CN118497722B - High-temperature ionization control system in MPCVD diamond cultivation process - Google Patents
High-temperature ionization control system in MPCVD diamond cultivation process Download PDFInfo
- Publication number
- CN118497722B CN118497722B CN202410955835.6A CN202410955835A CN118497722B CN 118497722 B CN118497722 B CN 118497722B CN 202410955835 A CN202410955835 A CN 202410955835A CN 118497722 B CN118497722 B CN 118497722B
- Authority
- CN
- China
- Prior art keywords
- plasma
- data
- temperature
- pressure
- module
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 229910003460 diamond Inorganic materials 0.000 title claims abstract description 25
- 239000010432 diamond Substances 0.000 title claims abstract description 25
- 238000000259 microwave plasma-assisted chemical vapour deposition Methods 0.000 title claims abstract description 22
- 238000000034 method Methods 0.000 title claims abstract description 19
- 230000008569 process Effects 0.000 title claims abstract description 16
- 238000003745 diagnosis Methods 0.000 claims abstract description 50
- 238000004891 communication Methods 0.000 claims abstract description 26
- 238000013528 artificial neural network Methods 0.000 claims abstract description 22
- 238000005516 engineering process Methods 0.000 claims abstract description 20
- 230000003993 interaction Effects 0.000 claims abstract description 16
- 239000012495 reaction gas Substances 0.000 claims abstract description 4
- 239000000284 extract Substances 0.000 claims abstract description 3
- 230000002085 persistent effect Effects 0.000 claims description 34
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 claims description 30
- 239000000758 substrate Substances 0.000 claims description 26
- 230000006870 function Effects 0.000 claims description 25
- 230000001427 coherent effect Effects 0.000 claims description 23
- 238000004458 analytical method Methods 0.000 claims description 21
- 238000004364 calculation method Methods 0.000 claims description 21
- 239000007789 gas Substances 0.000 claims description 21
- 239000011159 matrix material Substances 0.000 claims description 21
- UFHFLCQGNIYNRP-UHFFFAOYSA-N Hydrogen Chemical compound [H][H] UFHFLCQGNIYNRP-UHFFFAOYSA-N 0.000 claims description 18
- 229910052739 hydrogen Inorganic materials 0.000 claims description 18
- 239000001257 hydrogen Substances 0.000 claims description 18
- 238000011176 pooling Methods 0.000 claims description 15
- 238000013527 convolutional neural network Methods 0.000 claims description 9
- 238000011156 evaluation Methods 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 9
- 230000001105 regulatory effect Effects 0.000 claims description 8
- 230000001276 controlling effect Effects 0.000 claims description 7
- 230000004913 activation Effects 0.000 claims description 6
- 230000008033 biological extinction Effects 0.000 claims description 6
- 238000013178 mathematical model Methods 0.000 claims description 6
- 238000012544 monitoring process Methods 0.000 claims description 6
- 210000002569 neuron Anatomy 0.000 claims description 6
- 238000001228 spectrum Methods 0.000 claims description 6
- 238000001914 filtration Methods 0.000 claims description 5
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 claims description 3
- 229910052799 carbon Inorganic materials 0.000 claims description 3
- 230000015556 catabolic process Effects 0.000 claims description 3
- 210000004027 cell Anatomy 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 230000004927 fusion Effects 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 239000007787 solid Substances 0.000 claims description 3
- 238000010937 topological data analysis Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 238000012423 maintenance Methods 0.000 claims 1
- 238000004519 manufacturing process Methods 0.000 abstract description 4
- 210000002381 plasma Anatomy 0.000 description 103
- 230000006872 improvement Effects 0.000 description 8
- 230000008713 feedback mechanism Effects 0.000 description 2
- 230000002688 persistence Effects 0.000 description 2
- 230000036470 plasma concentration Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Landscapes
- Plasma Technology (AREA)
- Chemical Vapour Deposition (AREA)
Abstract
The invention relates to the technical control field of MPCVD high-temperature ionization control, in particular to a high-temperature ionization control system in the MPCVD diamond cultivation process. The data acquisition unit is responsible for the acquisition and storage of high-precision sensor data; the plasma diagnosis unit extracts and analyzes the data, and uses a topological neural network plasma diagnosis technology to diagnose the plasma state; the plasma control unit adjusts microwave power and reaction gas flow according to the diagnosis result to maintain the stability of plasma; the interaction and communication unit provides an interface between the user and the control system, displays real-time data and control states, and realizes communication between each unit and external equipment. The control system monitors the feedback plasma state in real time by using the topology neural network technology, dynamically adjusts each parameter by using the self-adaptive control algorithm, and improves the quality and the production efficiency of the diamond film.
Description
Technical Field
The invention relates to the technical field of MPCVD high-temperature ionization control, in particular to a high-temperature ionization control system in the MPCVD diamond cultivation process.
Background
The high temperature ionization control system in MPCVD diamond cultivation process utilizes topological neural network plasma diagnosis technology to realize high-precision and real-time plasma parameter monitoring and control, aims at ensuring the stability of plasma and optimizing the growth condition of diamond film, and combines self-adaptive control and intelligent feedback mechanism to realize dynamic adjustment of microwave power, gas flow, reaction chamber pressure and substrate temperature.
