Search

CN-121978960-A - RGA control system based on digital twinning and control method thereof

CN121978960ACN 121978960 ACN121978960 ACN 121978960ACN-121978960-A

Abstract

The invention discloses an RGA control system based on digital twin and a control method thereof, and relates to the technical field of mass spectrometers, wherein the RGA control system comprises a terminal layer, an edge layer and a cloud layer, wherein the terminal layer comprises an enhanced sensor array and a function execution module; the enhanced sensor array is used for collecting working state data, the edge layer is used for sequentially carrying out disturbance prediction and feedforward compensation by utilizing a lightweight digital twin model and a self-adaptive feedforward algorithm based on the working state data and current control parameters to obtain an optimized control instruction, the function execution module is used for obtaining an ion flow signal by utilizing the optimized control instruction, and the cloud layer is used for finely adjusting key parameters of the high-fidelity digital twin model through an online learning algorithm according to the ion flow signal to obtain an RGA control result. The invention realizes the technical upgrade from the traditional passive feedback control to the active feedforward compensation, and simultaneously realizes the breakthrough promotion of RGA instrument performance from multiple dimensions by depending on the continuous self-learning capability of the model and the hardware software compensation strategy.

Inventors

  • SHEN XIANZHONG
  • LI WENFENG
  • LI GANG

Assignees

  • 深圳市海瑞思自动化科技有限公司

Dates

Publication Date
20260505
Application Date
20260309

Claims (10)

  1. 1. The RGA control system based on digital twinning is characterized by comprising a terminal layer, an edge layer and a cloud layer, wherein the terminal layer comprises an enhanced sensor array and a function execution module; the enhanced sensor array is used for collecting working state data; the edge layer is used for sequentially carrying out disturbance prediction and feedforward compensation by utilizing a lightweight digital twin model and a self-adaptive feedforward algorithm based on the working state data and the current control parameters to obtain an optimized control instruction; The function execution module is used for acquiring an ion flow signal by utilizing the optimization control instruction; and the cloud layer is used for fine-tuning key parameters of the high-fidelity digital twin model through an online learning algorithm according to the ion flow signal to obtain an RGA control result.
  2. 2. The digital twinning-based RGA control system of claim 1, wherein the function execution module comprises a reconfigurable radio frequency drive module and an intelligent signal chain system-in-chip, the reconfigurable radio frequency drive module comprises a dual micro control unit chip, a direct digital frequency synthesis ‌ signal generator and a digital power amplifier, and the dual micro control unit chip comprises a real-time control chip and an intelligent computing chip.
  3. 3. The digital twinning-based RGA control system of claim 1, wherein the enhanced sensor array comprises an ion source, quadrupole rods, detectors, distributed high precision temperature sensors, micro-vibration sensors, and internal pressure gradient sensors.
  4. 4. The digital twinning-based RGA control system of claim 1, wherein the edge layer employs a multi-core ARM processor or a heterogeneous system-on-chip.
  5. 5. The digital twinning-based RGA control system of claim 1, wherein the high-fidelity digital twinning model comprises a physical submodel and a degradation and interference submodel.
  6. 6. A control method of a digital twin based RGA control system, applying the digital twin based RGA control system of any of claims 1 to 5, comprising: S1, collecting working state data; s2, based on the working state data and the current control parameters, orderly performing disturbance prediction and feedforward compensation by using a lightweight digital twin model and a self-adaptive feedforward algorithm to obtain an optimized control instruction; s3, acquiring an ion flow signal by utilizing the optimization control instruction; S4, fine-tuning key parameters of the high-fidelity digital twin model through an online learning algorithm according to the ion flow signal to obtain an RGA control result.
  7. 7. The method of claim 6, wherein S2, based on the operating state data and the current control parameters, sequentially performs disturbance prediction and feedforward compensation by using a lightweight digital twin model and an adaptive feedforward algorithm, and obtains an optimized control instruction, and the method comprises: Inputting the working state data and the current control parameters into a lightweight digital twin model for disturbance prediction, and obtaining a predicted spectrogram deviation; performing reverse calculation by utilizing a self-adaptive feedforward algorithm according to the predicted spectrogram deviation to obtain a control parameter fine adjustment quantity; and carrying out feedforward compensation according to the control parameter fine adjustment quantity and the current control parameter to obtain an optimized control instruction.
  8. 8. The method of claim 6, wherein S3, using the optimization control command, obtaining an ion flow signal comprises: Analyzing the optimized control instruction to obtain an analyzed optimized control instruction; Generating an RF digital waveform and a DC digital level according to the analyzed optimization control instruction, and performing digital-to-analog conversion to obtain an analog signal; Performing power amplification processing by using the analog signal to obtain a weak ion current signal; And performing mass scanning and signal acquisition operation based on the weak ion current signal to acquire an ion current signal.
  9. 9. The method of claim 7, wherein S4, fine tuning key parameters of a high-fidelity digital twin model according to the ion flow signal by an online learning algorithm to obtain an RGA control result comprises: Obtaining a predicted spectrogram by utilizing the deviation of the predicted spectrogram; Obtaining residual errors of a real spectrogram and a predicted spectrogram according to the ion flow signal and the predicted spectrogram; Inputting the residual errors of the real spectrogram and the predicted spectrogram into a high-fidelity digital twin model, and performing model parameter fine adjustment through an online learning algorithm to obtain fine-adjusted model parameters; And acquiring RGA control results based on the trimmed model parameters and RGA operation data.
  10. 10. The method of claim 9, wherein the high-fidelity digital twin model comprises a physical sub-model and a degradation and interference sub-model.

