MPCVD system stability monitoring method and system for discrete fractional order model

Abstract

The application provides a kind of MPCVD system stable monitoring method of discrete fractional order model, the method comprises: real-time acquisition of the operating state data of MPCVD equipment, constructs state vector, state vector centralization and normalization processing;Based on the state data, establish Nabla discrete fractional order hybrid time delay model;According to the system matrix identified, construct linear matrix inequality group, and call solver by industrial control computer to carry out online feasibility determination;When the linear matrix inequality group exists feasible solution, it is judged that the current system satisfies stability condition;When the linear matrix inequality group does not exist feasible solution, generate control adjustment and output to microwave power supply power controller, automatic pressure regulating valve and / or mass flow controller.The application develops around "multi-source state acquisition-state data preprocessing-Nabla discrete fractional order hybrid time delay modeling-LMI stability online determination-power, pressure and flow coordinated regulation" this link.

Description

Technical Field

[0001] This invention specifically relates to a method and system for stability monitoring of a discrete fractional-order MPCVD system. Background Technology

[0002] Microwave plasma chemical vapor deposition (MPCVD) is an important technology for preparing high-quality single-crystal diamond, polycrystalline diamond thin films, and cutting-edge semiconductor materials such as silicon carbide. During the long growth cycle of MPCVD, the morphology, volume, and density of the plasma spheres in the chamber have a direct impact on deposition quality, process consistency, and equipment operational safety.

[0003] MPCVD chambers are highly nonlinear and complex dynamic systems. During actual operation, changes in adjustable parameters such as microwave power, chamber pressure, and gas flow rate affect the plasma state and substrate temperature through thermal inertia, gas transport hysteresis, and chamber response hysteresis, resulting in significant mixed time-delay characteristics. Simultaneously, the chemical reaction kinetics and thermal diffusion processes within the plasma exhibit significant historical cumulative effects, causing the system to display memory characteristics. Traditional integer-order PID control methods or conventional state feedback methods have limited descriptive capabilities for this type of system, making it difficult to balance model accuracy, decision rigor, and online intervention effectiveness.

[0004] Existing MPCVD control schemes mostly focus on temperature regulation, power compensation, or closed-loop control of a single variable, lacking a unified mathematical framework for online assessment and control intervention of operational stability. This paper proposes establishing a discrete fractional-order model for MPCVD systems that can simultaneously express memory and time-delay effects, and using linear matrix inequalities for stability assessment. The assessment results are then directly converted into actuator adjustment quantities, demonstrating clear engineering application value. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by providing a method for monitoring the stability of MPCVD systems using a discrete fractional-order model. This method enables online determination of the operational stability of MPCVD equipment and automatically drives the actuators to adjust when the system is at risk of instability.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is: to provide a method for stability monitoring of an MPCVD system with a discrete fractional-order model, comprising the following steps:

[0007] S1: Real-time acquisition of the operating status data of the MPCVD equipment and the corresponding single-crystal diamond substrate infrared temperature, construction of state vector, state vector centering and normalization processing;

[0008] S2: Based on the data collected in step S1 and introducing an initial memory compensation mechanism in the fractional-order difference for the data not collected below 300℃, establish the Nabla discrete fractional-order hybrid time delay model of the MPCVD system;

[0009] S3: Construct a system of linear matrix inequalities for stability determination based on the model, and use a solver to perform online feasibility determination;

[0010] S4: When the system of linear matrix inequalities has a feasible solution, determine that the current system satisfies the stability condition and maintain the current operating condition;

[0011] S5: When the system of linear matrix inequalities does not have a feasible solution, it is determined that the current system does not meet the stability condition, and a control adjustment quantity is generated;

[0012] S6: Convert the control adjustment quantity into an execution signal and send it to the microwave power controller, automatic pressure regulating valve and / or mass flow controller to regulate the MPCVD system.

