactor in the prior art.

[0004] The technical solution adopted by the present invention to solve the above technical problems is:

[0005] A practical finite-time command filtering backstepping control method for an MPCVD reactor, wherein the specific process of the method is as follows:

[0006] Step 1: Based on the state variable x of the MPCVD reactor i , output signal y i and control signal u i , establish a state-space model for process control of MPCVD reactor;

[0007] Step 2: Based on the state space model of the MPCVD reactor process control, define the extended state variable w i =u i , a state space model of a two-dimensional nonlinear system for process control of an MPCVD reactor with extended state variables is established, and based on the state space model of a two-dimensional nonlinear system for process control of an MPCVD reactor with extended state variables, the error variable z is defined. 1,i 、z 2,i ;

[0008] Step 3: Use the error variable z from step 2 1,i 、z 2,i Design Lyapunov function V;

[0009] Step 4: Using the Lyapunov function V in step 3, take the first-order derivative with respect to time to obtain

[0010] Step 5: Introduce practical finite time command filter estimation, according to the first-order derivative of the Lyapunov function in step 4 Using the backstepping method, the virtual control function α is designed i and controller input u i , and then complete the backstepping control.

[0011] Furthermore, the specific steps of step one are:

[0012] First, a non-parametric uncertain state space model with unknown control direction for the process control of the MPCVD reactor is established as follows:

[0013]

[0014] Among them, T, P and represent the temperature of the MPCVD reactor substrate, the reaction chamber pressure, the CH4 gas flow rate, and the H2 gas flow rate, respectively. T 、m、C Mu ,T0,k c and R h denote the temperature attenuation coefficient, the mass of the heater resistor, the specific heat capacity of the reaction substrate, the initial temperature of the reaction substrate, the cooling coefficient, and the heater resistor, respectively. and k a 、C a , ρ a , ΔT a Represent the heat exchange coefficient, specific heat, density and temperature difference of CH4, H2 and air respectively, R g 、V r 、k v 、T r 、k t Represent the gas constant, the volume of the reactor, the voltage coefficient of the vacuum pump, the reaction temperature and the temperature coefficient, R0, C0, A, K respectively P Respectively represent the equivalent resistance, equivalent capacitance, amplification factor and proportional coefficient of the valve drive circuit,

[0015] Unknown equation is the unknown nonlinearity of the system, u h 、u c 、u v 、u e Represent the heater voltage, water pump voltage, vacuum pump voltage and electronic valve voltage respectively, where Indicates the electronic valve voltage of methane, Indicates the electronic valve voltage of hydrogen, represents the first derivative of T, represents the first derivative of P, represents the first derivative of Q;

[0016] Take x1=T, x2=P, The state space model of the process control of the MPCVD reactor is established as:

[0017]

[0018] y i =x i

[0019] Where β1 = 1, β2 = -R g (T r -k t T)k v / V,β3=β4=AK P / (R0C0),

[0020]

[0021] u2=u v ,

[0022] Among them, x i represents the state variables of the nonparametric uncertain state space model with unknown control direction for process control of MPCVD reactor, β1,...,β4 represent the unknown control direction function of the system, θ2,...,θ4 represent the unknown nonlinear function of the system, Represents x i The first derivative of u i Represents the controller, y i Indicates system output.

[0023] Furthermore, when the state variables x1,...x4 are bounded, θ1,...,θ4, is bounded, θ1,...,θ4, The first-order derivative of each with respect to time is bounded;

[0024] The θ1,...,θ4, Represent θ1(x i (t),t),...,θ4(x i (t),t),

[0025] Furthermore, the system outputs y i is bounded, the system output y i The first derivative with respect to time is bounded.

[0026] Furthermore, the state space model of the two-dimensional nonlinear system for process control of the MPCVD reactor with extended state variables is expressed as:

[0027]

[0028] y i =x i ,

[0029] Among them, γ i represents the equivalent control direction function.

[0030] Furthermore, the error variable z 2,i =w i -α i .

[0031] Furthermore, the Lyapunov function V is expressed as:

[0032]

[0033] Furthermore, in step 4 Expressed as:

[0034]

[0035] in, represents the target signal y d The first derivative of Express the virtual control function α i The first derivative of Instead of the unknown function, represents w i The first derivative of .

