Method, system, and non-transitory machine-readable storage medium for performing radio frequency matching control using a model-based digital twin

Abstract

A method includes generating an RF signal from production equipment to excite a processing chamber associated with the production equipment. The method further includes receiving current tracking data associated with the RF signal from one or more sensors associated with the production equipment. The method further includes updating the impedance value of a digital replica associated with the production equipment based on the current tracking data. The method further includes acquiring one or more outputs representing predicted data from the digital replica. The method further includes causing one or more correction actions associated with the production equipment to be performed based on the predicted data.

Description

Technical Field

[0001] This disclosure relates to electronic components, and more specifically to radio frequency matching control using model-based digital clones. Prior Technology

[0002] Products can be manufactured by performing one or more manufacturing processes using production equipment. For example, semiconductor manufacturing equipment can be used to produce semiconductor components (such as substrates, wafers, etc.) through semiconductor manufacturing processes. In the conventional radio frequency (RF) plasma etching process used in many semiconductor component manufacturing stages, RF signals can be supplied to the substrate processing chamber via RF energy. RF signals can be generated and supplied in continuous wave or pulse wave modes. Due to the impedance mismatch between the RF energy and the plasma formed in the processing chamber, some RF signals are reflected back to the RF energy, resulting in inefficient use of RF signals and wasted energy, potentially damaging the RF energy, and possibly causing inconsistencies / non-repeatability problems related to substrate processing. Therefore, an RF matching control system capable of achieving highly repeatable RF signal transmission is needed. Summary of the Invention

[0003] The following is a brief overview of this disclosure, providing a basic understanding of some aspects of it. This overview is not a comprehensive summary of this disclosure. It is not intended to identify any essential or key elements of this disclosure, nor to describe any category of any particular implementation or claim of this disclosure. Its sole purpose is to present some concepts of this disclosure in a simple form as a prelude to the more detailed description that follows.

[0004] In one embodiment of this disclosure, the method includes generating an RF signal from production equipment to excite a processing chamber associated with the production equipment. The method further includes receiving current tracking data associated with the RF signal from one or more sensors associated with the production equipment. The method further includes updating the impedance value of a digital replica associated with the production equipment based on the current tracking data. The method further includes acquiring one or more outputs representing predicted data from the digital replica. The method further includes performing one or more correction actions associated with the production equipment based on the predicted data.

[0005] In another embodiment of this disclosure, the system includes: a memory; and a processing device coupled to the memory to generate an RF signal from a production device to activate a processing chamber associated with the production device. The processing device further receives current tracking data associated with the RF signal from one or more sensors associated with the production device. The processing device further updates the impedance value of a digital copy associated with the production device based on the current tracking data. The processing device further acquires one or more outputs representing predicted data from the digital copy. The processing device further causes one or more correction actions associated with the production device to be performed based on the predicted data.

[0006] In another embodiment of this disclosure, a non-transitory machine-readable storage medium stores instructions. When executed, the instructions cause a processing device to perform operations, including causing production equipment to generate RF signals to excite a processing chamber associated with the production equipment. The operations further include receiving current tracking data related to the RF signals from one or more sensors associated with the production equipment. The operations further include updating the impedance value of a digital copy associated with the production equipment based on the current tracking data. The operations further include acquiring one or more outputs representing predicted data from the digital copy. The operations further include causing one or more correction actions associated with the production equipment to be performed based on the predicted data. Simple Explanation of the Diagram

[0007] This disclosure is illustrated by examples and is not limited to the figures in the accompanying drawings.

[0008] Figure 1 is a block diagram illustrating a system architecture according to certain embodiments.

[0009] Figure 2 is a block diagram illustrating an exemplary embodiment of the production equipment in more detail according to certain embodiments.

[0010] Figure 3 is a flowchart of generating predictive data to trigger correction actions according to certain embodiments.

[0011] Figures 4A through 4C are illustrative optimized cross-sections according to certain embodiments.

[0012] Figure 5 is a diagram illustrating multiple tuning paths associated with two variable capacitors according to certain embodiments.

[0013] Figure 6 is a block diagram of a computer system according to certain embodiments. Implementation

[0014] This document describes a technique involving RF matching control using model-based digital doppelgangers. Products (e.g., wafers) can be manufactured using production equipment. For example, the production equipment can execute a recipe to produce wafers by supplying RF energy in the form of an RF signal to a processing chamber to generate plasma for etching. Typically, the RF signal is coupled to the plasma in the processing chamber using a fixed or adjustable matching network that minimizes reflected RF signals by more tightly matching the impedance of the plasma to the impedance of the RF energy source. The matching network maximizes the amount of energy coupled to the plasma by effectively coupling the output of the RF source to the plasma. Thus, the matching network makes the total impedance (e.g., plasma impedance, chamber impedance, and matching network impedance) equal to or similar to the output impedance of the RF power delivery.

[0015] In some conventional systems, RF matching control profiles for matching networks are designed based on lookup tables. The lookup table can be associated with processing parameters, such as the type of input frequency used to excite the processing chamber (e.g., 2.2 MHz, 13.56 MHz, etc.), the type of processing gas used (e.g., H₂, He, Ar, O₂, NF₃, etc.), and the type of processing chamber used. One or more variable tuning elements of the matching network, such as variable capacitors, variable inductors, or variable resistors, can be adjusted using the control profile. When the parameters of the matching network or processing chamber change (due to temperature, aging effects, corrosion, component failure, etc.), the control profile selected using the lookup table may not yield optimal results.

[0016] The components, systems, and methods disclosed herein utilize current sensor data to provide real-time optimization of the control profile for maximizing power delivery with minimal tuning time. In one embodiment, a processing device causes a production device to generate an RF signal to excite a processing chamber associated with the production device. The processing device may receive current tracking data (e.g., voltage, current, etc.) associated with the RF signal from one or more sensors associated with the production device. The processing device may then update the impedance value of a digital replica associated with the production device based on the current tracking data. An output representing predicted data may be obtained from the digital replica. The output representing the predicted data may be generated using a trained machine learning model, a heuristic model, or a rule-based model. Based on the predicted data, the processing device may perform one or more corrective actions associated with the production device.

[0017] In some embodiments, the matching network may include two adjustable (tunable) capacitors. Prediction data may include one or more tuning settings for the capacitors. The tuning settings may be a function of the reflection coefficient related to the reflected RF signal and timing parameters. Correction actions may include adjusting one or both of the adjustable capacitors based on the prediction data.

[0018] In some embodiments, the processing device can update the digital copy in real time using current tracking data. For example, in response to a determination based on current tracking data that the digital copy does not meet an accuracy threshold, the processing logic can perform optimization of the digital copy.

[0019] In some embodiments, the matching network may include an input impedance sensor and an output impedance sensor. Updating the impedance value of the digital replica may include updating the processing chamber model associated with the digital replica using current tracking data from the output impedance sensor. Updating the impedance value of the digital replica may further include updating the matching network model associated with the digital replica using current tracking data from both the input impedance sensor and the output impedance sensor.