The traditional high-temperature ionization control system in the MPCVD diamond cultivation process cannot accurately monitor and feed back the plasma state in real time, and due to the lack of accurate monitoring and control of plasma concentration, pressure in a reaction cavity and temperature conditions, the problems of unstable plasma, uneven quality of a diamond film and low production efficiency can be caused, so that the high-temperature ionization control system in the MPCVD diamond cultivation process is designed.
Disclosure of Invention
The invention aims to provide a high-temperature ionization control system in the MPCVD diamond cultivation process, so as to solve the problems of unstable plasmas, uneven quality of diamond films and low production efficiency caused by lack of accurate monitoring and control of plasma concentration, pressure in a reaction cavity and temperature conditions in the background technology.
To achieve the above object, the present invention aims to provide a high temperature ionization control system in MPCVD diamond cultivation process, comprising:
The data acquisition unit acquires and stores high-precision sensor data;
The system also comprises a plasma diagnosis unit, wherein the plasma diagnosis unit extracts real-time sensor data in the data acquisition unit, calculates and analyzes plasma data by using a topological neural network plasma diagnosis technology, and diagnoses the plasma state;
the plasma control unit is used for adjusting and controlling the microwave power to generate and maintain plasma and controlling the flow of methane and hydrogen reaction gas according to the diagnosis result of the plasma diagnosis unit;
The environment regulation and control unit is used for analyzing the pressure data in the reaction cavity, the substrate temperature data and the diagnosis result state of the plasma diagnosis unit in the data acquisition unit and regulating the pressure in the reaction cavity and the substrate temperature;
The system also comprises an interaction and communication unit, wherein the interaction and communication unit provides an interface for interaction between a user and the control system, displays real-time data and control states, and realizes communication between the units and between the control system and external equipment.
The data acquisition unit comprises a sensor interface module and a data acquisition and storage module;
The sensor interface module is connected with and integrated with a high-precision sensor, and is used for monitoring different parameters in the system, wherein the high-precision sensor comprises an infrared temperature sensor, a capacitive pressure sensor, a mass flow controller, a spectrum analyzer and a laser-induced breakdown spectrum;
The data acquisition and storage module is used for acquiring high-precision sensor data in real time, storing the acquired sensor data into SSD, and storing and managing the sensor data by using PostgreSQL; wherein the sensor data includes substrate temperature, reaction chamber pressure, methane flow, hydrogen flow, plasma electron density, and plasma temperature.
As a further improvement of the technical scheme, the plasma diagnosis unit comprises a data processing module, a parameter calculation and analysis module and a reporting and feedback module;
The data processing module is used for extracting relevant plasma data in the data acquisition unit and carrying out primary processing of filtering, denoising and calibration on the plasma data; the plasma data comprises substrate temperature, reaction chamber pressure, methane flow, hydrogen flow, plasma electron density and plasma temperature;
the parameter calculation and analysis module uses a topological neural network plasma diagnosis technology to realize real-time calculation and analysis of plasma data parameters;
The report and feedback module is used for generating a diagnosis report of the parameter calculation and analysis module and feeding back an analysis result to the interaction and communication unit so as to be displayed to a user for real-time check through the display.
As a further improvement of the technical scheme, the topological neural network plasma diagnosis technology is realized based on the fusion of topological data analysis and the deep neural network technology and is used for diagnosing and analyzing the plasma state in the MPCVD diamond cultivation process in real time; the steps involved in the real-time calculation and analysis of the plasma data parameters by the parameter calculation and analysis module using the topological neural network plasma diagnosis technology are as follows:
s2.1, constructing an original data vector :
Wherein, Is the substrate temperature; the reaction chamber temperature; is the pressure of the reaction chamber; is methane flow; Is the hydrogen flow; Is the plasma electron density; Is the plasma temperature;
S2.2, calculating a distance matrix between the original plasma data Constructing Vietoris-Rips complex shape, and calculating topological characteristics under different scales, including connected components, rings and cavities, by using persistent coherent calculation;
S2.3, converting the persistent coherent result into numerical characteristics processed by the neural network, and constructing a topological characteristic vector :
Wherein, Is a topological feature vector; the numerical characteristics extracted from the persistent coherent including the total number of the persistent coherent, the persistent coherent number of each dimension and the persistent time of each coherent class;
S2.4, carrying out topological feature vector by using convolutional neural network And (5) analyzing and calculating to comprehensively diagnose the plasma state.
As a further improvement of the present technical solution, the distance matrix between the original plasma data is calculated in S2.2The Vietoris-Rips complex form is constructed, and the specific steps and mathematical formulas involved in calculating topological features under different scales by using persistent coherence are as follows:
S2.2.1 building a distance matrix Calculating a distance matrix between data pointsFor constructing Vietoris-Rips complex:
Wherein, Is the firstData points; Is the first Data points; Representing euclidean distance;
s2.2.2 based on distance matrix Selecting a scale parameterConstruction Vietoris-Rips ComplexIf (if)Vertex thenAnd a vertexAn edge is arranged between the two edges;
s2.2.3 calculating persistent coherence for identifying stable topological features in data, and calculating topological features under different scales by using the persistent coherence, wherein the topological features comprise connected components, rings and cavities:
Wherein, Is the firstA dimension coherent group; Is the first Generation time of the homonym; Is the first The death time of the same class;
s2.2.4 constructing a persistent bar code to show the persistence of topological features, wherein each bar represents the generation and extinction of a homonym.