[0036] Furthermore, the specific steps of step five are:

[0037] Step 51: Use the virtual control function to rewrite the first-order derivative of the Lyapunov function in step 4 Expressed as:

[0038]

[0039] Unknown Define a practical finite-time command filter estimate z 1,i :

[0040]

[0041] Among them, z 1,i , represents the input and output of the practical finite-time filter, η 1,irepresents the state of the practical finite-time filter, S() represents the hyperbolic tangent function, e 1,i represents η 1,i and z 1,i The difference,

[0042] e 1,i =η 1,i -z 1,i ,S(e 1,i )=tanh(b 1,i e 1,i ), b 1,i >0, μ 1,i , τ 1,i represents the filter parameters, represents z 1,i The first derivative of and Deviation ξ 1,i for:

[0043]

[0044] That is, the first derivative of the Lyapunov function Expressed as:

[0045]

[0046] in, Indicates u i The first derivative of

[0047] Step 52: According to the first derivative of the Lyapunov function in step 5 Using backstepping and adaptive law, design virtual control function

[0048] in, κ 1,i , σ 1,i represents the filter parameters;

[0049] Step 53: Design a practical finite-time filter to estimate α i :

[0050]

[0051] in, η 2,i denote the output and state of the practical finite-time filter, e 2,i represents η 2,i and α i The difference,

[0052] e 2,i =η 2,i -α i ,S(e2,i )=tanh(b 2,i e 2,i )b 2,i >0, μ 2,i , τ 2,i represents the filter parameters, and The error ξ 2,i for:

[0053]

[0054] Step 54: Design control signal u i :

[0055]

[0056] in, κ 2,i , σ 2,i represents the filter parameters, τ represents the integral variable, k 2,i >0 indicates design parameters.

[0057] Furthermore, the virtual control function α i Expressed as:

[0058]

[0059] The beneficial effects of the present invention are:

[0060] This application proposes a practical finite-time command filter to estimate unknown nonlinearity and proposes an equivalent augmented matrix for the MPCVD reactor to solve the problem of unknown control direction function; realizes the design of feedback controller u i , so that the system output can track the given target signal within a small error range, thereby solving the problems of uncertainty and slow convergence in the MPCVD reactor process control system in the existing technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 is the temperature response curve of the system;

[0062] Figure 2 is the gas flow response curve;

[0063] Figure 3 It is the CH4 gas flow response curve;

[0064] Figure 4 It is the H2 gas flow response curve;

[0065] Figure 5 is the input voltage response of the PFCFB controller;

[0066] Figure 6 is the input voltage response of the TCFB controller. DETAILED DESCRIPTION

[0067] It should be noted that, unless there is any conflict, the various embodiments disclosed in this application can be combined with each other.

[0068] Specific implementation method 1: refer to Figure 1 Specifically describing this embodiment, a practical finite-time command filtering backstepping control method for an MPCVD reactor described in this embodiment includes:

[0069] Step 1: Based on the state variable x of the MPCVD reactor i , output signal y i and control signal u i , establish a state-space model for process control of MPCVD reactor;

[0070] Step 2: Based on the state space model of the MPCVD reactor process control, define the extended state variable w i =u i , a state space model of a two-dimensional nonlinear system for process control of an MPCVD reactor with extended state variables is established, and based on the state space model of a two-dimensional nonlinear system for process control of an MPCVD reactor with extended state variables, the error variable z is defined. 1,i 、z 2,i ;

[0071] Step 3: Use the error variable z from step 2 1,i 、z 2,i Design Lyapunov function V;

[0072] Step 4: Using the Lyapunov function V in step 3, take the first-order derivative with respect to time to obtain

[0073] Step 5: Introduce practical finite time command filter estimation, according to the first-order derivative of the Lyapunov function in step 4 Using the backstepping method, the virtual control function α is designed i and controller input u i , and then complete the backstepping control.

[0074] This application proposes an innovative and practical finite-time command filter backstepping control method. The MPCVD reactor system is abstracted as a nonlinear system with unknown control direction functions and nonlinearities. To handle the unknown nonlinearities of the reactor system, a practical finite-time command filter is designed to construct a nonlinear estimation. Furthermore, to address the design challenges posed by the unknown control direction function of the system, an equivalent enhanced system for the MPCVD reactor system is proposed. Furthermore, based on the practical finite-time Lyapunov stability criterion, the tracking error of the reactor system is guaranteed to be practical finite-time stable.