[0020] The present disclosure offers technical advantages, significantly reducing the time required to achieve optimal settings, energy consumption, bandwidth usage, and processor overhead. For example, conventional systems perform iterative trial-and-error processes to improve the repeatability of delivering optimal RF power. This trial-and-error process requires more time, energy consumption, bandwidth, and processor overhead for each iteration (e.g., generating instructions via trial and error, transmitting instructions, receiving feedback, and generating updated instructions via trial and error). The present disclosure reduces time requirements, energy consumption, bandwidth, and processor overhead by using signal processing, digital replication, and machine learning models to acquire predictive data and perform corrective actions based on this data, thus avoiding iterative trial and error. The present disclosure enables the prediction and matching of optimal parameter settings related to the network, avoiding inconsistent and abnormal products, unplanned user time, and damage to production equipment.

[0021] Figure 1 is a block diagram of an exemplary system 100 (exemplary system architecture) according to certain embodiments. System 100 includes an analog system 110, a client device 120, a sensor 126, a production device 130, a metering device 128, and data storage 140. The analog system 110 may include a digital representation server 170, a server machine 180, and a prediction server 112.

[0022] Production equipment 130 may include an RF generator 132, a matching network 134, a processing chamber 136, and a controller 138. An RF signal may be generated by the RF generator 132 and transmitted to the matching network 134. An RF signal may then be applied to the processing chamber 136 to ignite and maintain plasma for the etching process. In some embodiments, the RF generator 132 may generate one or more low RF signals (e.g., 2.2 MHz, 13.56 MHz, etc.) that excite the processing chamber 136, one or more high RF signals (e.g., 24 MHz, 60 MHz, 100 MHz, etc.), or any combination thereof. The RF generator 132 may be capable of pulsed RF signals at a desired pulse rate, duty cycle, and phase angle. The controller 138 may be connected to the RF power generator 132 and the RF matching network 134, and may control (e.g., start, switch, shut down, etc.) the RF signal of the RF generator 132. Additionally, the pulse rate, duty cycle, and phase angle of the RF signal can be adjusted using the controller 138.

[0023] Matching network 134 minimizes reflected RF energy by matching the impedance of the plasma used in the etching process with the impedance (e.g., 50 ohms) of the RF signal provided by RF generator 132. Matching network 134 may include an arrangement of capacitors, inductors, and resistors. To change its parameters, matching network 134 may include one or more controllable, adjustable, or variable-tuning elements, such as variable capacitors, inductors, or resistors, or combinations thereof. For example, matching network 134 may include a variable parallel capacitor or a variable series capacitor, both of which can be used as variable-tuning elements. Variable (e.g., tuned or tunable) capacitors may be electrically operated vacuum capacitors operated by controller 138, or any other controllable variable capacitor. In embodiments, matching network 134 can tune one or more of the controllable variable-tuning elements such that the impedance associated with matching network 134 and / or processing chamber 136 increases / decreases towards the impedance (e.g., 50 ohms) associated with RF generator 132.

[0024] RF matching network 134 may also include one or more inductors 135, which may be any type of RF voltage / current measurement element (e.g., sensor, probe, etc.) capable of providing sensor data 142 (e.g., sensor values, characteristics, tracking data). In some embodiments, inductors 135 perform electrical measurements on electrical feed conductors (e.g., electrical feed lines) coupled to RF generator 132, matching network 134, and / or processing chamber 136. Inductors 135 sense the properties of the electrical feed conductors (e.g., impedance, magnetic fluctuations in the electrical feed conductors, current, voltage, resistance, etc.) and convert these properties into sensor data 142 (e.g., tracking data, historical tracking data 146, current tracking data 154). Inductors 135 may also measure the value of one or more currents, alternating current (AC) magnitude, phase, waveform (e.g., AC waveform, pulse waveform), direct current (DC), non-sinusoidal AC waveform, voltage, or the like.

[0025] In this embodiment, the inductor 135 includes an input (source) impedance sensor and an output (load) impedance sensor. The input impedance sensor determines the impedance associated with the RF signal at the matching network 134. The output impedance sensor determines the impedance of the RF signal at the destination (e.g., processing chamber 136). The production apparatus 130 will be described in more detail with reference to Figure 2.

[0026] Sensor 126 can provide sensor data 142 (e.g., sensor values, characteristics, tracking data) related to manufacturing equipment 130 (e.g., related to the production of corresponding products such as wafers using manufacturing equipment 130). Sensor data 142 can be used for equipment health and / or product health (e.g., product quality). Manufacturing equipment 130 can produce products by following a recipe or performing operations over a period of time. Sensor data 142 received over a period of time (e.g., corresponding to at least a portion of a recipe or operation) can be referred to as tracking data received over time from different sensors 126 (e.g., historical tracking data 146, current tracking data 154).

[0027] Sensor 126 may include additional sensors that provide other types of sensor data 142. In some embodiments, sensor data 142 may include values ​​for one or more of the following: temperature (e.g., heater temperature), spacing (SP), pressure, high-frequency radio frequency (HFRF), electrostatic chuck (ESC) voltage, current, flow, power, voltage, etc. Sensor data 142 may be associated with or represent manufacturing parameters such as hardware parameters (e.g., settings of manufacturing equipment 130 or components (e.g., size, type, etc.)) or process parameters of the manufacturing equipment. Sensor data 142 may be available when manufacturing equipment 130 performs a manufacturing process (e.g., when the equipment reads data while processing a product). Sensor data 142 may be different for each product (e.g., each wafer).

[0028] In some embodiments, sensor data 142 (e.g., a collection of historical tracking data 146, historical component data 148, current tracking data 154, current component data 156, etc.) may be processed (e.g., via client device 120 and / or via components of simulation system 110). Processing of sensor data 142 may include generating features. In some embodiments, features are patterns (e.g., ramps, widths, heights, peaks, etc.) in sensor data 142 or combinations of sensor values ​​in sensor data 142 (e.g., impedance measured from voltage and current). Sensor data 142 may include features that components of simulation system 110 may use to perform simulation processing and / or acquire simulation data 167 and / or prediction data 168 for correction actions.

[0029] Data storage 140 may be memory (e.g., random access memory), a drive (e.g., a hard drive, flash drive), a database system, or another type of component or element capable of storing data. Data storage 140 may include multiple storage components (e.g., multiple drives or multiple databases), which may encompass multiple computing devices (e.g., multiple server computers). Data storage 140 may store sensor data 142, performance data 160, library data 166, analog data 168, and predictive data 169. Sensor data 142 may include historical sensor data 144 and current sensor data 152. Historical sensor data may include a collection of historical tracking data 146, historical component data 148, and historical component identifiers 150. Current sensor data 152 may include a collection of current tracking data 154, current component data 156, and current component identifiers 158. Performance data 160 may include historical performance data 162 and current performance data 164. Historical sensor data 144 and historical performance data 162 can be historical data. Current sensor data 144 can be current data used to generate simulation data 167 and prediction data 168 (e.g., for performing correction actions). Current performance data 164 can also be current data (e.g., for retraining a post-trained machine learning model).