[0013] Preferably, the system state vector in step S1 is ,in, express Vioclimatic space For discrete time steps, the state vector includes at least one or more of the following: chamber pressure, sill temperature, and microwave input power data;

[0014] The system state vector is centered as

[0015] ;

[0016] in:

[0017] : 3D real vector, , indicating the system state;

[0018] : A 1D real vector, the centered state vector. ;

[0019] : 3D real vector, front The state mean vector of each sampling point;

[0020] : Vioclimatic space;

[0021] : Dimension of the state vector;

[0022] Discrete time step;

[0023] : The number of valid sampling points in the initial stage of the identification window;

[0024] The inverse of the normalized scaling factor;

[0025] : A 3D real vector, a state vector that is centered and then normalized. .

[0026] Preferably, the Nabla discrete fractional-order hybrid time-delay model described in step S2 is expressed as follows:

[0027] ;

[0028] In step S2, after introducing an initial memory compensation mechanism in the fractional difference method for the data not collected below 300℃, the model is represented as follows:

[0029] ;

[0030] in:

[0031] Caputo-type Nabla fractional difference operator;

[0032] : 3D real vector, , indicating the system state;

[0033] : 3D real vector, System delay The state of the step;

[0034] : 3D real vector, This indicates that the system starts from... The cumulative state term delayed from step r to step r;

[0035] : The current state coupling matrix;

[0036] : Discrete time-delay state matrix;

[0037] : Distributed time-delay state matrix;

[0038] : Vioclimatic space;

[0039] : 3D real-valued matrix space;

[0040] : Dimension of the state vector;

[0041] Discrete time step;

[0042] : No. step;

[0043] Fractional order This characterizes the strength of the MPCVD system's memory of historical states;

[0044] Time delay steps ;

[0045] Initial memory compensation term.

[0046] Preferably, the system of linear matrix inequalities for stability determination in step S3 includes:

[0047] There exists a positive definite matrix symmetric matrix and free matrix Make the following linear matrix inequalities hold simultaneously:

[0048] ;

[0049] as well as

[0050] ;

[0051] in:

[0052] : 3D real-valued matrix space;

[0053] : 3D real-valued matrix space;

[0054] : 3D real-valued matrix space;

[0055] : 3D positive definite matrix;

[0056] : 3D positive definite matrix;

[0057] : 3D positive definite matrix;

[0058] : 3D positive definite matrix;

[0059] : 3D symmetric matrix;

[0060] : 2D symmetric matrix;

[0061] : 3D free matrix;

[0062] : 3D free matrix;

[0063] : 3D free matrix;

[0064] :matrix Negative definite;

[0065] : The block matrix is ​​positive definite;

[0066] : Symmetrical block placeholder.

[0067] Preferably, the The structure satisfies:

[0068] ;

[0069] ;

[0070] ;

[0071] ;

[0072] in:

[0073] : Represents a matrix With matrix The sum of;

[0074] : Represents any matrix;

[0075] : Represents a matrix transpose;

[0076] : indicates the first Unit block vectors;

[0077] , , , Auxiliary vector;

[0078] , Auxiliary matrix;

[0079] : A 3rd order zero matrix;

[0080] : 3rd order identity matrix;

[0081] : Number of time delay steps;

[0082] : The square of the time delay steps;

[0083] : Represents a matrix;

[0084] : Current state coupling matrix;

[0085] Discrete time-delay state matrix;

[0086] Distributed time-delay state matrix;

[0087] Positive definite matrix;

[0088] Positive definite matrix;

[0089] Positive definite matrix;

[0090] Positive definite matrix;

[0091] Symmetric matrix;

[0092] Symmetric matrix;

[0093] Free matrix;

[0094] Free matrix;

[0095] : Free matrix.

[0096] Preferably, the online feasibility determination method in step S4 is as follows: when the system of linear matrix inequalities has a solution for decision variables that satisfies the conditions, the system is determined to be asymptotically stable under the current operating conditions; when the system of linear matrix inequalities does not have a solution for decision variables that satisfies the conditions, the system is determined to be at risk of instability under the current operating conditions.