[0075] In step 1, the state variable x is controlled according to the MPCVD reactor process. i , output signal y i and control signal u i , a state space model for MPCVD reactor process control is established. The specific process is as follows:

[0076] The state space model of the system with unknown control direction for MPCVD reactor process control is established as:

[0077]

[0078] Among them, T[℃], P[Pa] and They are the temperature of the MPCVD reactor substrate, the reaction chamber pressure, the CH4 gas flow rate and the H2 gas flow rate.

[0079] Coefficient D T ,m[kg],C Mu [J / (kg·℃)],T0[℃],k c and R h [Ω] represents the temperature attenuation coefficient, the mass of the heater resistor, the specific heat capacity of the reaction substrate, the initial temperature of the reaction substrate, the cooling coefficient, and the heater resistor, respectively.

[0080] and k a ,C a ,ρ a ,ΔT a They represent the heat exchange coefficient, specific heat, density and temperature difference of CH4, H2 and air respectively.

[0081] Coefficient R g ,V r ,k v , T r [℃],k t They represent the gas constant, the volume of the reactor, the voltage coefficient of the vacuum pump, the reaction temperature and the temperature coefficient respectively.

[0082] Coefficients R0, C0, A, K PThey represent the equivalent resistance, equivalent capacitance, amplification factor and proportional coefficient of the valve drive circuit respectively.

[0083] Unknown equation is the unknown nonlinearity of the system, u h ,u c ,u v , They represent heater voltage, water pump voltage, vacuum pump voltage and electronic valve voltage respectively.

[0084] Take x1=T,x2=P, The state space model of MPCVD reactor process control is established as follows:

[0085]

[0086] Where β1 = 1, β2 = -R g (T r -k t T)k v / V,β3=β4=AK P / (R0C0),

[0087]

[0088]

[0089] u2=u v ,

[0090] where β2,...,β4,θ1,...,θ4, are all unknown functions. Controller u i The design goal is to make the system output tracking error practically finite time stable.

[0091] where x i represents the state variables of system (1), β1,...,β4 are the unknown control direction functions of the system, θ2,...,θ4 are the unknown nonlinear functions of the system, is x i The first derivative of u i is the controller, y i Output of the system.

[0092] β2,...,β4,θ1,...,θ4, With uncertainty; β2,...,β4,θ1,...,θ4, is nonlinear; β2,...,β4,θ1,...,θ4, Only with x i About not involving ui .

[0093] “Having unknown control direction functions” means that β2, ..., β4 are unknown, and such a system is more difficult to design.

[0094] The control objective is to design a feedback controller u i , so that the system outputs y i Able to track the given system target signal y within a small error range di .

[0095] When the state variables x1,...x4 are bounded, θ1,...,θ4, is bounded, θ1,...,θ4, The first-order derivative of each with respect to time is bounded;

[0096] The θ1,...,θ4, Represent θ1(x i (t),t),...,θ4(x i (t),t),

[0097] Target output y i Is bounded, the target output y i The first derivative with respect to time is bounded.

[0098] In step 2, define the extended state variable w i =u i , a state space model for MPCVD reactor process control with extended state variables, defining the error variable z 1,i , z 2,i ; The specific process is:

[0099] 1) Define the extended state variable w i =u i ,

[0100] 2) Establish a system state space model with extended state variables:

[0101]

[0102] where γ i Represents the equivalent control direction function, taking Replace unknown function.

[0103] 3) Define the error variable z 2,i =w i -α i

[0104] where αi Represents a virtual control function.

[0105] In step 3, we use the error variable z from step 2 1,i , z 2,i Design the Lyapunov function V; the specific process is:

[0106] Use the error variables z1 and z2 in step 2 to design the Lyapunov function V

[0107]

[0108] In step 4, use the Lyapunov function V in step 3 to find the first-order derivative with respect to time

[0109]

[0110] Use the Lyapunov function V in step 3 to find the first-order derivative with respect to time:

[0111] in is the target signal y d The first derivative of Represents the virtual control function α i The first derivative of .