[0030] Performance data 160 may include information related to manufacturing equipment 130 and / or the products manufactured by manufacturing equipment 130. In some embodiments, performance data 160 may include the lifespan (e.g., downtime) of components of manufacturing equipment 130, production parameters of manufacturing equipment 130, maintenance of manufacturing equipment 130, energy usage of components of manufacturing equipment 130, and indications of differences or similarities among components of manufacturing equipment 130 (e.g., components of the same part number). Performance data 160 may include indications of differences among components of manufacturing equipment (e.g., components of the same type and part number). Performance data 160 may indicate whether differences (e.g., jitter, ramp, peaks, etc.) affect product-to-product variation. Performance data 160 may indicate whether differences provide improved wafers (e.g., better matching of RF generators, better tuning of feedback loops, newer firmware, better chips). Performance data 160 may be related to the quality of products manufactured by manufacturing equipment 130. Metering device 128 can provide performance data 160 (e.g., wafer property data, yield, metering data) related to the product (e.g., wafer) produced by production equipment 130. Performance data 160 may include values ​​for one or more of film property data (e.g., wafer space film properties), dimensions (e.g., thickness, height, etc.), dielectric constant, dopant concentration, density, defects, etc. Performance data 160 may be for finished or semi-finished products. Performance data 160 may be different for each product (e.g., each wafer). Performance data 160 may indicate whether the product meets critical quality requirements (e.g., defective, defect-free, etc.). Performance data 160 may indicate the reasons for not meeting critical quality requirements. In some embodiments, performance data 160 includes historical performance data 162, which corresponds to historical property data of the product (e.g., a product produced using production parameters associated with historical tracking data 146). For supervised and / or unsupervised machine learning, sensor data 142, performance data 160, and library data 166 may be used.

[0031] The simulation system 110 may include a digital representation server 170, a server machine 180, and a prediction server 112. Each of the prediction server 112, the digital representation server 170, and the server machine 180 may include one or more computing devices, such as rack-mount servers, router computers, server computers, personal computers, mainframe computers, laptop computers, tablet computers, desktop computers, graphics processing units (GPUs), and application-specific integrated circuits (ASICs) (e.g., tensor processing units (TPUs)).

[0032] The digital representation server 170 may be an algorithmic model that simulates the production equipment 130. As an example, the digital representation server 170 may be a digital replica model of the production equipment 130 (e.g., a digital clone). The digital representation server 170 may use supervised machine learning, semi-supervised machine learning, unsupervised machine learning, or any combination thereof to generate a virtual representation of the physical components and the dynamics of how the production equipment 130 operates. The digital representation server 170 may be updated via reinforcement learning using periodic updates of sensor 126, inductor 135, sensor data 142, performance data 160, library data 166, and / or any other data related to generating and maintaining the digital replica model of the production equipment 130.

[0033] The digital representation server 170 may include a mating network model 172 and a processing chamber model 174. The mating network model 172 may be related to the dynamics of the physical components and the mating network 134. The processing chamber model 174 may be related to the dynamics of the physical components and the processing chamber 136.

[0034] In some embodiments, the digital representation server 170 may generate analog data 167. Analog data 167 may include data for determining how production equipment 130 (e.g., matching network 134, processing chamber 136, etc.) should operate based on current or analog parameters. Analog data 167 may include electrical parameter data (e.g., impedance, voltage, current, resistance, reflection, signal reflection, etc.) associated with the matching network model 172 and the processing chamber model 174. Analog data 167 may further include predictive properties data of a digital replica model of production equipment 130 (e.g., products to be produced or produced using current tracking data 154). Analog data 167 may further include predictive metering data (e.g., virtual metering data) of products to be produced or produced using current tracking data 154. Analog data 167 may further include indications of anomalies (e.g., defective products, defective components, defective production equipment 130, defective energy usage, etc.) and one or more causes of the anomalies. Analog data 167 may further include indications of the end-of-life of components of production equipment 130. The simulation data can be comprehensive, covering every mechanical or electronic aspect of the production equipment.

[0035] The prediction server 112 may include a prediction component 114. In some embodiments, the prediction component 114 may receive simulation data 167 and current tracking data 154 (e.g., processing chamber flow, processing chamber pressure, RF power, etc.) and generate an output (e.g., prediction data 168) to perform correction actions related to the production equipment 130. In some embodiments, the prediction component 114 may use one or more trained machine learning models 190 to determine the output to perform correction actions based on the simulation data 167 and the current tracking data 154. In some embodiments, the prediction component 114 receives the simulation data 167 and the current tracking data 154, performs signal processing to decompose the current tracking data 154 into a set of current component data 156 mapped to a current component identifier 158, provides the set of current component data 156 and the current component identifier 158 as input to the trained machine learning model 190, and obtains an output representing the prediction data 168 from the trained machine learning model 190. The trained machine learning model 190 may include a single model or multiple models. In some embodiments, the trained machine learning model 190 may use other data from the data storage 140 (e.g., library data 166, performance data 160, sensor data 142, etc.).

[0036] In some embodiments, the simulation system 110 further includes a server machine 180. The server machine 180 may use a dataset generator to generate one or more datasets (e.g., a set of data inputs and a set of target outputs) to train, validate, and / or test machine learning models 190. Specifically, the server machine 180 may include a training engine 182, a validation engine 184, a selection engine 185, and / or a testing engine 186. Engines (e.g., training engine 182, validation engine 184, selection engine 185, and testing engine 186) may refer to hardware (e.g., circuitry, special-purpose logic, programmable logic, microcode, processing device, etc.), software (e.g., instructions running on a processing device, a general-purpose computer system, or a special-purpose machine), firmware, microcode, or a combination thereof. The training engine 182 is capable of training one or more machine learning models 190 using one or more sets of features associated with the training set from the dataset generator 172. Training engine 182 can generate multiple post-trained machine learning models 190, each of which corresponds to a different set of features from the training set (e.g., sensor data from different sets of sensors). For example, a first post-trained machine learning model has been trained using all features (e.g., X1 to X5), a second post-trained machine learning model has been trained using a first subset of features (e.g., X1, X2, X4), and a third post-trained machine learning model has been trained using a second subset of features that partially overlaps with the first subset of features (e.g., X1, X3, X4, and X5).

[0037] The validation engine 184 can validate the trained machine learning model 190 using a corresponding set of features from the validation set generated by the dataset generator. For example, a first trained machine learning model 190 trained using a first set of features from the training set can be validated using a first set of features from the validation set. The validation engine 184 can determine the accuracy of each of the trained machine learning models 190 based on the corresponding set of features from the validation set. The validation engine 184 can discard trained machine learning models 190 whose accuracy does not meet a critical accuracy requirement. In some embodiments, the selection engine 185 can select one or more trained machine learning models 190 whose accuracy meets the critical accuracy requirement. In some embodiments, the selection engine 185 can select the trained machine learning model 190 with the highest accuracy among the trained machine learning models 190.