As a further improvement of the technical scheme, in S2.4, the topology feature vector is performed by using a convolutional neural networkThe specific steps and mathematical formulas involved in the comprehensive diagnosis of the plasma state are as follows:
s2.4.1 input layer of convolutional neural network receives topology feature vector The convolution layer applies a convolution kernel to slide over the input feature vector to generate a feature map:
Wherein, Is the firstThe weights of the convolution kernels; Is the first Offset of the individual convolution kernels; an index that is a convolution kernel; is the length of the convolution kernel;
S2.4.2, nonlinear activation function ReLU, dimension of feature mapping is reduced by downsampling at the pooling layer using maximum pooling:
Wherein, Is the size of the pooling window;
S2.4.3, the full connection layer expands the output of the pooling layer into a one-dimensional vector, is connected to the output layer, and calculates through linear transformation and combination with an activation function:
Wherein, Is the firstA weight vector of the individual neurons; expanding the output vector of the pooling layer into a one-dimensional vector; Is the first Bias of individual neurons;
S2.4.4 and output layer generation predicted plasma parameters :
Wherein, To predict plasma parameters; a weight matrix for an output layer; the output of the full connection layer; is a bias vector for the output layer;
S2.4.5, using the mean square error as a loss function, measuring the error between the predicted value and the true value:
Wherein, As a loss function; is a predicted value; Is a true value; is the number of samples;
s2.4.6, calculating the gradient of the loss function relative to the weight by a back propagation algorithm, and updating the weight:
Wherein, Is the learning rate; Gradient of weight matrix for loss function;
s2.4.7, performing plasma state estimation based on predicted plasma parameters:
Wherein, Is the plasma state evaluation result; Is a state evaluation function.
As a further improvement of the technical proposal, the S2.4.7Is the result of the plasma state evaluation, the plasma stateThe method comprises five states:
State one, plasma stability: indicating the stability of the plasma, whether there are fluctuations and instabilities, including stable, slightly unstable and unstable states;
State two, temperature deviation: the deviation between the predicted temperature and the target temperature comprises three states of normal temperature, higher temperature and lower temperature;
State three, pressure deviation: predicting deviation of pressure and target pressure, wherein the deviation comprises three states of normal pressure, high pressure and low pressure;
state four, gas flow deviation: predicting deviation of the gas flow and a set value, wherein the deviation comprises three states of normal flow, higher flow and lower flow;
state five, electron density: reflecting whether the electron density of the plasma is in a set range, including three states of normal electron density, too high electron density and too low electron density.
As a further improvement of the technical scheme, the plasma control unit comprises a microwave power adjusting module and a gas flow control module;
The microwave power adjusting module adjusts microwave power entering the reaction cavity by using a solid microwave source so as to generate and maintain plasma;
The gas flow control module is used for adjusting the proportion of methane and hydrogen and controlling the carbon concentration and the hydrogen concentration in the plasma.
As a further improvement of the technical scheme, the environment regulation and control unit comprises a pressure control module and a temperature control module;
wherein, the pressure control module acquires real-time pressure data in the reaction cavity from the data acquisition unit And regulating the pressure in the reaction cavity according to the pressure data fed back by the capacitive pressure sensor by using a pressure PID controller, wherein a mathematical model formula related to the pressure PID control is as follows:
Wherein, Is set to a pressure; To measure pressure; As a result of the error in the error, ;、、Respectively proportional, integral and differential coefficients;
Wherein the temperature control module acquires temperature data of the substrate from the data acquisition unit And regulating the temperature of the substrate according to temperature data fed back by an infrared temperature sensor by using a temperature PID controller, wherein a mathematical model formula related to the temperature PID control is as follows:
Wherein, Is a set temperature; to measure temperature; As a result of the error in the error, ;、、Respectively proportional, integral and differential coefficients.
As a further improvement of the technical scheme, the interaction and communication unit comprises a user interface module and a network communication module;
The user interface module provides a graphical interface and a control panel for a user to view and operate various functions of the system, including real-time data and control states;
The graphical interface displays real-time data from the data acquisition unit, including temperature, pressure, gas flow and plasma parameters; displaying the diagnosis result of the plasma diagnosis unit, the control result of the plasma control unit and the environment regulation unit;
the control panel is used for manually adjusting microwave power, gas flow, reaction cavity pressure and substrate temperature by a user;
the network communication module is used for realizing network communication between the control system and external equipment; external devices include user computers, cell phones, and other terminals.
Compared with the prior art, the invention has the beneficial effects that:
1. in the high-temperature ionization control system in the MPCVD diamond cultivation process, the topological neural network plasma diagnosis technology is used for calculating and analyzing plasma data, the plasma state is diagnosed, key parameters such as stability, temperature, pressure and gas flow of the plasma can be monitored and fed back in real time, the plasma is ensured to be always in an optimal working state, and unstable conditions are avoided.