[0112] In step 5, a practical finite-time command filter is introduced, and the first-order derivative of the Lyapunov function in step 4 is obtained. Using the backstepping method, the virtual control function α is designed i and controller input u i ; The specific process is:

[0113] 1) Use the virtual control function to rewrite the first-order derivative of the Lyapunov function in step 4

[0114]

[0115] Among them, γ i is the equivalent control direction function,

[0116] 2) Define the practical finite-time command filter estimate z 1,i :

[0117]

[0118] where z 1,i , are the input and output of the practical finite-time filter, η 1,i is the state of the practical finite-time filter. In addition, e 1,i =η 1,i -z1,i ,S(e 1,i )=tanh(b 1,i e 1,i ), b 1,i >0,μ 1,i , τ 1,i is a positive parameter. The deviation of is defined as follows:

[0119]

[0120] That is, the first derivative of the Lyapunov function

[0121]

[0122] 3) According to the first-order derivative of the Lyapunov function in step 5 Design a virtual control function using backstepping and practical finite time filter estimation

[0123] in κ 1,i ,σ 1,i is a positive parameter.

[0124] That is, the first derivative of the Lyapunov function

[0125]

[0126] 4) Design a practical finite-time filter to estimate α i :

[0127]

[0128] where α i , η 2,i are the input, output and state of the practical finite-time filter respectively. In addition, e 2,i =η 2,i -α i ,S(e 2,i )=tanh(b 2,i e 2,i )b 2,i >0,μ 2,i ,τ 2,i is a positive parameter, The error is:

[0129]

[0130] 5) Design control signal u:

[0131]

[0132] Among them, κ 2,i , σ 2,i are positive filter parameters, τ is the integral variable, and k 2,i > 0 is the design parameter.

[0133] That is, the first derivative of the Lyapunov function

[0134]

[0135] Equation (11) is the tracking controller based on the backstepping control of the practical finite-time filter.

[0136] The desired virtual control function

[0137] Other steps and parameters are the same as those in any one of the specific embodiments one to nine.

[0138] Next, it will be proved that the tracking controller of the process control method for the MPCVD reactor based on the improved adaptive backstepping control can make the system tracking error converge to a small neighborhood near the origin. The proof process is as follows:

[0139] Lemma 1. If there exist constants n1, n2, M > 0, 0 < c < ∞, 0 < p < 1, and a continuous function V(x), where V(x(0)) ≤ M satisfies the following inequality:

[0140]

[0141] Among them, obeys the practical finite-time stability. Therefore, the system solution residual set can be obtained as:

[0142]

[0143] where ρ ∈ (0, 1), and T is the set time as follows:

[0144]

[0145] According to Lemma 1, a practical finite-time command filter is given below to approximate the time derivative

[0146]

[0147] Among them μ and τ are positive filter parameters. η is the filter state, is the filter output, S(x) = tanh(bx), b > 0.

[0148] Lemma 2. If the input signal and is bounded, there exists a constant T f >0, satisfy:

[0149]

[0150] where ξ represents the practical finite-time filter error.

[0151] For i=1,2,3,4, we can get

[0152]

[0153] When x∈Ω x It is not difficult to conclude is bounded, so we have:

[0154]

[0155] where ξ 1,i ,ξ 2,i represents the error using a finite-time filter, is a positive parameter. So we can get

[0156]

[0157] where |ε 1,i |,|ε 2,i |≤1, ζ 1,i ,ζ 2,i >0,

[0158] By Lemma 1:

[0159]

[0160] Where j = 1, 2, ρ∈(0, 1), a sets the time T f satisfy

[0161]

[0162] in

[0163] V2(z j,i (0))≤M,i=1,...,4,j=1,2 (25).

[0164] Obviously 1,i is uniformly eventually bounded. When we apply practical adaptive backstepping control, as follows:

[0165]

[0166] It can be further concluded that:

[0167]

[0168] Take c / b≤V2(z j,i (0))≤M, Tracking error {V2≤λ}, setting time:

[0169]

[0170] Take parameters that satisfy:

[0171]

[0172] We can get:

[0173] T f ≤T e (30)

[0174] So far, it shows that the system tracking error converges to a smaller neighborhood near the origin, and the proof is complete.