[0038] The testing engine 186 can test the trained machine learning model 190 using a corresponding set of features from the test set generated by the dataset generator. For example, a first trained machine learning model 190 trained using a first set of features from the training set can be tested using a first set of features from the test set. The testing engine 186 can then determine the trained machine learning model 190 that has the highest accuracy among all trained machine learning models based on the test set.

[0039] Machine learning model 190 may refer to a model artifact built by training engine 182 using a training set that includes data inputs and corresponding target outputs (correct answers to individual training inputs). Patterns mapping data inputs to target outputs (correct answers) can be discovered in the dataset, and mappings for capturing these patterns are provided to machine learning model 190. Machine learning model 190 may use one or more of the following: Support Vector Machine (SVM), Radial Basis Function (RBF), clustering, supervised machine learning, semi-supervised machine learning, unsupervised machine learning, k-Nearest Neighbor algorithm (k-NN), linear regression, random forest, neural networks (e.g., artificial neural networks).

[0040] The prediction component 114 can provide simulated data 167 and current sensor data 152 to the trained machine learning model 190, and can run the trained machine learning model 190 on the input to obtain one or more outputs. The prediction component 114 can determine (e.g., extract) prediction data 168 from the output of the trained machine learning model 190.

[0041] For illustrative and not limiting purposes, this disclosure describes the training of one or more machine learning models 190, and the input of simulated data 167 and sensor data 142 into one or more trained machine learning models 190 to determine prediction data 168. In other embodiments, heuristic or rule-based models are used to determine prediction data 168 (e.g., without using trained machine learning models). Prediction component 114 may monitor historical sensor data 144 and historical performance data 162. Any information described regarding data from data storage 140 may be monitored or otherwise used in heuristic or rule-based models.

[0042] In some embodiments, a smaller number of machines provide the functionality of client device 120, prediction server 112, digital representation server 170, and server machine 180. For example, in some embodiments, digital representation server 170 and server machine 180 may be integrated into a single machine, while in other embodiments, digital representation server 170, server 180, and prediction server 112 may be integrated into a single machine. In some embodiments, client device 120 and prediction server 112 may be integrated into a single machine.

[0043] Generally speaking, the functions described in one embodiment as being performed by the client device 120, prediction server 112, digital representation server 170, and server machine 180 may also be performed on the prediction server 112 as appropriate in other embodiments. Additionally, different or multiple components operating together may perform functions belonging to a specific component. For example, in some embodiments, the prediction server 112 may determine a correction action based on prediction data 168. In another example, the client device 120 may determine the prediction data 168 based on the output from a trained machine learning model.

[0044] Additionally, different or multiple components operating together can perform the functions of a specific component. One or more of the prediction server 112, the digital display server 170, and the server machine 180 can be accessed as a service provided to other systems or components via an appropriate application programming interface (API).

[0045] In this embodiment, a "user" may be referred to as a single individual. However, other embodiments of this disclosure cover an entity controlled by a plurality of users and / or automated sources. For example, a group of individual users combined into a group of managers may be considered a "user".

[0046] The embodiments disclosed herein can be applied to data quality assessment, feature enhancement, model evaluation, virtual metrology (VM), predictive maintenance (PdM), limit optimization, or the like.

[0047] While the embodiments disclosed herein discuss the generation of predictive data 168 for performing calibration actions in a production facility (e.g., a semiconductor manufacturing facility), the embodiments can also be broadly applied to characterization and monitoring components. The embodiments can be broadly used for characterization and monitoring based on different types of data.

[0048] The client device 120, production equipment 130, sensor 126, metering device 128, prediction server 112, data storage 140, digital display server 170 and server machine 180 can be interconnected via network 105 to generate prediction data 168 to perform correction actions.

[0049] In some embodiments, network 105 is a public network that provides client device 120 with access to prediction server 112, data storage 140, and other public computing devices. In some embodiments, network 105 is a private network that provides client device 120 with access to manufacturing equipment 130, sensor 126, metering device 128, data storage 140, and other private computing devices. Network 105 may include one or more wide area networks (WANs), local area networks, wired networks (e.g., Ethernet), wireless networks (e.g., 802.11 networks or Wi-Fi networks), cellular networks (e.g., LTE networks), routers, hubs, switches, server computers, cloud computing networks, and / or combinations thereof.

[0050] Client device 120 includes computing devices such as personal computers (PCs), laptops, mobile phones, smartphones, tablets, laptops, network-connected televisions ("smart TVs"), network-connected media players (e.g., Bluetooth players), set-top boxes, over-the-top (OTT) streaming devices, operating boxes, etc. Client device 120 may include a calibration action component 122. Calibration action component 122 may receive user input (e.g., via a graphical user interface (GUI) displayed through client device 120) related to production equipment 130. In some embodiments, calibration action component 122 transmits instructions to analog system 110, receives output (e.g., prediction data 168) from analog system 110, determines a calibration action based on the output, and performs the calibration action. In some embodiments, the correction action component 122 acquires sensor data 142 (e.g., current tracking data 154) related to the production equipment 130 (e.g., from data storage 140, etc.) and provides the sensor data 142 (e.g., current tracking data 154) related to the production equipment 130 to the simulation system 110. In some embodiments, the correction action component 122 stores the sensor data 142 in the data storage 140, and the prediction server 112 retrieves the sensor data 142 from the data storage 140. In some embodiments, the prediction server 112 may store the output of the trained machine learning model 190 (e.g., prediction data 168) in the data storage 140, and the client device 120 may retrieve the output from the data storage 140. In some embodiments, the correction action component 122 receives instructions for correction actions from the simulation system 110, causing the correction actions to be performed. Each client device 120 may include an operating system that allows a user to perform one or more of the following: generating, observing, or editing data (e.g., instructions related to the production equipment 130, correction actions related to the production equipment 130, etc.).

[0051] Executing a production process that results in defective products can cause losses in terms of time, energy, products, components, production equipment 130, identification of defects, and disposal of defective products. By inputting sensor data 142 (e.g., production parameters that are being used or will be used to produce the products), receiving outputs of predictive data 168, and performing corrective actions based on the predictive data 168, system 100 can have the technical advantage of avoiding the costs of producing, identifying, and disposing of defective products.

[0052] Executing a production process that leads to the failure of components 124 in production equipment can result in losses in terms of downtime, product damage, equipment damage, and the need for quick replacement parts. By inputting sensor data 142 (e.g., production parameters currently used or to be used in product production), receiving output predictive data 168, and performing corrective actions based on the predictive data 168 (e.g., predictive maintenance, such as component replacement, handling, cleaning, etc.), system 100 can offer the following technical advantages: avoiding the costs of one or more of the following: unexpected component failures, unplanned downtime, productivity loss, unexpected equipment failures, product scrap, or the like.