[0097] Preferably, the control adjustment amount is generated in step S5 mainly through the following methods:

[0098] The Nabla discrete fractional-order hybrid time-delay model described in step S2, after adding the controller, is the model after...

[0099] ;

[0100] in:

[0101] Caputo-type Nabla fractional difference operator

[0102] : 3D real vector, , indicating the system state;

[0103] : 3D real vector, System delay The state of the step;

[0104] : 3D real vector, This indicates that the system starts from... The cumulative state term delayed from step r to step r;

[0105] : Control the input vector;

[0106] State feedback control law;

[0107] : The current state matrix of the system;

[0108] : Discrete time-delay state matrix;

[0109] : Distributed time-delay state matrix;

[0110] : Input the allocation matrix;

[0111] : , The state feedback gain matrix;

[0112] : Vioclimatic space;

[0113] : 3D real-valued matrix space;

[0114] : Dimension of the state vector;

[0115] Discrete time step;

[0116] : No. step;

[0117] Fractional order It represents the strength of memory;

[0118] Time delay steps ;

[0119] Instructions issued to the executor;

[0120] Initial state value;

[0121] : Actuator conversion coefficient.

[0122] Preferably, the control adjustment amount in step S5 is used to adjust the microwave power setting value, chamber pressure setting value and / or process gas flow rate setting value of the MPCVD equipment to change the subsequent operating state of the system, and after adjustment, steps S1 to S4 are re-executed until the system meets the stability conditions.

[0123] Preferably, the process of converting the control adjustment amount into an execution signal in step S6 is as follows:

[0124] ;

[0125] Where ucmd(r) is the correction signal vector sent to the actuator, ∆u(r) is the control adjustment vector, ωout(r) is the current setpoint vector of the actuator at the moment of instability risk, and ϱ is the actuator conversion coefficient matrix, ensuring that the issued instructions do not exceed the working range allowed by the MPCVD equipment;

[0126] Furthermore, a stability monitoring system for an MPCVD system based on a discrete fractional-order model is provided. This system mainly includes: a data acquisition module for real-time acquisition of the operating status data of the MPCVD equipment and the corresponding single-crystal diamond substrate infrared temperature and for constructing a state vector; a model construction module for establishing a Nabla discrete fractional-order hybrid time-delay model based on the acquired data; a stability determination module for constructing a system of linear matrix inequalities based on the model and performing online feasibility determination; a control module for generating control adjustment quantities when the current system is determined not to meet stability conditions; and an actuator module for receiving the control adjustment quantities and driving a microwave power controller, an automatic pressure regulating valve, and / or a mass flow controller to perform adjustments.

[0127] Compared with existing technologies, the present invention has the following advantages: by adopting a discrete fractional time delay model, the memory effect and time delay effect of the MPCVD system can be characterized simultaneously; by constructing a stability criterion through the free matrix method, discrete time delay terms and distributed time delay terms can be uniformly incorporated into the stability analysis framework, thereby improving the theoretical rigor of the judgment results; by directly linking the stability monitoring results with the execution module, microwave power, chamber pressure and process flow can be coordinated and adjusted, which is beneficial to improving the stability of equipment operation and process continuity. Attached Figure Description

[0128] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, use the same reference numerals to denote the same or similar parts. The illustrative embodiments of this application and their descriptions are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0129] Figure 1 This is a flowchart of the present invention.

[0130] Figure 2 These are the system state response curves in this embodiment of the invention, with and without control. Detailed Implementation

[0131] The technical solution of the present invention will be further described below with reference to the accompanying drawings. The following embodiments are used to illustrate the technical concept and implementation process of the present invention.

[0132] In the following description, references to "an embodiment," "an embodiment," "an example," "example," etc., indicate that the described embodiment or example may include a particular feature, structure, characteristic, property, element, or limitation, but not every embodiment or example necessarily includes that particular feature, structure, characteristic, property, element, or limitation. Furthermore, the repeated use of the phrase "an embodiment according to this application," while possibly referring to the same embodiment, does not necessarily refer to the same embodiment.