[0175] The following examples are used to verify the beneficial effects of the present invention:

[0176] System parameters are:D T =0.5,k c =20,m=1,T0=25,C Mu =242.8,R h =10, k a =0.02,ΔT a =1200,C a =1005,ρ a =1.169,R g =8.314,T r =1500,V r =1,k v =0.001,k t =0.01, R0=50, C0=20, A=0.05,K P =0.1, The initial value of the state variable is x(0) = [25, 1200, 0, 0] T .

[0177] PFCFB controller parameters: γ1=1,γ2=-12,γ3=5e-6,γ4=5e-6,κ 1,1 =4,σ 1,1 =10,μ1,1 =100,t 1,1 =100,k 2,1 =2,s 2,1 =5.m 2,1 =100,t 2,1 =100,k 1,2 =40,s 1,2 =200.m 1,2 =400,t 1,2 =200,k 2,2 =20,s 2,2 =100.m 2,2 =400,t 2,2 =200,k 1,3 =4,s 1,3 =10.m 1,3 =200,t 1,3 =200,k 2,3 =2,s 2,3 =5.m 2,3 =200,t 2,3 =200,k 1,4 =4,s 1,4 =10.m 1,4 =200,t 1,4 =200,

[0178] k 2,4 =2,s 2,4 =5.m 2,4 =200,t 2,4 =200, The initial values ​​are:x 2,1 (0)=x 2,2 (0)=x 2,3 (0)=x 2,4 (0)=0,u1(0)=u2(0)=u3(0)=u4(0)=0,n 1,1 =-1075,h 1,2 =200,h 1,3 =-10 -7 ,or 1,4 =-0.5*10 -5 ,or 2,1 (0)=4310,h 2,2 (0)=683.3333,n 2,3 (0)=0.28,η 2,4 (0)=14,

[0179] The TCFB controller parameters are: γ1=1,γ2=-12,γ3=5e-6,γ4=5e-6,λ 1,1 =4,m 1,1 =100,λ 2,1 =2,m 2,1 =100,λ 1,2 =40,m 1,2 =200,λ 2,2 =20,m 2,2 =200,λ 1,3 =4,m 1,3 =200,λ 2,3 =2,m 2,3 =200,λ 1,4 =4,m 1,4 =200,λ 2,4 =2,m 2,4 =200.The initial values ​​are:x 2,1 (0) = x 2,2 (0) = x 2,3 (0) = x 2,4 (0)=0,u1(0)=u2(0)=u3(0)=u4(0)=0,s 1,1 =-1075,s 1,2 =200,s 1,3 =-10 -7 ,s 1,4 =-0.5*10 -5 ,s 2,1 (0)=4300,s 2,2 (0)=666.67,s 2,3 (0)=0.08,s 2,4 (0)=4,

[0180] The practical finite-time command filtering backstepping controller of the technical solution of the present application can achieve an exponentially faster tracking error convergence speed than the traditional command filtering backstepping.

[0181] It should be noted that the specific embodiments are merely explanations and illustrations of the technical solutions of the present invention and cannot be used to limit the scope of protection. Any minor changes made based on the claims and description of the present invention shall still fall within the scope of protection of the present invention.