[0053] The corrective action may be related to one or more of Computational Process Control (CPC), Statistical Process Control (SPC) (e.g., SPC for decision control of electronic components, SPC for predicting component lifespan, SPC for comparison with 3 sigma plots, etc.), Advanced Process Control (APC), model-based process control, preventive operation and maintenance, design optimization, production parameter updates, feedback control, machine learning modification, or the like. In some embodiments, the corrective action may include adjusting one or more of the controllable variable tuning elements of the matching network 134. This will be explained in further detail below.

[0054] In some embodiments, corrective actions include providing an alert (e.g., warning to stop or not execute a production process if forecast data 168 indicates a forecast anomaly, such as a product, component, or production equipment 130 malfunction). In some embodiments, corrective actions include providing feedback control (e.g., modifying production parameters in response to forecast data 168 indicating a forecast anomaly). In some embodiments, corrective actions include providing machine learning (e.g., modifying one or more production parameters based on forecast data 168). In some embodiments, performing corrective actions includes triggering updates to one or more production parameters.

[0055] Production parameters may include hardware parameters (e.g., replacement of components, use of certain components, replacement of processing wafers, firmware updates, etc.) and / or process parameters (e.g., temperature, pressure, flow, rate, current, voltage, airflow, lift speed, etc.). In some embodiments, corrective actions include initiating preventative operational maintenance (e.g., replacing, processing, cleaning, etc., components of production equipment 130). In some embodiments, corrective actions include initiating design optimization (e.g., updating production parameters, production processes, production equipment 130, etc., to obtain an optimized product). In some embodiments, corrective actions include updating the formulation (e.g., production equipment 130 is in idle mode, hibernation mode, warm-up mode, etc.).

[0056] Figure 2 is a block diagram illustrating an exemplary embodiment of the production equipment 130 in more detail. For illustrative purposes, the components in Figure 2 are used as examples. Therefore, embodiments of this disclosure may use different combinations of components such as insulators, capacitors, resistors, etc. As discussed above, the production equipment 130 includes an RF generator 132, a matching network 134, a processing chamber 136, and a controller 138.

[0057] Matching network 126 may include a first capacitor 226 (also referred to as "C1") and a second capacitor 228 (also referred to as "C2"). Each of the first capacitor 226 and the second capacitor 228 may be a variable capacitor (tunable or adjustable capacitor) capable of tuning the total impedance of matching network 134. For example, embodiments of this disclosure may perform matching tuning by adjusting either or both of the first capacitor 226 and the second capacitor 228, such that the output impedance associated with processing chamber 136 is adjusted to the input impedance associated with RF generator 132 (e.g., 50 ohms). This maximizes the power delivery from RF generator 132 to processing chamber 136. In some embodiments, the first capacitor 226 may further be a parallel capacitor, while the second capacitor 228 may be a series capacitor. Series capacitors can be used in transmission lines for series compensation to improve power handling capabilities. Parallel capacitors can be applied to electrical systems to perform multiple tasks in a single application. Furthermore, in series capacitors, the generation of unused power is proportional to the square of the load current. In a parallel capacitor, the generation of ineffective power is proportional to the square of the voltage.

[0058] Matching network 134 may include input sensor 222 and output sensor 224. Input sensor 222 determines the impedance associated with matching network 134. Output sensor 224 determines the impedance of the RF signal processed at chamber 136.

[0059] Matching network 126 can be associated with a first impedance 212 (also referred to as "Z1") and a second impedance 214 (also referred to as "Z2"). The first impedance 212 can be associated with a first capacitor 226, a resistor 234 (also referred to as "R1"), and an inductor 232 (also referred to as "L1") in parallel paths within matching network 134. The first impedance 212 can be represented by the following Equation 1, where term j represents the imaginary unit, and term w represents the angular frequency and is expressed as 2*π*f, where f is the frequency of the RF signal (e.g., 2.2 MHz, 13.56 MHz, 100 MHz, etc.): Equation 1

[0060] The second impedance 214 may be associated with the second capacitor 228. In some embodiments, the second impedance 214 may also be associated with one or more passive elements (e.g., inductors, resistors, etc.). The second impedance 214 can be expressed by the following Equation 2: Equation 2

[0061] The processing chamber 136 can be associated with a third impedance 216 (also referred to as "Z3"), which is related to the impedance of the processing chamber components, the transfer line impedance, the feed line impedance (from the matching network 134 to the processing chamber 136), etc. As an example, Z3 can be represented by the real part Rp 238 and the imaginary part Lp 236 of the impedance associated with the processing chamber 136. Rp and Lp can be measured in real time using an input impedance sensor 224. The third impedance 216 can be represented by the following Equation 3: Equation 3

[0062] Therefore, the total load impedance of production equipment 130 can be expressed by the following equation 4: Equation 4

[0063] Figure 3 is a flowchart of a method 300 for generating predictive data to trigger a correction action according to certain embodiments. Method 300 can be executed using processing logic including hardware (e.g., circuitry, dedicated logic, programmable logic, microcode, processing device, etc.), software (e.g., instructions running on a processing device, a general-purpose computer system, or a special-purpose machine), firmware, microcode, or combinations thereof. In some embodiments, method 300 can be partially executed by an analog system 110. In some embodiments, non-transitory storage media stores instructions that, when executed by a processing device (e.g., analog system 110, server machine 180, prediction server 112, etc.), cause the processing device to execute method 300.

[0064] For the sake of brevity, method 300 is depicted and described as a series of operations. However, the operations according to this disclosure may occur in various orders and / or simultaneously, and may occur together with other operations not presented and described herein. Furthermore, not all the illustrated operations are performed to implement method 300 according to the disclosed object. Additionally, those skilled in the art will understand and appreciate that method 300 may alternatively be represented as a series of interrelated states using state diagrams or events.

[0065] In block 310, the processing logic of implementation method 300 causes RF generator 132 to generate RF signals to excite the processing chamber.

[0066] In block 312, the processing logic updates the impedance value in the digital replica. Specifically, the processing logic may update the impedance associated with the matching network model 172 based on the current tracking data generated from the input sensor 222, and update the impedance associated with the processing chamber model 174 based on the current tracking data 154 generated from the output sensor 224.

[0067] In block 314, the processing logic can determine whether the digital replica meets the accuracy threshold based on the current impedance value. For example, the processing logic can use a cost function formula (e.g., mean squared difference) to determine whether the matching network model 172 and the processing chamber model 174 are within the tolerance limits.

[0068] As an example, the processing logic can be represented by Equation 5 below, where term J1 represents the tolerance limit, term Zinput represents the impedance determined by input sensor 222, and term angle represents the frequency and phase of the RF signal: Equation 5

[0069] It should be understood that the closer term J1 is to 0, the more accurate the physical component parameters of the matching network model 172 and the dynamics of the matching network 134. Terms related to Equation 5 may be restricted. In some embodiments, the values ​​of the first capacitor 226 and the second capacitor 228 may have upper limits (C1 max and C2 max) and lower limits (C1 min and C2 min) as indicated below: C1 min < C1 < C1 max, C2 min < C2 < C2 max. In response to the digital copy meeting the critical criteria (e.g., both the matching network model 172 and / or the processing chamber model 174 are within the tolerance limits), the processing logic proceeds to block 318. In response to the digital copy not meeting the critical criteria, the processing logic proceeds to block 316.