[0133] Example 1: A Stability Monitoring Method for an MPCVD System with a Discrete Fractional Model

[0134] This embodiment provides a stability monitoring method for an MPCVD system with a discrete fractional-order model. It includes the following steps:

[0135] S1: Real-time acquisition of microwave input power data and corresponding infrared temperature data of the single-crystal diamond substrate, constructing the system state vector as follows: ,in, express Vioclimatic space For discrete time steps, the state vector includes at least one or more of the following: chamber pressure, sill temperature, and microwave input power data;

[0136] The system state vector is centered as

[0137] ;

[0138] in, To identify the number of valid sampling points before the initial stage of the window;

[0139] The normalized state vector is

[0140] ;

[0141] in, This is the normalized scaling factor.

[0142] In this embodiment, a 2.45GHz microwave source is used, with an input power limit set to 9500W. Real-time equipment operation data is collected during the single-crystal diamond MPCVD growth experiment. Output temperature is provided by an infrared temperature measurement link, output pressure by a cavity pressure sensor, and output power by a microwave source feedback signal. The set power and set pressure are read from the current set values ​​of the equipment controller. The control unit performs time alignment, outlier removal, and necessary smoothing processing on the collected data. Due to the three quantities of the system... All of them have a relatively high static operating level. To highlight the dynamic deviation relationship, the mean of 10 samples at the beginning of the identification window is selected as the reference operating point, and the deviation state is defined.

[0143] ;

[0144] in, Indicates the output temperature deviation. This indicates the output pressure deviation. This represents the output power deviation; to ensure dimensional consistency between the identification model, controller parameters, and equipment control quantities, a normalized state vector is further constructed.

[0145] ;

[0146] in, , These represent the normalized scaling factors for temperature deviation, pressure deviation, and output power deviation, respectively, using their respective standard deviations as the normalized scaling factors.

[0147] S2: Based on the data obtained after processing in step S1, the dynamic evolution process of MPCVD plasma is modeled as a Nabla discrete fractional-order hybrid time-delay system using a data-driven method:

[0148] ;

[0149] in, Used to characterize the strength of an MPCVD system's memory of historical states. The matrix is ​​used to characterize the effective discrete time delay caused by power regulation, gas path regulation, and cavity response. Corresponding to the coupling effect of the current state, the matrix Corresponding to the discrete time-delay state, the matrix The distribution lag effect caused by the accumulation of states over a period of time corresponds to the following: For data not collected below 300℃, after introducing an initial memory compensation mechanism in the fractional difference, the model is expressed as follows:

[0150]

[0151] in, This is the initial memory compensation term, used to recover historical effects lost due to observation window truncation.

[0152] In this embodiment, the data acquired in S1 is processed, and the model parameters are obtained by calling the program in the model building module as follows:

[0153]

[0154] ;

[0155] ;

[0156] ;

[0157] ;

[0158] Reference working point ; the above matrix , , When the open-loop stability criterion is substituted to solve for the feasibility of LMI, no feasible solution is found. Therefore, it is determined that the current open-loop model does not meet the preset stability conditions, and the system enters the control preparation state.

[0159] S3: Matrix extracted by the industrial control computer based on the current working point. , , A system of linear matrix inequalities for stability determination is constructed using the free matrix method, and a solver is invoked to perform feasibility analysis; the criteria include:

[0160] ;

[0161] as well as

[0162] ;

[0163] Among them, there exists a positive definite matrix. symmetric matrix and free matrix This makes the system of linear matrix inequalities hold; The structure satisfies:

[0164] ,

[0165] , , ,

[0166] in, express 3D real-valued matrix space, express 3D real-valued matrix space, express 3D real-valued matrix space, Represents a 3rd order zero matrix. This represents a 3rd order identity matrix.

[0167] The results obtained through the stability determination module are as follows:

[0168] ;

[0169] The feasibility assessment index for the linear matrix inequality system is as follows: .when When, it indicates that the system of linear matrix inequalities has no feasible solution, and the system does not satisfy the stability condition; when When the condition is met, it indicates that the system of linear matrix inequalities has a feasible solution and the system is asymptotically stable.