2. In the high-temperature ionization control system in the MPCVD diamond cultivation process, accurate dynamic adjustment of microwave power, gas flow, reaction cavity pressure and substrate temperature is realized through a control algorithm and an intelligent feedback mechanism, so that reaction conditions are optimized, and the growth quality and production efficiency of a diamond film are improved.
Drawings
Fig. 1 is an overall flow diagram of the present invention.
The meaning of each reference sign in the figure is:
1. A data acquisition unit; 2. a plasma diagnosis unit; 3. a plasma control unit; 4. an environment control unit; 5. and an interaction and communication unit.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Examples
Referring to FIG. 1, a high temperature ionization control system for MPCVD diamond cultivation is provided, comprising
The data acquisition unit 1 is used for acquiring and storing high-precision sensor data;
The data acquisition unit 1 comprises a sensor interface module and a data acquisition and storage module;
The sensor interface module is connected with and integrated with a high-precision sensor, and is used for monitoring different parameters in the system, wherein the high-precision sensor comprises an infrared temperature sensor, a capacitive pressure sensor, a mass flow controller, a spectrum analyzer and a laser-induced breakdown spectrum;
The data acquisition and storage module is used for acquiring high-precision sensor data in real time, storing the acquired sensor data into SSD, and storing and managing the sensor data by using PostgreSQL; wherein the sensor data includes substrate temperature, reaction chamber pressure, methane flow, hydrogen flow, plasma electron density, and plasma temperature.
The plasma diagnosis unit 2 is used for extracting real-time sensor data in the data acquisition unit 1, calculating and analyzing plasma data by using a topological neural network plasma diagnosis technology, and diagnosing a plasma state;
The plasma diagnosis unit 2 comprises a data processing module, a parameter calculation and analysis module and a reporting and feedback module;
The data processing module is used for extracting relevant plasma data in the data acquisition unit 1 and carrying out primary processing of filtering, denoising and calibration on the plasma data; the plasma data comprises substrate temperature, reaction chamber pressure, methane flow, hydrogen flow, plasma electron density and plasma temperature;
the plasma data is subjected to filtering and denoising treatment, kalman filtering is applied to filter the original data, noise is removed, and the stability of the data is improved;
the parameter calculation and analysis module uses a topological neural network plasma diagnosis technology to realize real-time calculation and analysis of plasma data parameters;
The report and feedback module is used for generating a diagnosis report of the parameter calculation and analysis module, and feeding back an analysis result to the interaction and communication unit 5 so as to be displayed to a user for real-time check through a display.
The topological neural network plasma diagnosis technology is realized based on the fusion of topological data analysis and the deep neural network technology and is used for diagnosing and analyzing the plasma state in the MPCVD diamond cultivation process in real time; the steps involved in the real-time calculation and analysis of the plasma data parameters by the parameter calculation and analysis module using the topological neural network plasma diagnosis technology are as follows:
s2.1, constructing an original data vector :
Wherein, Is the substrate temperature; the reaction chamber temperature; is the pressure of the reaction chamber; is methane flow; Is the hydrogen flow; Is the plasma electron density; Is the plasma temperature;
S2.2, calculating a distance matrix between the original plasma data Constructing Vietoris-Rips complex shape, and calculating topological characteristics under different scales, including connected components, rings and cavities, by using persistent coherent calculation;
S2.3, converting the persistent coherent result into numerical characteristics processed by the neural network, and constructing a topological characteristic vector :
Wherein, Is a topological feature vector; the numerical characteristics extracted from the persistent coherent including the total number of the persistent coherent, the persistent coherent number of each dimension and the persistent time of each coherent class;
S2.4, carrying out topological feature vector by using convolutional neural network And (5) analyzing and calculating to comprehensively diagnose the plasma state.
Calculating a distance matrix between the original plasma data in S2.2The Vietoris-Rips complex form is constructed, and the specific steps and mathematical formulas involved in calculating topological features under different scales by using persistent coherence are as follows:
S2.2.1 building a distance matrix Calculating a distance matrix between data pointsFor constructing Vietoris-Rips complex:
Wherein, Is the firstData points; Is the first Data points; Representing euclidean distance;
s2.2.2 based on distance matrix Selecting a scale parameterConstruction Vietoris-Rips ComplexIf (if)Vertex thenAnd a vertexAn edge is arranged between the two edges;
in this complex form, any two points AndThe edges between exist if and only if the distance between them. Formally, for each oneComplex shapeComprising all simple paths of raw data points connected by edges, wherein the presence of an edge is based on a distance threshold;
S2.2.3 calculating persistent coherence for identifying stable topological features in data, and calculating topological features under different scales by using the persistent coherence, wherein the topological features comprise connected components, rings and cavities:
Wherein, Is the firstA dimension coherent group; Is the first Generation time of the homonym; Is the first The death time of the same class;
persistent coherence is a measure of Vietoris-Rips complex shape follow-up As it increases, it is a tool how its topology changes. It is obtained by computing k-dimensional holes (k=0, 1,2,) that are complex at different scales and tracking these holes withIncreased birth (birth) and death (death) processes. For each ofWe will get a set of k-dimensional coherent generators (i.e. the loops of the k-chain), and their corresponding birth and death distances.