Description

RGA control system based on digital twinning and control method thereof Technical Field The invention relates to the technical field of mass spectrometers, in particular to an RGA control system based on digital twinning and a control method thereof. Background The residual gas analyzer (Residual Gas Analyzer, RGA) is used as a core device for gas component analysis in a vacuum environment, gas molecules are converted into charged ions through an ionization source, signals are captured by a detector after being separated by a quadrupole electric field, so that gas component concentration analysis is realized, and the accuracy, stability and reliability of gas detection are directly determined by the performance of the residual gas analyzer. In the prior art, a part of RGA control system adopts a double micro control unit (Microcontroller Unit, MCU) chip architecture, wherein a hardware part comprises a Radio Frequency (RF)/Direct Current (DC) power supply circuit, an ion source and detector power supply circuit, a signal acquisition processing circuit, a vacuum and system monitoring circuit and the like, and is responsible for ion generation, screening, detection and environmental state monitoring, and a control part relies on an embedded processor to execute a feedback control algorithm to complete core functions such as quality scanning, ion flow signal processing, equipment parameter calibration and the like, thus being a basic architecture for guaranteeing basic operation of RGA. Along with the promotion of industrial environment and the diversified expansion of scientific research scenes, the inherent limitation of the traditional RGA control system is increasingly prominent, the high-precision and intelligent application requirements under the complex environment are difficult to meet, firstly, the traditional system adopts passive feedback control logic, after detection deviation occurs due to environmental interference or aging of core devices (filaments, electron multipliers and electrodes), the stability of the system is greatly reduced under non-ideal working conditions and the detection precision of laboratory level is difficult to maintain due to the fact that the control parameters are regulated to carry out post correction, secondly, the hardware parameters of the traditional RGA control system are mostly in fixed configuration, if different application scenes such as a 'high-speed mode' required by vacuum leak detection and a 'ultra-high-resolution mode' required by scientific research analysis are required to be adapted, the hardware modules are required to be manually replaced or complex parameter debugging is required, the operation is complicated, and the adaptation efficiency is low, thirdly, the periodic calibration and the parameter tuning of the system are required to be completed by expert users with professional knowledge, the common users are difficult to independently operate, the aging process of the core devices is lack of effective real-time monitoring and early warning mechanism, the service life is easy to be prolonged, the real-time service life is difficult to be easily prolonged, the real-time performance is difficult to be prolonged, the overall performance of the system is difficult to be cooperated with the overall performance of the overall control system is difficult to be matched with the overall performance and the overall performance of the system is difficult to be matched with the overall performance to be in real-time, and the overall performance is difficult to be matched with the overall performance, and has the overall performance is difficult to be optimized. In the prior art, although some RGA products attempt to improve reliability and compatibility through modularized design, digital control and standardized interfaces, core problems such as 'passive response interference', 'complex operation and maintenance', 'high maintenance cost' and the like are not fundamentally solved. The digital twin technology is used as a key technology for realizing real-time mapping, data synchronization and collaborative optimization of physical entities and virtual models, has strong predictive maintenance, interference prejudgment and active control capability in the fields of intelligent manufacturing, industrial control and the like, but is not effectively integrated into an RGA control system, and cannot prejudge the change of the running state of the physical RGA control system through the virtual models and implement control compensation in advance. Accordingly, there is a need for a digital twin-based RGA control system and control method thereof to solve the deficiencies in the prior art. Disclosure of Invention The invention aims to provide an RGA control system based on digital twinning and a control method thereof, which are used for solving the problems of lack of active interference compensation, low intelligent degree, insufficient stability, low scene adapta