[0070] In block 316, the processing logic optimizes the digital copy. In an embodiment, the processing logic can use Equation 5 to determine the optimal values ​​of the resistive element R1 and / or the inductor element L1 used in the matching network model 172. For example, the processing logic can update the matching network model 172 by adjusting one or more parameters (e.g., L1, R1, C1, C2, etc.) until term J1 meets the accuracy threshold. In an embodiment, the processing logic can use an update algorithm to update the processing chamber model 174 by adjusting one or more parameters (e.g., Lp, Rp, etc.). The adjusted parameters can be stored as simulation data 167.

[0071] In block 318, the processing logic acquires one or more outputs representing prediction data from the simulation system 110. For example, the processing logic may determine the optimal values ​​of the first capacitor 226 and the second capacitor 228 to minimize reflected RF signals. In some embodiments, the processing logic may use prediction unit 114 to perform a closed-loop simulation using simulation data 167 and model 190. Prediction unit 114 may generate one or more optimization profiles that indicate the values ​​for adjusting the first capacitor 226 and the second capacitor 228 to minimize reflected RF signals. In some embodiments, prediction unit 114 may further use a tuning time parameter to minimize the time required to reach the optimal value.

[0072] As an example, the processing logic can use Equation 6, represented as follows, to generate the optimized profile, where: Equation 6

[0073] Term J2 represents the cost function, which can be set as the total impedance of production equipment 130 (e.g., Z total), and term w1 represents the weighting coefficient related to the reflection coefficient. The term T_tune represents the tuning time, the term w_2 represents a weighting coefficient related to the tuning time, and the term Z_0 represents the target impedance (e.g., the impedance of the RF generator 132, e.g., 50 ohms). In some embodiments, Equation 6 has the following limitations: (C1 speed limit) Limitation 1 (C2 speed limit) Restriction 2 Let T tune be defined as γ < 0.05, where Restriction 3

[0074] The prediction server can use Equation 6 to generate one or more optimized profiles for the first capacitor 226 and the second capacitor 228. The optimized profiles can indicate the reflection coefficient value for each tuning combination of the first capacitor 226 and the second capacitor 228. The optimized profiles can further indicate the tuning path to reach the target capacitor value. The target capacitor value can be a setting parameter for the first capacitor 226 and the second capacitor 228 having the lowest achievable reflection coefficient.

[0075] Figures 4A through 4C are illustrative optimized profiles according to certain embodiments. Specifically, Figure 4A represents the reflection coefficient (γ) values ​​for each initial tuning value (e.g., 0 to 100%) of the first capacitor 226 and the second capacitor 228. Point 410 represents the lowest achievable reflection coefficient of the optimized profile, which is associated with the first capacitor 226 tuned to approximately 86% and the second capacitor tuned to approximately 30% (target capacitor value).

[0076] Figure 4B illustrates the duration required to reach the target capacitor values ​​for each set of initial values ​​for the first capacitor 226 and the second capacitor 228. For example, if the first capacitor 226 has a 50% tuning value and the second capacitor 228 has a 0% tuning value (e.g., point 430), then the calibration action unit 122 needs 1.25 seconds to tune the first capacitor 226 and the second capacitor 228 to the target capacitor values ​​(e.g., 86% and 30%, respectively).

[0077] Figure 4C illustrates a gradient-based matched tuning optimization profile. The gradient-based matched tuning optimization profile can be defined using the following cost function: Equation 7

[0078] The control parameters of the optimization model can be determined by taking the partial derivatives of the cost function in Equation 7. For example, the following equation represents the partial derivatives of the cost function: Equation 8 Equation 9

[0079] Partial derivatives yield the gradient control values ​​for a given gradient tuning. For example, the gradient control law can produce the following gradient control: Equation 10 Equation 11

[0080] The positive parameters k1, k2, g1, and g2 are the controller gains. The values ​​of c1 and c2 are the critical values ​​for the transition from gradient control to linear control. For example, the transition between gradient control and linear control can be used to remove or otherwise mitigate chatter or noise in the steady state of the optimized profile.

[0081] In some embodiments, the optimized profile can be based on an RF matching tuning algorithm based on a subcommensurable determinant. In the context of controlling RF matching, the subcommensurable determinant can predict capacitance changes (e.g., the first capacitor C1 226 and the second capacitor C2 228 in Figure 2) caused by past and future iterations of the optimized profile. As previously noted, a matching network (e.g., matching network 134 in Figure 1) can be run to minimize reflected RF energy by matching the impedance of the plasma used in the etching process with the impedance of the RF signal. For example, the desired real-valued impedance (e.g., 50 ohms, 75 ohms, etc.) can be achieved by plasma matching. Using Equation 4, the input impedance can be expressed as follows: Equation 12

[0082] According to Equation 12, the subcomparable matrix can be calculated as follows. Equation 13

[0083] In some embodiments, a subcomparable controller may be defined as follows: Equation 14

[0084] Where C(t) = [C1 C2], e(t) = [50 - abs(Z); -angle(Z)], and K is a positive definite gain matrix (e.g., Equation 13). In other embodiments, the subcommensurable determinant Jc can be modified to adjust the loop gain of the amplitude-phase control system.

[0085] In other embodiments, a damping constant λ can be used to prevent singularities from occurring in the inverse calculation of the subcomparable matrix Jc, further resulting in the determination of the inverse subcomparable determinant being driven by processing logic. In another embodiment, a non-direct matrix solver can be used to compute and update tuning values ​​without directly solving the inverse subcomparable matrix. For example, methods such as matrix least squares solutions, iterative solvers, and matrix decomposition solvers (e.g., upper and lower triangular solvers, singular value decomposition, and the like) can be used, which will be further discussed in other embodiments.

[0086] In some embodiments, distributed control of the optimized profile can be used via Singular Value Decomposition (SVD). For example, the SVD of a subcommensurable matrix can be represented as follows: Equation 15

[0087] U and V are orthogonal matrices, and D is a diagonal matrix. For example, D contains only non-zero values ​​along its diagonal elements (e.g., along the diagonal 𝜎 𝑖=𝑑 i,j). In the absence of singularities, the diagonal elements decrease along the columns and rows of the matrix (e.g., 𝜎1≧𝜎2…>0). Because D is set as a diagonal matrix, the control of the optimization model is decentralized. For example, the amplitude control loop and phase control loop of the impedance can be independently tuned by adjusting one or more gains of the diagonal matrix D.

[0088] In some embodiments, the RF matching circuit optimization model may not meet critical accuracy. For example, the optimization model described herein (e.g., Figures 4A through 4C) may depend on values ​​(e.g., theoretical or generalized values) that incorrectly represent the unique characteristics of a particular RF matching circuit. In some embodiments, the following calibrations can be used to measure parameters and values ​​more accurately to determine and predict a more accurate optimization profile.