S2.2.4 constructing a persistent bar code to show the persistence of topological features, wherein each bar represents the generation and extinction of a homonym.
The persistent barcode is a visual representation of the persistent coherent result. For each k-dimensional hole, one "bar" in the persistent barcode corresponds to one interval [ b, d ], where b is the birth distance of the hole and d is the extinction distance of the hole (d= infinity if the hole is not extinction). These intervals (bars) are graphically presented in the form of horizontal line segments, with the start representing birth, the end representing extinction (if any), and the persistent bar code visually demonstrates the stability of different topological features as a function of scale, often used to identify important structures in the data.
In S2.4, the topological feature vector is carried out by utilizing a convolutional neural networkThe specific steps and mathematical formulas involved in the comprehensive diagnosis of the plasma state are as follows:
s2.4.1 input layer of convolutional neural network receives topology feature vector The convolution layer applies a convolution kernel to slide over the input feature vector to generate a feature map:
Wherein, Is the firstThe weights of the convolution kernels; Is the first Offset of the individual convolution kernels; an index that is a convolution kernel; is the length of the convolution kernel;
S2.4.2, nonlinear activation function ReLU, dimension of feature mapping is reduced by downsampling at the pooling layer using maximum pooling:
Wherein, Is the size of the pooling window;
S2.4.3, the full connection layer expands the output of the pooling layer into a one-dimensional vector, is connected to the output layer, and calculates through linear transformation and combination with an activation function:
Wherein, Is the firstA weight vector of the individual neurons; expanding the output vector of the pooling layer into a one-dimensional vector; Is the first Bias of individual neurons;
S2.4.4 and output layer generation predicted plasma parameters :
Wherein, To predict plasma parameters; a weight matrix for an output layer; the output of the full connection layer; is a bias vector for the output layer;
S2.4.5, using the mean square error as a loss function, measuring the error between the predicted value and the true value:
Wherein, As a loss function; is a predicted value; Is a true value; is the number of samples;
s2.4.6, calculating the gradient of the loss function relative to the weight by a back propagation algorithm, and updating the weight:
Wherein, Is the learning rate; Gradient of weight matrix for loss function;
s2.4.7, performing plasma state estimation based on predicted plasma parameters:
Wherein, Is the plasma state evaluation result; Is a state evaluation function.
In said S2.4.7Is the result of the plasma state evaluation, the plasma stateThe method comprises five states:
State one, plasma stability: indicating the stability of the plasma, whether there are fluctuations and instabilities, including stable, slightly unstable and unstable states;
State two, temperature deviation: the deviation between the predicted temperature and the target temperature comprises three states of normal temperature, higher temperature and lower temperature;
State three, pressure deviation: predicting deviation of pressure and target pressure, wherein the deviation comprises three states of normal pressure, high pressure and low pressure;
state four, gas flow deviation: predicting deviation of the gas flow and a set value, wherein the deviation comprises three states of normal flow, higher flow and lower flow;
state five, electron density: reflecting whether the electron density of the plasma is in a set range, including three states of normal electron density, too high electron density and too low electron density.
The plasma control unit 3 is used for adjusting and controlling microwave power to generate and maintain plasma and controlling the flow of methane and hydrogen reaction gas according to the diagnosis result of the plasma diagnosis unit 2;
The plasma control unit 3 comprises a microwave power adjusting module and a gas flow control module;
The microwave power adjusting module adjusts microwave power entering the reaction cavity by using a solid microwave source so as to generate and maintain plasma;
The gas flow control module is used for adjusting the proportion of methane and hydrogen and controlling the carbon concentration and the hydrogen concentration in the plasma.
The device also comprises an environment regulation and control unit 4, wherein the environment regulation and control unit 4 analyzes the pressure data in the reaction cavity, the temperature data of the substrate and the diagnosis result state of the plasma diagnosis unit 2 in the data acquisition unit 1, and regulates the pressure in the reaction cavity and the temperature of the substrate;
the environment regulation and control unit 4 comprises a pressure control module and a temperature control module;
Wherein, the pressure control module acquires real-time pressure data in the reaction cavity from the data acquisition unit 1 And regulating the pressure in the reaction cavity according to the pressure data fed back by the capacitive pressure sensor by using a pressure PID controller, wherein a mathematical model formula related to the pressure PID control is as follows:
Wherein, Is set to a pressure; To measure pressure; As a result of the error in the error, ;、、Respectively proportional, integral and differential coefficients;
Wherein the temperature control module acquires temperature data of the substrate from the data acquisition unit 1 And regulating the temperature of the substrate according to temperature data fed back by an infrared temperature sensor by using a temperature PID controller, wherein a mathematical model formula related to the temperature PID control is as follows:
Wherein, Is a set temperature; to measure temperature; As a result of the error in the error, ;、、Respectively proportional, integral and differential coefficients.