[0170] S4: When the system of linear matrix inequalities in step S3 has a feasible solution, it is determined that the system under the current operating parameters meets the stability condition; when the system of linear matrix inequalities does not have a feasible solution, it is determined that the system does not meet the stability condition, and the control module generates a control adjustment quantity.

[0171] In this embodiment, the result obtained by S3 This indicates that the system of linear matrix inequalities has no feasible solution, therefore the MPCVD system does not satisfy the stability condition.

[0172] S5: When the system of linear matrix inequalities has no feasible solution, the system is determined to not meet the stability condition, triggering a control command. The industrial control computer then calls the solver to solve the matrix. and The control module generates the control adjustment amount. and Mainly sourced from:

[0173] The Nabla discrete fractional-order hybrid time-delay model described in step S2, after adding the controller, is the model after...

[0174] ;

[0175] in, To control the input vector, Assign a matrix to the input; the control input adopts the following state feedback form. ;in, This is the state feedback gain matrix. After the controller intervenes, the current system state matrix is ​​determined by... Become This forms a closed-loop controlled model; the closed-loop matrix is ​​then substituted into step S4, and the particle swarm optimization algorithm is used to solve for the closed-loop matrix. and This ensures that the system of linear matrix inequalities has a solution for the decision variables that satisfies the conditions.

[0176] In this embodiment, the control input is defined as:

[0177] ;

[0178] Input assignment matrix The following structural constraint is used to characterize the different action paths of the set power channel and the set pressure channel on the three state components:

[0179] ;

[0180] in, Represents the set of real numbers. This indicates the weight of the positive effect of the set power on the temperature state. This represents the weight of the coupling effect of set pressure on temperature state; its sign can vary depending on the operating conditions. This indicates the weight of the positive effect of the set pressure on the pressure state. This represents the weight of the positive effect of the set power on the output power state; the matrix that makes the closed-loop criterion hold is found through the particle swarm optimization algorithm. and The feasible solutions are as follows:

[0181] ;

[0182] ;

[0183] ;

[0184] ; ; ; ; ; ; ; ; ;

[0185] In this embodiment, the obtained result This indicates that the system of linear matrix inequalities has a feasible solution, therefore the MPCVD system is stable; Figure 2 As shown, the system is unstable and exhibits a divergent trend when no control is applied; after the controller is applied, the system's states return to stability within a finite time.

[0186] S6: The control adjustment quantity obtained in step S5 is converted into an analog signal and / or a digital signal, and sent to the microwave power controller, automatic pressure regulating valve, and / or mass flow controller of the MPCVD equipment to adjust the system operating status; the process of converting the control adjustment quantity into an execution signal is as follows:

[0187] ;

[0188] Wherein, ucmd(r) is the correction signal vector sent to the actuator, ∆u(r) is the control adjustment vector, ωout(r) is the current setpoint vector of the actuator at the moment of instability risk, and ϱ is the actuator conversion coefficient matrix, ensuring that the issued instructions do not exceed the working range allowed by the MPCVD equipment.

[0189] In this embodiment, the controller sends control commands to the microwave power controller and the automatic voltage regulating valve, and obtains the feedback gain matrix. Then, the two-dimensional control adjustment amount is calculated based on the real-time state deviation.