[0089] The processing logic can follow a calibration procedure to update optimization parameters (e.g., improve the accuracy and precision of the optimization profile). The processing logic can use RF matching auto-tuning mode to find the tuning positions of the first capacitor (e.g., parallel capacitor C1) and the second capacitor (e.g., series capacitor C2). The processing logic can use perturbations of the tuning impedance positions to calculate the subcomparable determinant of the measurement. For example, the processing logic can move C1 / C2 to the matching position (e.g., by setting RF matching to manual control). While keeping C2 constant, the processing logic can calculate the impedance of C1δ (e.g., δc1 = 0.5). In this example, within a short time (e.g., 10 seconds), the processing logic moves C1 to C1 +δc1 and reads the input impedance R1. The processing logic can also move C1 to C1 –δc1 within a short time (e.g., 10 seconds) and read the matched input impedance R2. While keeping C1 constant, the processing logic can calculate the impedance of C2δ (e.g., δc2 = 0.5). In this example, within a short time (e.g., 10 seconds), the processing logic moves C2 to C2+δc2 and reads the input impedance R3. The processing logic can then move C2 to C2–δc2 within a short time (e.g., 10 seconds) and read the matched input impedance R4. The processing logic can then use the following equations and parameter values ​​described in this calibration example to calculate the subcommensurable determinant. Equations 16 to 20

[0090] As discussed in this paper, the subcomparable determinant identified by Equation 20 can then be used in conjunction with other optimized profiles (such as SVD, damping, etc.).

[0091] In some embodiments, the processing logic may continue to generate predictive data (e.g., optimized profiles) based on weighted coefficients until the data meets critical criteria. Specifically, the weighted coefficients may be set to 1 (e.g., w1 + w2 = 1 or 100%). Users may modify the weighted coefficients based on a preference for minimizing reflection power relative to tuning time. For example, if a user prioritizes minimizing the matched tuning time, the user may set w2 to a high value. If a user prioritizes minimizing reflection power, the user may set w1 to a high value.

[0092] At block 320, based on the predicted data, the processing logic causes a correction action to be performed. For example, the processing logic may select a tuning path from an optimized profile. The controller 138 may adjust the first capacitor 226 and / or the second capacitor 228 based on the tuning path. Figure 5 is a diagram illustrating multiple tuning paths associated with the first capacitor 226 and / or the second capacitor 228. Specifically, Figure 5 shows tuning paths to target capacitor values ​​for each initial tuning value (e.g., 0 to 100%) of the first capacitor 226 and the second capacitor.

[0093] Figure 6 is a block diagram illustrating a computer system 600 according to certain embodiments. In some embodiments, the computing system 600 may be connected (e.g., via a network, such as a local area network (LAN), intranet, extranet, or internet) to other computer systems. The computer system 600 may operate in a client-server environment at the capacity of a server or client computer, or as a peer point computer in a peer-to-peer or distributed network environment. The computer system 600 may be provided by a personal computer (PC), tablet PC, set-top box (STB), personal digital assistant (PDA), cellular phone, web device, server, network router, switch, or bridge, or any other device capable of executing a set of instructions (sequentially or otherwise) specifying actions to be performed by the device. Furthermore, the term "computer" should include any connected computer that individually or collectively executes a set (or more) of instructions to perform any one or more of the methods disclosed herein.

[0094] In another embodiment, the computer system 600 may include a processing device 602, volatile memory 604 (e.g., random access memory (RAM)), non-volatile memory 606 (e.g., read-only memory (ROM) or electrically erasable programmable ROM (EEPROM)), and data storage device 616, which can communicate with each other via bus 608.

[0095] The processing device 602 may be provided by one or more processors, such as general-purpose processors (e.g., Complex Instruction Set Computing (CISC) microprocessors, Reduced Instruction Set Computing (RISC) microprocessors, Very Long Instruction Word (VLIM) microprocessors, microprocessors implementing other types of instruction sets, or microprocessors implementing combined types of instruction sets) or special-purpose processors (e.g., Application Specific Integrated Circuit (ASIC), Field Programmable Gate Array (FPGA), Digital Signal Processor (DSP), or Network Processor).

[0096] The computer system 600 may further include a network interface device 622 (e.g., coupled to a network 674). The computer system 600 may also include a video display unit 610 (e.g., an LCD), an alphanumeric input device 612 (e.g., a keyboard), a cursor control device 614 (e.g., a mouse), and a signal generating device 620.

[0097] In some embodiments, the data storage device 616 may include a non-transitory computer-readable storage medium 624 that stores instructions 626 encoding any or more of the methods or functions described herein, including encoding components of Figure 1 (e.g., correction action component 122, prediction component 114, etc.) and instructions for implementing the methods described herein.

[0098] Instruction 626 may also reside wholly or partially within volatile memory 604 and / or processing device 602 during execution by computer system 600; therefore, in some embodiments, volatile memory 604 and processing device 602 also constitute machine-readable storage media.

[0099] Although computer-readable storage medium 624 is shown as a single medium in the illustrative examples, the term "computer-readable storage medium" should include single or multiple media (e.g., centralized or distributed databases and / or associated caches and servers) that store one or more sets of executable instructions. The term "computer-readable storage medium" should also include tangible media capable of storing or encoding a set of instructions that, when executed by a computer, cause the computer to perform any or more of the methods described herein. The term "computer-readable storage medium" should include, but is not limited to, solid-state memory, optical media, and magnetic media.

[0100] The methods, elements, and features described herein may be implemented by separate hardware components or integrated into the functionality of other hardware components such as ASICs, FPGAs, DSPs, or similar devices. Additionally, in some embodiments, the methods, elements, and features may be implemented by functional circuitry within a firmware module or hardware device. Furthermore, the methods, elements, and features may be implemented by any combination of hardware components and computer program elements, or by a computer program.

[0101] Unless otherwise specified, terms such as “receive,” “execute,” “provide,” “acquire,” “trigger,” “access,” “determine,” “add,” “use,” “train,” or similar terms refer to actions or processes performed or implemented by computer systems that manipulate data represented as physical (electronic) quantities in computer system registers and memory, and convert them into data similarly represented as physical quantities in computer system memory or registers or other such information storage, transmission, or display devices. Furthermore, the terms “first,” “second,” “third,” “fourth,” etc., used herein are used as labels to distinguish different elements and do not have an sequential meaning associated with their numerical symbols.

[0102] The examples described herein also relate to apparatuses used to perform the methods described herein. Such apparatuses are specifically constructed to perform the methods described herein, or the apparatus includes a general-purpose computer system selectively programmed by computer programs stored within the computer system. Such computer programs may be stored in a computer-readable tangible storage medium.

[0103] The methods and illustrative examples described herein are not fixedly related to any particular computer or other device. It is convenient to use various general-purpose systems, or to construct more specialized devices, to perform the methods and / or their individual functions, routines, subroutines, or operations described herein, based on the teachings described herein. The above description illustrates examples of structures used in various such systems.

[0104] The above description is intended to be illustrative and not restrictive. Although this disclosure has been described with reference to specific illustrative examples and implementations, it should be understood that this disclosure is not limited to the described examples and implementations. The scope of this disclosure should be determined with reference to the following claims and the full scope of their authorized equivalents.

[0105] 100: System 105: Internet 110: Simulation System 112: Prediction Server 114: Predicted Components 120: Client device 122: Correction of moving parts 126: Matching network 128: Measuring equipment 130: Production equipment 132: RF Generator 134: Matching network 135: Inductor 136: Processing Chamber 138: Controller 140: Data Storage 142: Sensor Data 144: Historical Sensor Data 146: Historical Tracking Data 148: Historical Component Data 150: Historical Component Identifier 152: Current sensor data 154: Current Tracking Data 156: Current Component Information 158: Current component identifier 160: Performance Data 162: Historical Performance Data 164: Current Performance Data 166: Programming Library Materials 167: Simulation Data 168: Forecast Data 170: Digital Display Server 172: Matching Network Model 174: Processing Chamber Model 180: Server Machine 182: Training Engine 184: Verification Engine 185: Select Engine 186: Test Engine 190: Machine Learning Model 212: First impedance 214: Second impedance 216: Third impedance 222: Input Sensor 224: Input Impedance Sensor 226: First capacitor 228: Second capacitor 232: Inductor 234: Resistive element 236: Imaginary part Lp 238: Real Part Rp 300: Method 310: Square 312: Square 314: Square 316: Square 318: Square 320: Square 410: point 420: points 430: points 600: Computer System 602: Processing device 604: Volatile Memory 606: Non-volatile memory 608: Busbar 610: Video display unit 612: Alphanumeric Input Device 614: Vernier control device 618: Data storage device 620: Signal generating device 622: Network Interface Device 624: Computer-readable storage media 626: Instruction 674: Network C1: First capacitor C2: Second capacitor LP: Imaginary part L1: Inductor RP: Real Part R1: Resistor element

[0106] Domestic storage information (please note in order of storage institution, date, and number) none Overseas storage information (please note in the order of storage country, institution, date, and number) none

Claims

1. A method for radio frequency matching control using a model-based digital replica, the method comprising the steps of: generating an RF signal from a production device to excite a processing chamber associated with the production device; receiving current tracking data associated with the RF signal from one or more sensors associated with the production device; updating the impedance value of a digital replica associated with the production device based on the current tracking data, wherein the digital replica includes a model reflecting a virtual representation of how physical components and the production device operate; acquiring one or more outputs representing predicted data from the digital replica; and performing one or more correction actions associated with the production device based on the predicted data.

2. The method as described in claim 1 further comprises the following steps: in response to a determination based on the current tracking data that the digital copy does not meet an accuracy threshold, performing an optimization of the digital copy.

3. The method as described in claim 1, wherein updating the impedance values ​​of the digital replica comprises: updating a processing chamber model associated with the digital replica using tracking data from an output impedance sensor, wherein the output impedance sensor is associated with a matching network of the production equipment.

4. The method as described in claim 1, wherein the step of updating the impedance values ​​of the digital replica comprises the following steps: updating a matching network model associated with the digital replica using tracking data from an input impedance sensor and an output impedance sensor, wherein the input impedance sensor and the output impedance sensor are associated with a matching network of the production equipment.

5. The method as described in claim 1, wherein a trained machine learning model is used to generate one or more outputs indicating the predicted data.

6. The method as described in claim 1, wherein the prediction data includes one or more tuning settings of one or more variable capacitors associated with a matching network of the production equipment.

7. The method as described in claim 1, wherein the correction action includes adjusting one or more variable capacitors associated with a matching network of the production equipment based on the prediction data.

8. A system for radio frequency matching control using a model-based digital replica, the system comprising: a memory; and a processing device coupled to the memory, the processing device: causing a production device to generate an RF signal to excite a processing chamber associated with the production device; receiving current tracking data associated with the RF signal from one or more sensors associated with the production device; updating the impedance value of a digital replica associated with the production device based on the current tracking data, wherein the digital replica includes a model reflecting a virtual representation of how physical components and the production device operate; acquiring one or more outputs representing predicted data from the digital replica; and performing one or more correction actions associated with the production device based on the predicted data.

9. The system as described in claim 8, wherein the processing device further: in response to a determination based on the current tracking data that the digital copy does not meet an accuracy threshold, performs an optimization of the digital copy.

10. The system as claimed in claim 8, wherein, in order to update the impedance values ​​of the digital replica, the processing apparatus further: updates a processing chamber model associated with the digital replica using tracking data from an output impedance sensor, wherein the output impedance sensor is associated with a matching network of the production equipment.

11. The system as claimed in claim 8, wherein, in order to update the impedance values ​​of the digital replica, the processing apparatus further: updates a matching network model associated with the digital replica using tracking data from an input impedance sensor and an output impedance sensor, wherein the input impedance sensor and the output impedance sensor are associated with a matching network of the production equipment.

12. The system as described in claim 8, wherein a trained machine learning model is used to generate one or more outputs indicating predicted data.

13. The system as described in claim 8, wherein the predictive data includes one or more tuning settings of one or more variable capacitors associated with a matching network of the production equipment.

14. The system as described in claim 8, wherein the correction action includes adjusting one or more variable capacitors associated with a matching network of the production equipment based on the predicted data.

15. A non-transitory machine-readable storage medium storing instructions that, when executed, cause a processing device to perform operations, the operations comprising: causing a production apparatus to generate an RF signal to excite a processing chamber associated with the production apparatus; receiving current tracking data associated with the RF signal from one or more sensors associated with the production apparatus; updating the impedance value of a digital copy associated with the production apparatus based on the current tracking data, wherein the digital copy includes a model reflecting a virtual representation of how physical components and the production apparatus operate; acquiring one or more outputs representing predicted data from the digital copy; and performing one or more correction actions associated with the production apparatus based on the predicted data.

16. The non-transitory machine-readable storage medium as described in claim 15, wherein such operation further comprises: in response to a determination based on the current tracking data that the digital copy does not meet an accuracy threshold, performing an optimization of the digital copy.

17. The non-transitory machine-readable storage medium as described in claim 15, wherein updating the impedance values ​​of the digital copy comprises: updating a processing chamber model associated with the digital copy using tracking data from an output impedance sensor, wherein the output impedance sensor is associated with a matching network of the production equipment.

18. The non-transitory machine-readable storage medium as described in claim 15, wherein updating the impedance values ​​of the digital copy comprises: updating a matching network model associated with the digital copy using tracking data from an input impedance sensor and an output impedance sensor, wherein the input impedance sensor and the output impedance sensor are associated with a matching network of the production equipment.

19. The non-transitory machine-readable storage medium as described in claim 15, wherein one or more outputs indicative of predicted data are generated using a trained machine learning model.

20. The non-transitory machine-readable storage medium as described in claim 15, wherein the predictive data includes one or more tuning settings of one or more variable capacitors associated with a matching network of the production equipment, wherein the correction action includes adjusting one or more variable capacitors associated with a matching network of the production equipment based on the predictive data.

Images