The system also comprises an interaction and communication unit 5, wherein the interaction and communication unit 5 provides an interface for interaction between a user and the control system, displays real-time data and control states, and realizes communication between the units and between the control system and external equipment;
The interaction and communication unit 5 comprises a user interface module and a network communication module;
The user interface module provides a graphical interface and a control panel for a user to view and operate various functions of the system, including real-time data and control states;
The graphical interface displays real-time data from the data acquisition unit 1, including temperature, pressure, gas flow and plasma parameters; displaying the diagnosis result of the plasma diagnosis unit 2, the control result of the plasma control unit 3 and the environment control unit 4;
the control panel is used for manually adjusting microwave power, gas flow, reaction cavity pressure and substrate temperature by a user;
the network communication module is used for realizing network communication between the control system and external equipment; external devices include user computers, cell phones, and other terminals.
The foregoing has shown and described the basic principles, principal features and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the above-described embodiments, and that the above-described embodiments and descriptions are only preferred embodiments of the present invention, and are not intended to limit the invention, and that various changes and modifications may be made therein without departing from the spirit and scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (5)
- High temperature ionization control system in MPCVD diamond cultivation process, its characterized in that: comprising the following steps:The data acquisition unit (1), the said data acquisition unit (1) gathers and stores the high-accuracy sensor data;The plasma diagnosis unit (2), the said plasma diagnosis unit (2) extracts the real-time sensor data in the data acquisition unit (1), use the topological neural network plasma diagnosis technology to calculate and analyze the plasma data, diagnose the plasma state;A plasma control unit (3), wherein the plasma control unit (3) regulates and controls the generation and maintenance of microwave power and controls the flow of methane and hydrogen reaction gas according to the diagnosis result of the plasma diagnosis unit (2);The environment regulation and control unit (4), the pressure data in the reaction cavity, the substrate temperature data and the diagnosis result state of the plasma diagnosis unit (2) in the data acquisition unit (1) are analyzed by the environment regulation and control unit (4), and the pressure in the reaction cavity and the substrate temperature are regulated;the interaction and communication unit (5) is used for providing an interface for interaction between a user and the control system, displaying real-time data and control states and realizing communication between the units and between the control system and external equipment;The topological neural network plasma diagnosis technology is realized based on the fusion of topological data analysis and the deep neural network technology and is used for diagnosing and analyzing the plasma state in the MPCVD diamond cultivation process in real time; the steps involved in the real-time calculation and analysis of the plasma data parameters by the parameter calculation and analysis module using the topological neural network plasma diagnosis technology are as follows:s2.1, constructing an original data vector :Wherein, Is the substrate temperature; the reaction chamber temperature; is the pressure of the reaction chamber; is methane flow; Is the hydrogen flow; Is the plasma electron density; Is the plasma temperature;S2.2, calculating a distance matrix between the original plasma data Constructing Vietoris-Rips complex shape, and calculating topological characteristics under different scales, including connected components, rings and cavities, by using persistent coherent calculation;S2.3, converting the persistent coherent result into numerical characteristics processed by the neural network, and constructing a topological characteristic vector :Wherein, Is a topological feature vector; the numerical characteristics extracted from the persistent coherent including the total number of the persistent coherent, the persistent coherent number of each dimension and the persistent time of each coherent class;S2.4, carrying out topological feature vector by using convolutional neural network Analyzing and calculating to comprehensively diagnose the plasma state;calculating a distance matrix between the original plasma data in S2.2 The Vietoris-Rips complex form is constructed, and the specific steps and mathematical formulas involved in calculating topological features under different scales by using persistent coherence are as follows:S2.2.1 building a distance matrix Calculating a distance matrix between data pointsFor constructing Vietoris-Rips complex:Wherein, Is the firstData points; Is the first Data points; Representing euclidean distance;s2.2.2 based on distance matrix Selecting a scale parameterConstruction Vietoris-Rips ComplexIf (if)Vertex thenAnd a vertexAn edge is arranged between the two edges;s2.2.3 calculating persistent coherence for identifying stable topological features in data, and calculating topological features under different scales by using the persistent coherence, wherein the topological features comprise connected components, rings and cavities:Wherein, Is the firstA dimension coherent group; Is the first Generation time of the homonym; Is the first The death time of the same class;S2.2.4, constructing a durable bar code to display the durability of the topological feature, wherein each bar represents the generation and extinction of a homonym;In S2.4, the topological feature vector is carried out by utilizing a convolutional neural network The specific steps and mathematical formulas involved in the comprehensive diagnosis of the plasma state are as follows:s2.4.1 input layer of convolutional neural network receives topology feature vector The convolution layer applies a convolution kernel to slide over the input feature vector