Claims

1. A practical finite-time command filtering backstepping control method for an MPCVD reactor, characterized by: The specific process of the method is: Step 1: Based on the state variable x of the MPCVD reactor i , output signal y i and control signal u i , establish a state-space model for process control of MPCVD reactor; Step 2: Based on the state space model of the MPCVD reactor process control, define the extended state variable w i =u i , a state space model of a two-dimensional nonlinear system for process control of an MPCVD reactor with extended state variables is established, and based on the state space model of a two-dimensional nonlinear system for process control of an MPCVD reactor with extended state variables, the error variable z is defined. 1,i 、z 2,i ; Step 3: Use the error variable z from step 2 1,i 、z 2,i Design Lyapunov function V; Step 4: Using the Lyapunov function V in step 3, take the first-order derivative with respect to time to obtain Step 5: Introduce practical finite time command filter estimation, according to the first-order derivative of the Lyapunov function in step 4 Using the backstepping method, the virtual control function α is designed i and controller input u i , and then complete the backstepping control; The specific steps of step one are: First, a non-parametric uncertain state space model with unknown control direction for the process control of the MPCVD reactor is established as follows: Among them, T, P and represent the temperature of the MPCVD reactor substrate, the reaction chamber pressure, the CH4 gas flow rate, and the H2 gas flow rate, respectively. T 、m、C Mu ,T0,k c and R h denote the temperature attenuation coefficient, the mass of the heater resistor, the specific heat capacity of the reaction substrate, the initial temperature of the reaction substrate, the cooling coefficient, and the heater resistor, respectively. and k a 、C a , ρ a , ΔT a Represent the heat exchange coefficient, specific heat, density and temperature difference of CH4, H2 and air respectively, R g 、V r 、k v 、T r 、k t Represent the gas constant, the volume of the reactor, the voltage coefficient of the vacuum pump, the reaction temperature and the temperature coefficient, R0, C0, A, K respectively P Respectively represent the equivalent resistance, equivalent capacitance, amplification factor and proportional coefficient of the valve drive circuit, Unknown equation is the unknown nonlinearity of the system, u h 、u c 、u v 、u e Represent the heater voltage, water pump voltage, vacuum pump voltage and electronic valve voltage respectively, where Indicates the electronic valve voltage of methane, Indicates the electronic valve voltage of hydrogen, represents the first derivative of T, represents the first derivative of P, represents the first derivative of Q; Take x1=T, x2=P, The state space model of the process control of the MPCVD reactor is established as: and i =x i where, β1 = 1, β2 = -R g (T r -k t T)k v / V, β3 = β4 = AK P / (R0C0), u2=u v ,u3=u e1 ,u4=u e2 Among them, x i represents the state variables of the nonparametric uncertain state space model with unknown control direction for process control of MPCVD reactor, β1,...,β4 represent the unknown control direction function of the system, θ2,...,θ4 represent the unknown nonlinear function of the system, Represents x i The first derivative of u i Represents the controller, y i Indicates system output; When the state variables x1,...x4 are bounded, θ1,...,θ4, is bounded, θ1,...,θ4, The first-order derivative of each with respect to time is bounded; The θ1,...,θ4, Represent θ1(x i (t),t),...,θ4(x i (t),t), The system output y i is bounded, the system output y i The first-order derivative with respect to time is bounded; The state space model of the two-dimensional nonlinear system for process control of the MPCVD reactor with extended state variables is expressed as: and i =x i , Among them, γ i represents the equivalent control direction function; The error variable z 2,i =w i -α i ; The specific steps of step five are: Step 51: Use the virtual control function to rewrite the first-order derivative of the Lyapunov function in step 4 Expressed as: Unknown Define a practical finite-time command filter estimate z 1,i : Among them, z 1,i , represents the input and output of the practical finite-time filter, η 1,i represents the state of the practical finite-time filter, S() represents the hyperbolic tangent function, e 1,i represents η 1,i and z 1,i The difference, e 1,i =η 1,i -z 1,i ,S(e 1,i )=tanh(b 1,i e 1,i ), b 1,i >0, μ 1,i , τ 1,i represents the filter parameters, represents z 1,i The first derivative of and Deviation ξ 1,i for: That is, the first derivative of the Lyapunov function Expressed as: in, Indicates u i The first derivative of Step 52: According to the first derivative of the Lyapunov function in step 5 Using backstepping and adaptive law, design virtual control function in, κ 1,i , σ 1,i represents the filter parameters; Step 53: Design a practical finite-time filter to estimate α i : in, η 2,i denote the output and state of the practical finite-time filter, e 2,i represents η 2,i and α i The difference, e 2,i =η 2,i -α i ,S(e 2,i )=tanh(b 2,i e 2,i )b 2,i >0, μ 2,i , τ 2,i represents the filter parameters, and The error ξ 2,i for: Step 54: Design control signal u i : in, κ 2,i , σ 2,i represents the filter parameters, τ represents the integral variable, k 2,i >0 indicates design parameters.

2. A practical finite time command filtering backstepping control method for an MPCVD reactor according to claim 1, characterized in that The Lyapunov function V is expressed as:

3. A practical finite time command filter backstepping control method for an MPCVD reactor according to claim 2, characterized in that In step 4 Expressed as: in, represents the target signal y d The first derivative of Express the virtual control function α i The first derivative of Instead of the unknown function, Indicates w i The first derivative of .

4. A practical finite time command filter backstepping control method for an MPCVD reactor according to claim 3, characterized in that The virtual control function α i Expressed as:

Images