[0190] Based on the reference operating point calculated in step 2, to avoid excessive disturbance caused by the control adjustment directly acting on the device's execution end, a safe adjustment range for the device is set; specifically, the maximum permissible single adjustment power of the microwave power supply is 100W, and the maximum permissible single adjustment pressure of the automatic pressure regulating valve is... When a risk of system instability is detected, that moment is recorded as... And select the initial state value at that moment. 5000; Actuator conversion factor Let the initial state deviation be denoted as

[0191] ;

[0192] The following state feedback is used to calculate the two-dimensional control adjustment:

[0193] ;

[0194] Among them, the first control adjustment amount for:

[0195] Second control adjustment amount for:

[0196] Furthermore, the first control adjustment amount Convert to microwave power supply power adjustment amount; record The setpoint for the microwave power supply signal is calculated as follows:

[0197] ;

[0198] At this time, the microwave power supply adjustment amount is

[0199] ;

[0200] Satisfy the single maximum allowable adjustment power constraint;

[0201] Similarly, the second control adjustment amount Convert to reaction chamber pressure adjustment amount, and record. The set value for the reaction chamber pressure is calculated as follows:

[0202] ;

[0203] At this time, the single adjustment amount of the reaction chamber pressure is:

[0204] ;

[0205] The maximum permissible adjustment pressure constraint for a single operation must be met.

[0206] Finally, the actuator issues a command to adjust the power setting to 4950W and the pressure to 131 Torr. After adjustment, steps S1 to S6 are repeated.

[0207] The embodiments described above are merely one implementation of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the claims.

Claims

1. A method for stability monitoring of an MPCVD system with a discrete fractional-order model, characterized in that, Includes the following steps: S1: Real-time acquisition of the operating status data of the MPCVD equipment and the corresponding single-crystal diamond substrate infrared temperature, construction of state vector, state vector centering and normalization processing; S2: Based on the data collected in step S1 and introducing an initial memory compensation mechanism in the fractional-order difference for the data not collected below 300℃, establish the Nabla discrete fractional-order hybrid time delay model of the MPCVD system; S3: Construct a system of linear matrix inequalities for stability determination based on the model, and use a solver to perform online feasibility determination; S4: When the system of linear matrix inequalities has a feasible solution, determine that the current system satisfies the stability condition and maintain the current operating condition; S5: When the system of linear matrix inequalities does not have a feasible solution, it is determined that the current system does not meet the stability condition, and a control adjustment quantity is generated; S6: Convert the control adjustment quantity into an execution signal and send it to the microwave power controller, automatic pressure regulating valve and / or mass flow controller to regulate the MPCVD system.

2. The method for stability monitoring of an MPCVD system using a discrete fractional-order model according to claim 1, characterized in that, The system state vector mentioned in step S1 is ,in, express Vioclimatic space For discrete time steps, the state vector includes at least one or more of the following: chamber pressure, sill temperature, and microwave input power data; The system state vector is centered as ； in: : 3D real vector, , indicating the system state; : A 1D real vector, the centered state vector. ; : 3D real vector, front The state mean vector of each sampling point; : Vioclimatic space; : Dimension of the state vector; Discrete time step; : The number of valid sampling points in the initial stage of the identification window; The inverse of the normalized scaling factor; : A 3D real vector, a state vector that is centered and then normalized. .

3. The method for stability monitoring of an MPCVD system using a discrete fractional-order model according to claim 1, characterized in that, The Nabla discrete fractional-order hybrid time-delay model described in step S2 is expressed as follows: ； In step S2, after introducing an initial memory compensation mechanism in the fractional difference method for the data not collected below 300℃, the model is represented as follows: ； in: Caputo-type Nabla fractional difference operator; : 3D real vector, , indicating the system state; : 3D real vector, System delay The state of the step; : 3D real vector, This indicates that the system starts from... The cumulative state term delayed from step r to step r; : The current state coupling matrix; : Discrete time-delay state matrix; : Distributed time-delay state matrix; : Vioclimatic space; : 3D real-valued matrix space; : Dimension of the state vector; Discrete time step; : No. step; Fractional order This characterizes the strength of the MPCVD system's memory of historical states; Time delay steps ; Initial memory compensation term.

4. The method for stability monitoring of an MPCVD system using a discrete fractional-order model according to claim 1, characterized in that, The stability determination linear matrix inequalities in step S3 include: There exists a positive definite matrix symmetric matrix and free matrix Make the following linear matrix inequalities hold simultaneously: ; as well as ； in: : 3D real-valued matrix space; : 3D real-valued matrix space; : 3D real-valued matrix space; : 3D positive definite matrix; : 3D positive definite matrix; : 3D positive definite matrix; : 3D positive definite matrix; : 2D symmetric matrix; : 2D symmetric matrix; : 3D free matrix; : 3D free matrix; : 3D free matrix; :matrix Negative definite; : The block matrix is ​​positive definite; : Symmetrical block placeholder.

5. The method for stability monitoring of an MPCVD system using a discrete fractional-order model according to claim 4, characterized in that, The The structure satisfies: ； ； ； ； in: : Represents a matrix With matrix The sum of; : Represents any matrix; : Represents a matrix Transpose of; : indicates the first Unit block vectors; , , , Auxiliary vector; , Auxiliary matrix; : A 3rd order zero matrix; : 3rd order identity matrix; : Number of time delay steps; : The square of the time delay steps; : Represents a matrix; : Current state coupling matrix; Discrete time-delay state matrix; Distributed time-delay state matrix; Positive definite matrix; Positive definite matrix; Positive definite matrix; Positive definite matrix; Symmetric matrix; Symmetric matrix; Free matrix; Free matrix; : Free matrix.

6. The method for stability monitoring of an MPCVD system using a discrete fractional-order model according to claim 1, characterized in that, The online feasibility determination method in step S4 is as follows: when the system of linear matrix inequalities has a solution for decision variables that satisfies the conditions, the system is determined to be asymptotically stable under the current operating conditions; when the system of linear matrix inequalities does not have a solution for decision variables that satisfies the conditions, the system is determined to be at risk of instability under the current operating conditions.

7. The method for stability monitoring of an MPCVD system using a discrete fractional-order model according to claim 1, characterized in that, In step S5, the control adjustment amount is mainly generated in the following ways: The Nabla discrete fractional-order hybrid time-delay model described in step S2, after adding the controller, is the model after... ； in: Caputo-type Nabla fractional difference operator; : 3D real vector, , indicating the system state; : 3D real vector, System delay The state of the step; : 3D real vector, This indicates that the system starts from... The cumulative state term delayed from step r to step r; : Control the input vector; State feedback control law; : The current state matrix of the system; : Discrete time-delay state matrix; : Distributed time-delay state matrix; : Input the allocation matrix; : , The state feedback gain matrix; : Closed-loop system state matrix; : Vioclimatic space; : 3D real-valued matrix space; : Dimension of the state vector; Discrete time step; : No. step; Fractional order It represents the strength of memory; Time delay steps .

8. The method for stability monitoring of an MPCVD system using a discrete fractional-order model according to claim 1, characterized in that, The control adjustment in step S5 is used to adjust the microwave power setting, chamber pressure setting and / or process gas flow setting of the MPCVD equipment to change the subsequent operating state of the system, and after adjustment, steps S1 to S4 are re-executed until the system meets the stability conditions.

9. The method for stability monitoring of an MPCVD system using a discrete fractional-order model according to claim 1, characterized in that, The process of converting the control adjustment quantity into an execution signal in step S6 is as follows: ； in, This is the correction signal vector sent to the actuator. To control the adjustment vector, This represents the actuator's current setpoint vector at the moment of instability risk. This is the actuator conversion coefficient matrix, ensuring that the issued instructions do not exceed the operating range allowed by the MPCVD device.

10. A stability monitoring system for an MPCVD system with a discrete fractional-order model, characterized in that, include: The data acquisition module is used to acquire the operating status data of the MPCVD equipment and the corresponding single-crystal diamond substrate infrared temperature in real time and construct a state vector. The model building module is used to build a Nabla discrete fractional-order hybrid time-delay model based on the collected data. The stability determination module is used to construct a system of linear matrix inequalities based on the model and perform online feasibility determination. The control module is used to generate control adjustment quantities when it is determined that the current system does not meet the stability conditions; An actuator module is used to receive the control adjustment amount and drive the microwave power controller, automatic pressure regulating valve and / or mass flow controller to perform the adjustment.

Images