to generate a feature map:Wherein, Is the firstThe weights of the convolution kernels; Is the first Offset of the individual convolution kernels; an index that is a convolution kernel; is the length of the convolution kernel;S2.4.2, nonlinear activation function ReLU, dimension of feature mapping is reduced by downsampling at the pooling layer using maximum pooling:Wherein, Is the size of the pooling window;S2.4.3, the full connection layer expands the output of the pooling layer into a one-dimensional vector, is connected to the output layer, and calculates through linear transformation and combination with an activation function:Wherein, Is the firstA weight vector of the individual neurons; expanding the output vector of the pooling layer into a one-dimensional vector; Is the first Bias of individual neurons;S2.4.4 and output layer generation predicted plasma parameters :Wherein, To predict plasma parameters; a weight matrix for an output layer; the output of the full connection layer; is a bias vector for the output layer;S2.4.5, using the mean square error as a loss function, measuring the error between the predicted value and the true value:Wherein, As a loss function; is a predicted value; Is a true value; is the number of samples;s2.4.6, calculating the gradient of the loss function relative to the weight by a back propagation algorithm, and updating the weight:Wherein, Is the learning rate; Gradient of weight matrix for loss function;s2.4.7, performing plasma state estimation based on predicted plasma parameters:Wherein, Is the plasma state evaluation result; is a state evaluation function;In said S2.4.7 Is the result of the plasma state evaluation, the plasma stateThe method comprises five states:State one, plasma stability: indicating the stability of the plasma, whether there are fluctuations and instabilities, including stable, slightly unstable and unstable states;State two, temperature deviation: the deviation between the predicted temperature and the target temperature comprises three states of normal temperature, higher temperature and lower temperature;State three, pressure deviation: predicting deviation of pressure and target pressure, wherein the deviation comprises three states of normal pressure, high pressure and low pressure;state four, gas flow deviation: predicting deviation of the gas flow and a set value, wherein the deviation comprises three states of normal flow, higher flow and lower flow;state five, electron density: reflecting whether the electron density of the plasma is in a set range, wherein the set range comprises three states of normal electron density, overhigh electron density and overlow electron density;the environment regulation and control unit (4) comprises a pressure control module and a temperature control module;wherein the pressure control module acquires real-time pressure data in the reaction cavity from the data acquisition unit (1) And regulating the pressure in the reaction cavity according to the pressure data fed back by the capacitive pressure sensor by using a pressure PID controller, wherein a mathematical model formula related to the pressure PID control is as follows:Wherein, Is set to a pressure; To measure pressure; As a result of the error in the error, ;、、Respectively proportional, integral and differential coefficients;wherein the temperature control module acquires temperature data of the substrate from the data acquisition unit (1) And regulating the temperature of the substrate according to temperature data fed back by an infrared temperature sensor by using a temperature PID controller, wherein a mathematical model formula related to the temperature PID control is as follows:Wherein, Is a set temperature; to measure temperature; As a result of the error in the error, ;、、Respectively proportional, integral and differential coefficients.
- 2. A high temperature ionization control system during MPCVD diamond growth according to claim 1, wherein: the data acquisition unit (1) comprises a sensor interface module and a data acquisition and storage module;The sensor interface module is connected with and integrated with a high-precision sensor, and is used for monitoring different parameters in the system, wherein the high-precision sensor comprises an infrared temperature sensor, a capacitive pressure sensor, a mass flow controller, a spectrum analyzer and a laser-induced breakdown spectrum;The data acquisition and storage module is used for acquiring high-precision sensor data in real time, storing the acquired sensor data into SSD, and storing and managing the sensor data by using PostgreSQL; wherein the sensor data includes substrate temperature, reaction chamber pressure, methane flow, hydrogen flow, plasma electron density, and plasma temperature.
- 3. A high temperature ionization control system during MPCVD diamond growth according to claim 1, wherein: the plasma diagnosis unit (2) comprises a data processing module, a parameter calculation and analysis module and a reporting and feedback module;The data processing module is used for extracting relevant plasma data in the data acquisition unit (1) and carrying out primary processing of filtering, denoising and calibration on the plasma data; the plasma data comprises substrate temperature, reaction chamber pressure, methane flow, hydrogen flow, plasma electron density and plasma temperature;the parameter calculation and analysis module uses a topological neural network plasma diagnosis technology to realize real-time calculation and analysis of plasma data parameters;The report and feedback module is used for generating a diagnosis report of the parameter calculation and analysis module and feeding back an analysis result to the interaction and communication unit (5) so as to be displayed to a user for real-time check through a display.
- 4. A high temperature ionization control system during MPCVD diamond growth according to claim 1, wherein: the plasma control unit (3) comprises a microwave power adjustment module and a gas flow control module;The microwave power adjusting module adjusts microwave power entering the reaction cavity by using a solid microwave source so as to generate and maintain plasma;The gas flow control module is used for adjusting the proportion of methane and hydrogen and controlling the carbon concentration and the hydrogen concentration in the plasma.
- 5. A high temperature ionization control system during MPCVD diamond growth according to claim 1, wherein: the interaction and communication unit (5) comprises a user interface module and a network communication module;The user interface module provides a graphical interface and a control panel for a user to view and operate various functions of the system, including real-time data and control states;The graphical interface displays real-time data from the data acquisition unit (1), including temperature, pressure, gas flow and plasma parameters; displaying the diagnosis result of the plasma diagnosis unit (2), the control result of the plasma control unit (3) and the environment regulation unit (4);the control panel is used for manually adjusting microwave power, gas flow, reaction cavity pressure and substrate temperature by a user;the network communication module is used for realizing network communication between the control system and external equipment; external devices include user computers, cell phones, and other terminals.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410955835.6A CN118497722B (en) | 2024-07-17 | 2024-07-17 | High-temperature ionization control system in MPCVD diamond cultivation process |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410955835.6A CN118497722B (en) | 2024-07-17 | 2024-07-17 | High-temperature ionization control system in MPCVD diamond cultivation process |
Publications (2)
Publication Number | Publication Date |
---|---|
CN118497722A CN118497722A (en) | 2024-08-16 |
CN118497722B true CN118497722B (en) | 2024-10-15 |
Family
ID=92246813
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202410955835.6A Active CN118497722B (en) | 2024-07-17 | 2024-07-17 | High-temperature ionization control system in MPCVD diamond cultivation process |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN118497722B (en) |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102820323A (en) * | 2012-09-07 | 2012-12-12 | 温州大学 | Nanometer silicon carbide/crystal silicon carbide double graded junction fast recovery diode and preparation method thereof |
CN107218896A (en) * | 2017-07-26 | 2017-09-29 | 大连理工大学 | The method for measuring vacuum ion plating and plasma spray coating plated film thickness and uniformity |
Family Cites Families (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6116779A (en) * | 1997-03-10 | 2000-09-12 | Johnson; Shane R. | Method for determining the temperature of semiconductor substrates from bandgap spectra |
US9157730B2 (en) * | 2012-10-26 | 2015-10-13 | Applied Materials, Inc. | PECVD process |
CN109030976B (en) * | 2018-05-28 | 2020-11-24 | 上海大学 | Power electronic equipment monitoring and diagnosing system and method |
CN109092227B (en) * | 2018-08-30 | 2021-06-29 | 长沙新材料产业研究院有限公司 | MPCVD synthesis equipment and control method |
US11121303B2 (en) * | 2019-08-16 | 2021-09-14 | Microsoft Technology Licensing, Llc | Treatment during fabrication of a quantum computing device to increase channel mobility |
CN112684829A (en) * | 2020-12-22 | 2021-04-20 | 同济大学 | Temperature detection control system and method for MPCVD device |
CN115242334A (en) * | 2022-06-30 | 2022-10-25 | 哈尔滨工业大学 | Satellite on-orbit earth ionized layer plasma environment simulation method and device |
CN115662529A (en) * | 2022-10-08 | 2023-01-31 | 武汉大学 | Discharge plasma reaction path drawing and analyzing method based on node force relationship |
-
2024
- 2024-07-17 CN CN202410955835.6A patent/CN118497722B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102820323A (en) * | 2012-09-07 | 2012-12-12 | 温州大学 | Nanometer silicon carbide/crystal silicon carbide double graded junction fast recovery diode and preparation method thereof |
CN107218896A (en) * | 2017-07-26 | 2017-09-29 | 大连理工大学 | The method for measuring vacuum ion plating and plasma spray coating plated film thickness and uniformity |
Also Published As
Publication number | Publication date |
---|---|
CN118497722A (en) | 2024-08-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11436395B2 (en) | Method for prediction of key performance parameter of an aero-engine transition state acceleration process based on space reconstruction | |
Ye et al. | A novel time-delay multivariate grey model for impact analysis of CO2 emissions from China’s transportation sectors | |
Nagy et al. | Distributional uncertainty analysis using power series and polynomial chaos expansions | |
Lindemann et al. | Anomaly detection and prediction in discrete manufacturing based on cooperative LSTM networks | |
JP2021101388A (en) | Computer system and method for dynamic construction and online deployment of operation-centric first-principles process model for predictive analytics | |
CN109978403A (en) | A kind of quality management-control method, device and the equipment of Product Assembly process | |
JP2010506256A (en) | Univariate methods for monitoring and analyzing multivariate data | |
KR20080098332A (en) | Graphical user interface for presenting multivariate fault contributions | |
US20170249559A1 (en) | Apparatus and method for ensembles of kernel regression models | |
JP6337869B2 (en) | Plant model creation device, plant model creation method, and plant model creation program | |
Ahmad et al. | On median control charting under double sampling scheme | |
JP2011175540A (en) | Apparatus for constructing predictive/diagnostic model | |
CN110442911B (en) | High-dimensional complex system uncertainty analysis method based on statistical machine learning | |
CN113287104A (en) | Data sorting device | |
Li et al. | Application of combined model with DGM (1, 1) and linear regression in grain yield prediction | |
CN117556366B (en) | Data abnormality detection system and method based on data screening | |
CN115130658B (en) | Reservoir landslide displacement prediction method and device based on nonlinear combination model | |
JP2018073316A (en) | Plant diagnostic system and diagnostic method | |
WO2021252815A1 (en) | Activity level measurement using deep learning and machine learning | |
CN103279030B (en) | Dynamic soft measuring modeling method and device based on Bayesian frame | |
CN118497722B (en) | High-temperature ionization control system in MPCVD diamond cultivation process | |
CN112149054B (en) | Construction and application of orthogonal neighborhood preserving embedding model based on time sequence expansion | |
CN117553840A (en) | Instrument based on intelligent management and system thereof | |
Carvajal et al. | Estimation of calibration intervals using Bayesian inference | |
CN109522657B (en) | Gas turbine anomaly detection method based on correlation network and SVDD |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant |