A mass flow controller adaptive temperature compensation method and system suitable for wide temperature range

Abstract

The application discloses a mass flow controller adaptive temperature compensation method and system suitable for a wide temperature range, and belongs to the technical field of systems for controlling or regulating non-electric variables corresponding to G05D. The method constructs a segmented temperature flow error database, determines compensation coefficients by using a machine learning algorithm, and performs primary correction on original signals of a flow sensor. Meanwhile, secondary correction is performed on the corrected signals based on a physical model in combination with viscosity-temperature characteristic data corresponding to fluid types. In addition, when the temperature gradient between key components in the controller is out of limit, a feedforward compensation instruction is generated and superimposed on a valve driving signal to suppress thermal interference. The application solves the problem of control precision decline caused by temperature nonlinear drift, fluid property change and internal thermal gradient in a wide temperature range, and significantly improves the reliability and universality of the mass flow controller in extreme environments.

Description

Technical Field

[0001] This invention relates to the field of fluid measurement and control technology, and in particular to an adaptive temperature compensation method and system for a mass flow controller suitable for a wide temperature range. Background Technology

[0002] As a core device in G05D (systems controlling or regulating non-electrical variables) for precisely controlling the critical non-electrical variable of fluid mass flow rate, the performance of mass flow controllers directly determines the stability and repeatability of high-precision processes in semiconductor manufacturing, biomedicine, and other industries. Simultaneously, its internal closed-loop control logic is also a typical application of G05B (general control or regulation systems). In actual operating conditions, changes in ambient and fluid temperature are the main sources of interference affecting the control accuracy of G05D systems. Currently, mainstream mass flow controllers generally employ temperature compensation strategies based on fixed lookup tables or simple linear models. While these methods can maintain a certain level of accuracy near standard temperature points, in wide temperature range applications from -40℃ to +150℃, the failure to accurately characterize the complex nonlinear relationship between temperature and flow rate errors leads to input signal distortion in the G05B control loop. This causes the control valve's adjustment action to deviate from the actual requirements, severely weakening the overall control accuracy and reliability of the system.

[0003] Furthermore, existing technologies typically treat fluid properties (such as viscosity) as constants or make only coarse corrections, ignoring the physical nature of the dynamic viscosity changes with temperature in different gas media. This introduces systematic biases when converting the volumetric flow rate sensed by the sensor into mass flow rate, failing to meet the requirements for accurate measurement of non-electrical variables across multiple gases and wide temperature ranges. More critically, the flow sensor chip, control valve body, and circuit board within the mass flow controller exhibit significant internal temperature gradients due to differences in material heat capacity and thermal conductivity. This structural thermal interference disrupts the thermal equilibrium of the sensing unit and causes minute changes in the valve body's mechanical dimensions, thereby simultaneously affecting the sensing accuracy of the control loop and the actuator response characteristics, leading to flow overshoot, oscillation, or even loss of control. Existing technologies lack effective sensing and proactive compensation mechanisms for such internal thermal field inhomogeneities, making it difficult to guarantee control stability under complex operating conditions. Summary of the Invention

[0004] The purpose of this invention is to design an adaptive temperature compensation method and system for mass flow controllers applicable to a wide temperature range, aiming to solve the problems of control accuracy decay, poor fluid adaptability and internal thermal interference caused by temperature changes in the prior art, thereby improving the reliability and versatility of mass flow controllers in extreme temperature environments.

[0005] To achieve the above objectives, a first aspect of the present invention provides an adaptive temperature compensation method for a mass flow controller applicable to a wide temperature range, the method comprising: Obtain the ambient temperature of the environment where the mass flow controller is located and the temperature of the fluid flowing through the mass flow controller; Based on a pre-defined segmented temperature-flow error database, a machine learning algorithm is used to determine the target compensation coefficient corresponding to the ambient temperature and fluid temperature. Based on the target compensation coefficient, the original output signal of the flow sensor is corrected to obtain the corrected flow signal; Obtain viscosity-temperature characteristic data corresponding to the fluid type, and query the current fluid viscosity based on the fluid temperature; Using the current fluid viscosity, the corrected flow rate signal is further corrected to obtain the final mass flow rate signal; Detect the temperature gradient between different components inside the mass flow controller; If the temperature gradient is greater than a preset threshold, a gradient compensation command is generated and superimposed on the drive signal of the control valve. Based on the comparison between the final mass flow rate signal and the set flow rate value, a closed-loop control command is generated to adjust the opening of the control valve.

[0006] Furthermore, prior to the step of using a preset segmented temperature-flow error database, the method further includes: The target wide temperature range is divided into continuous sub-temperature zones; Within each sub-temperature zone, error data between the original output signal of the flow sensor and the standard flow value at different flow points is collected through high-precision calibration to construct the segmented temperature-flow error database.

[0007] Furthermore, the step of performing a secondary correction on the corrected flow rate signal using the current fluid viscosity includes: For thermal flow sensors, the corrected flow signal is adjusted based on the ratio of the fluid viscosity at the reference temperature to the current fluid viscosity.

[0008] Furthermore, the step of detecting the temperature gradient between different components inside the mass flow controller includes: Temperature sensors are installed on the flow sensing chip, the control valve body, and the signal processing circuit board, respectively. Read the values ​​from each temperature sensor and calculate the absolute value of the temperature difference between any two temperature sensor values; The maximum value among all absolute temperature differences is defined as the temperature gradient.

[0009] Furthermore, the method also includes: Record the operating data of the mass flow controller during operation. The operating data includes the set flow rate, the measured flow rate, the ambient temperature, the fluid temperature, and the control error.

[0010] Furthermore, the method also includes: When a new fluid type is detected or the control error continues to exceed the preset tolerance range, the self-learning mechanism is triggered.

[0011] Furthermore, the method also includes: Using an incremental learning algorithm, the compensation model within the local sub-temperature zone is fine-tuned online based on the working data, and the segmented temperature-flow error database is updated.

[0012] Furthermore, the machine learning algorithm employs a feedforward neural network model.

[0013] Furthermore, the step of determining the target compensation coefficient corresponding to the ambient temperature and fluid temperature using a machine learning algorithm based on a preset segmented temperature-flow error database includes: The target sub-temperature zone to which the fluid belongs is determined based on the fluid temperature. Load the machine learning model corresponding to the target sub-temperature region; The target compensation coefficient is calculated by using the fluid temperature and set flow rate as inputs through the machine learning model.

[0014] A second aspect of the invention provides an adaptive temperature compensation system for a mass flow controller suitable for a wide temperature range, the system comprising: The system includes a main control unit, a flow sensing unit, a control valve unit, and a temperature sensing unit; the main control unit includes: The first acquisition module is used to acquire the ambient temperature of the environment where the mass flow controller is located and the fluid temperature of the fluid flowing through the mass flow controller. The compensation coefficient determination module is used to determine the target compensation coefficient corresponding to the ambient temperature and fluid temperature based on a preset segmented temperature-flow error database and a machine learning algorithm. The first correction module is used to correct the original output signal of the flow sensor according to the target compensation coefficient to obtain the corrected flow signal. The second acquisition module is used to acquire viscosity-temperature characteristic data corresponding to the fluid type, and query the current fluid viscosity based on the fluid temperature. The second correction module is used to perform a second correction on the corrected flow rate signal using the current fluid viscosity to obtain the final mass flow rate signal; A gradient detection module is used to detect the temperature gradient between different components inside the mass flow controller. The gradient compensation module is used to generate a gradient compensation command if the temperature gradient is greater than a preset threshold, and to superimpose the gradient compensation command onto the drive signal of the control valve. The closed-loop control module is used to generate closed-loop control commands to adjust the opening degree of the control valve based on the comparison result between the final mass flow rate signal and the set flow rate value.

[0015] The beneficial technical effects of the present invention are at least as follows: To address the aforementioned issues, this invention provides an adaptive temperature compensation method and system for mass flow controllers applicable to a wide temperature range. By constructing a segmented temperature-flow error database and utilizing machine learning algorithms for nonlinear fitting, it can accurately capture complex temperature-error relationships within a wide temperature range, effectively solving the problem of insufficient accuracy in traditional fixed-coefficient methods. Simultaneously, by introducing a real-time fluid viscosity-temperature mapping module, dynamic compensation is performed on the physical properties of different fluids at different temperatures, ensuring the physical accuracy of mass flow calculations. Furthermore, by adding a temperature gradient compensation unit, interference caused by uneven thermal fields within the MFC is actively sensed and counteracted, improving the system's dynamic stability and robustness. Finally, the integrated self-learning mechanism endows the system with long-term evolution capabilities, enabling it to adapt to new operating conditions and resist component aging, thereby comprehensively improving the control accuracy, reliability, and versatility of the mass flow controller in wide temperature range, multi-fluid, and long-cycle application scenarios. Attached Figure Description

[0016] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0017] Figure 1 This is a flowchart of an adaptive temperature compensation method for a mass flow controller applicable to a wide temperature range, according to the present invention.

[0018] Figure 2 This is a framework diagram of an adaptive temperature compensation system for a mass flow controller applicable to a wide temperature range, according to the present invention.

[0019] Figure 3 for Figure 2 A schematic diagram of the structure of the main control unit 1. The system comprises: 1. Main control unit; 2. Flow sensing unit; 3. Control valve unit; 4. Temperature sensing unit; 11. First acquisition module; 12. Compensation coefficient determination module; 13. First correction module; 14. Second acquisition module; 15. Second correction module; 16. Gradient detection module; 17. Gradient compensation module; and 18. Closed-loop control module. Detailed Implementation

[0020] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0021] In one or more embodiments, such as Figure 1 As shown, an adaptive temperature compensation method for a mass flow controller applicable to a wide temperature range is disclosed, the method comprising the following: S1. Obtain the ambient temperature of the environment where the mass flow controller is located and the fluid temperature of the fluid flowing through the mass flow controller. S2. Based on a preset segmented temperature-flow error database, a machine learning algorithm is used to determine the target compensation coefficient corresponding to the ambient temperature and fluid temperature. S3. Based on the target compensation coefficient, the original output signal of the flow sensor is corrected to obtain the corrected flow signal; S4. Obtain viscosity-temperature characteristic data corresponding to the fluid type, and query the current fluid viscosity based on the fluid temperature; S5. Using the current fluid viscosity, the corrected flow rate signal is further corrected to obtain the final mass flow rate signal; S6. Detect the temperature gradient between different components inside the mass flow controller; S7. If the temperature gradient is greater than a preset threshold, a gradient compensation command is generated and superimposed on the drive signal of the control valve. S8. Based on the comparison result between the final mass flow rate signal and the set flow rate value, generate a closed-loop control command to adjust the opening of the control valve.

[0022] In step S2, to achieve high-precision nonlinear compensation, a segmented temperature-flow error database must first be constructed. Specifically, the target wide temperature range of -40℃ to +150℃ is divided into multiple continuous sub-temperature zones. Within each sub-temperature zone, high-precision calibration is performed using a national metrological standard device, and error data at different flow rates is collected. Subsequently, a lightweight feedforward neural network model is independently trained for each sub-temperature zone. During equipment operation, the main control unit 1 determines the target sub-temperature zone to which it belongs based on the real-time fluid temperature, loads the corresponding neural network model, and quickly calculates the accurate target compensation coefficient using the fluid temperature and set flow rate as input.

[0023] The construction of the segmented temperature-flow error database is the foundation for achieving high-precision compensation in this invention. In specific implementation, the target application temperature range is first defined as [-40℃, +150℃], and then divided into 19 continuous and non-overlapping sub-temperature zones, each spanning 10℃, i.e., [-40, -30), [-30, -20), ..., [140, 150]. For each sub-temperature zone... , Perform the following calibration procedure: a) Place the mass flow controller to be calibrated in a high-precision constant temperature test chamber that has been certified by the National Metrology Institute, and stably control the air temperature inside the chamber at the center point of the sub-temperature zone (e.g., for the [-40, -30) sub-temperature zone, the control point is -35℃), and keep it for at least 2 hours to ensure that the equipment reaches complete thermal equilibrium. b) Connect the fluid inlet of the mass flow controller to a national first-class standard sonic nozzle flow meter, and connect the outlet to the atmosphere or a back pressure control system to ensure stable test pressure; c) Introduce a specified standard gas (such as nitrogen with a purity of 99.999%), and sequentially set the target flow rate of the mass flow controller to 10%, 20%, 30%, ..., 100% of the full scale (FS), for a total of 10 flow rate points; d) For each flow point, wait for the system to reach steady state (defined as the standard flow meter reading fluctuation being less than ±0.05% FS within 60 consecutive seconds), and then synchronously record the original differential voltage signal output by the mass flow controller flow sensing unit 2. The true mass flow rate given by the standard flow meter ( ), and precise fluid temperature ( ); e) Based on the original flow-voltage mapping relationship established at the factory under the reference temperature (25°C), Convert to uncompensated flow rate value Calculate the relative error under this operating condition. ; f) Repeat the above process to cover all 19 sub-temperature zones and 10 flow points, ultimately forming a four-dimensional dataset containing 1900 data samples {( , , , )}.

[0024] Subsequently, a lightweight feedforward neural network (MLP) model is independently trained for each subset of data in sub-temperature region i. The model structure is fixed at three layers: the input layer receives two features—fluid temperature (…). ) and set flow rate ( The hidden layer contains 12 neurons with ReLU activation; the output layer contains 1 neuron with linear activation, directly outputting the target compensation coefficient under this condition. The training process uses mean squared error (MSE) as the loss function and employs the Adam optimizer with an initial learning rate of 0.001 and a batch size of 32. To prevent overfitting, an L2 weight decay regularization term with a coefficient of 0.01 is introduced. After training, tools such as TensorFlow Lite are used to quantize the floating-point model into 8-bit integer (INT8) format to significantly reduce memory usage and computational power consumption. The quantized model weights are then stored in the on-chip Flash memory of the main control unit 1, with each sub-temperature region corresponding to an independent model file.

[0025] During normal operation of the mass flow controller, the main control unit 1 executes step S2: First, it reads the real-time fluid temperature. The target sub-temperature region index i is determined using simple interval judgment logic (if-else or table lookup). Next, the quantized MLP model corresponding to index i is loaded from Flash into RAM. Then, the current... and the user-defined data usage value As input, the neural network inference engine on the microcontroller is invoked, and the accurate target compensation coefficient is calculated in less than 1 millisecond. In step S3, the main control unit 1 uses this coefficient to perform linear correction on the original output signal of the flow sensor, obtaining a preliminarily corrected flow signal. .

[0026] In step S4, a fluid property parameter library is pre-stored in the non-volatile memory inside the main control unit 1. This library stores a set of Sutherland formula parameters for each supported process gas (e.g., H2, N2, O2, Ar, CH4), including a reference temperature T0 (273.15 K), a reference viscosity μ0 (Pa·s), and a Sutherland constant S (K). When the user specifies the current fluid type through the device's human-machine interface or Modbus communication protocol, the main control unit 1 retrieves the corresponding (S, T0, μ0) triple from this library.

[0027] Main control unit 1 uses the retrieved parameters and real-time fluid temperature (Convert to Kelvin temperature) Calculate the current fluid viscosity using the Sutherland formula. For thermal mass flow controllers, the basic working principle follows a simplified form of King's Law, namely, the heat loss of the sensor is proportional to the mass flow rate of the fluid and the square root of the fluid's thermal properties. Among these, fluid viscosity μ is a key property parameter affecting heat transfer efficiency. Therefore, to ensure the physical consistency of the mass flow rate calculation, a secondary correction must be performed based on viscosity changes. The correction logic is based on the following physical inference: at the same mass flow rate, the higher the fluid viscosity, the lower its flow rate, resulting in a weaker cooling effect sensed by the sensor and a smaller output signal. Therefore, the flow signal after the initial correction needs to be amplified. The specific correction operation is as follows: the main control unit 1 reads the reference viscosity pre-calculated and stored at the factory reference temperature (298.15 K). Then calculate the correction factor. Ultimately, a physically accurate final mass flow rate signal is obtained. .

[0028] In steps S6 and S7, to address internal thermal interference, temperature sensors are respectively arranged on the flow sensing chip, the control valve body, and the signal processing circuit board. The main control unit 1 reads the temperature at each point in real time, calculates the absolute value of the temperature difference between any two points, and takes the maximum value as the temperature gradient. When this gradient exceeds a preset threshold, the system determines that there is significant thermal interference and generates a feedforward gradient compensation command, which is superimposed on the main drive signal of the control valve to suppress control deviations that may be caused by thermal expansion or thermal drift in advance.

[0029] In step S7, the preset threshold is set to an empirical value, such as 5℃. When the main control unit 1 detects ΔT > 5℃, the system determines that there is a non-negligible internal thermal interference. At this time, the gradient compensation module 17 is activated. The core of this module is an empirical feedforward compensation model. This model is based on offline experimental data to identify: when the valve body temperature... Higher than sensor temperature When the valve seat material (usually stainless steel or Hastelloy) undergoes thermal expansion, it increases the actual flow area of ​​the valve, resulting in a larger actual flow rate under the same valve opening electrical signal. To counteract this effect, the system generates a negative compensation variable Δu. The magnitude of this compensation variable is related to the temperature difference. Proportional, that is The proportional coefficient α is a constant determined through system identification experiments, with a typical value of 0.5% per ℃. This compensation amount Δu is directly superimposed on the main drive signal calculated by the PID closed-loop control module 18. This forms the final drive signal sent to the solenoid coil of the control valve. This feedforward compensation mechanism can actively suppress interference before it affects the closed loop.

[0030] In one or more embodiments, such as Figure 2 As shown, an adaptive temperature compensation system for a mass flow controller suitable for a wide temperature range is disclosed, the system comprising: The system includes a main control unit 1, a flow sensing unit 2, a control valve unit 3, and a temperature sensing unit 4. The main control unit 1 includes: a first acquisition module 11, a compensation coefficient determination module 12, a first correction module 13, a second acquisition module 14, a second correction module 15, a gradient detection module 16, a gradient compensation module 17, and a closed-loop control module 18.

[0031] In this embodiment, the temperature sensing unit 4 consists of three high-precision, low-latency digital temperature sensors. The first sensor is a PT1000 platinum resistance thermometer, whose temperature probe is directly embedded into the wall of the fluid channel via laser welding, within 2 mm of the heating zone of the thermal flow sensor chip, to ensure the most accurate reflection of the actual dynamic temperature of the fluid flowing across the sensor surface. The second sensor is a TMP117 digital temperature sensor, which uses an I²C communication interface and is mounted on the printed circuit board (PCB) of the main control unit 1, away from any heat-generating components. It is used to monitor the macroscopic ambient temperature inside the equipment. The third sensor, also a TMP117, is tightly bonded to the stainless steel outer surface of the proportional control valve (PCV) using thermally conductive silicone grease. It is used to monitor the localized temperature of the valve body caused by fluid friction and the operation of the solenoid coil. The main control unit 1 synchronously polls the three sensors at a period of 10 milliseconds to obtain high-fidelity temperature field distribution data.

[0032] There are two triggering conditions for the self-learning mechanism: First, the user switches to a completely new fluid type through the human-machine interface, whose Sutherland parameters are not pre-stored in the property library; second, the system detects that the absolute value of the average control error continues to exceed 0.8% FS within 1000 consecutive control cycles (about 100 seconds), which usually indicates component aging or operating condition drift.

[0033] Once triggered, the system initiates an incremental learning process. It does not discard the existing, extensively calibrated global model, but instead fine-tunes only the model for the local sub-temperature zone where the current operating condition is located. Specifically, the system uses newly acquired working data with true error labels as new training samples and employs the Online Gradient Descent algorithm to iteratively update the output layer weights of the original MLP model in a small number of iterations (e.g., 5 to 10 rounds). The updated model must pass an internal validation process: it is tested using a reserved set of validation data. Only when the new model's prediction error on this dataset is reduced by at least 10% compared to the old model will it be officially adopted and override the old model in Flash.

[0034] It is worth noting that the specific workflow of the adaptive temperature compensation system for a mass flow controller applicable to a wide temperature range provided in this embodiment of the invention is the same as that of the adaptive temperature compensation method for a mass flow controller applicable to a wide temperature range described in the above embodiment, and will not be repeated here.

[0035] This invention also provides an adaptive temperature compensation device for a mass flow controller suitable for a wide temperature range, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the steps described in the above embodiment of an adaptive temperature compensation method for a mass flow controller suitable for a wide temperature range, for example... Figure 1 The steps S1 to S4 described above; or, when the processor executes the computer program, it implements the functions of each module in the above system embodiments.

[0036] For example, the computer program may be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the adaptive temperature compensation device for a wide temperature range mass flow controller.

[0037] The adaptive temperature compensation device for a wide-temperature-range mass flow controller can be a computing device such as a desktop computer, laptop, handheld computer, or cloud server. This device may include, but is not limited to, a processor and memory. Those skilled in the art will understand that the device may also include input / output devices, network access devices, buses, etc.

[0038] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASACs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. This processor is the control center of the adaptive temperature compensation device for a wide-temperature-range mass flow controller, connecting all parts of the device via various interfaces and lines.

[0039] The memory can be used to store the computer program and / or modules. The processor implements various functions of the adaptive temperature compensation device for a mass flow controller suitable for a wide temperature range by running or executing the computer program and / or modules stored in the memory and calling the data stored in the memory. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function, etc.; the data storage area may store data created based on the operation of the air conditioning controller, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital card (SD card), flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage devices.

[0040] The module integrated into the adaptive temperature compensation device for a mass flow controller suitable for a wide temperature range, if implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0041] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0042] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. An adaptive temperature compensation method for a mass flow controller suitable for a wide temperature range, characterized in that, The method includes: Obtain the ambient temperature of the environment where the mass flow controller is located and the temperature of the fluid flowing through the mass flow controller; Based on a pre-defined segmented temperature-flow error database, a machine learning algorithm is used to determine the target compensation coefficient corresponding to the ambient temperature and fluid temperature. Based on the target compensation coefficient, the original output signal of the flow sensor is corrected to obtain the corrected flow signal; Obtain viscosity-temperature characteristic data corresponding to the fluid type, and query the current fluid viscosity based on the fluid temperature; Using the current fluid viscosity, the corrected flow rate signal is further corrected to obtain the final mass flow rate signal; Detect the temperature gradient between different components inside the mass flow controller; If the temperature gradient is greater than a preset threshold, a gradient compensation command is generated and superimposed on the drive signal of the control valve. Based on the comparison between the final mass flow rate signal and the set flow rate value, a closed-loop control command is generated to adjust the opening of the control valve.

2. The adaptive temperature compensation method for a mass flow controller applicable to a wide temperature range according to claim 1, characterized in that, Prior to the step of using a preset segmented temperature-flow error database, the method further includes: The target wide temperature range is divided into continuous sub-temperature zones; Within each sub-temperature zone, error data between the original output signal of the flow sensor and the standard flow value at different flow points is collected through high-precision calibration to construct the segmented temperature-flow error database.

3. The adaptive temperature compensation method for a mass flow controller applicable to a wide temperature range according to claim 1, characterized in that, The step of performing a secondary correction on the corrected flow rate signal using the current fluid viscosity includes: For thermal flow sensors, the corrected flow signal is adjusted based on the ratio of the fluid viscosity at the reference temperature to the current fluid viscosity.

4. The adaptive temperature compensation method for a mass flow controller applicable to a wide temperature range according to claim 1, characterized in that, The step of detecting the temperature gradient between different components inside the mass flow controller includes: Temperature sensors are installed on the flow sensing chip, the control valve body, and the signal processing circuit board, respectively. Read the values ​​from each temperature sensor and calculate the absolute value of the temperature difference between any two temperature sensor values; The maximum value among all absolute temperature differences is defined as the temperature gradient.

5. The adaptive temperature compensation method for a mass flow controller applicable to a wide temperature range according to claim 1, characterized in that, The method further includes: Record the operating data of the mass flow controller during operation. The operating data includes the set flow rate, the measured flow rate, the ambient temperature, the fluid temperature, and the control error.

6. The adaptive temperature compensation method for a mass flow controller applicable to a wide temperature range according to claim 5, characterized in that, The method further includes triggering a self-learning mechanism when a new fluid type is detected or the control error continues to exceed a preset tolerance range.

7. The adaptive temperature compensation method for a mass flow controller applicable to a wide temperature range according to claim 6, characterized in that, The method further includes: Using an incremental learning algorithm, the compensation model within the local sub-temperature zone is fine-tuned online based on the working data, and the segmented temperature-flow error database is updated.

8. The adaptive temperature compensation method for a mass flow controller applicable to a wide temperature range according to claim 1, characterized in that, The machine learning algorithm uses a feedforward neural network model.

9. The adaptive temperature compensation method for a mass flow controller applicable to a wide temperature range according to claim 1, characterized in that, The step of determining the target compensation coefficient corresponding to the ambient temperature and fluid temperature using a machine learning algorithm based on a preset segmented temperature-flow error database includes: The target sub-temperature zone to which the fluid belongs is determined based on the fluid temperature. Load the machine learning model corresponding to the target sub-temperature region; The target compensation coefficient is calculated by using the fluid temperature and set flow rate as inputs through the machine learning model.

10. An adaptive temperature compensation system for a mass flow controller suitable for a wide temperature range, characterized in that, The system includes a main control unit, a flow sensing unit, a control valve unit, and a temperature sensing unit; the main control unit includes: The first acquisition module is used to acquire the ambient temperature of the environment where the mass flow controller is located and the fluid temperature of the fluid flowing through the mass flow controller. The compensation coefficient determination module is used to determine the target compensation coefficient corresponding to the ambient temperature and fluid temperature based on a preset segmented temperature-flow error database and a machine learning algorithm. The first correction module is used to correct the original output signal of the flow sensor according to the target compensation coefficient to obtain the corrected flow signal. The second acquisition module is used to acquire viscosity-temperature characteristic data corresponding to the fluid type, and query the current fluid viscosity based on the fluid temperature. The second correction module is used to perform a second correction on the corrected flow rate signal using the current fluid viscosity to obtain the final mass flow rate signal; A gradient detection module is used to detect the temperature gradient between different components inside the mass flow controller. The gradient compensation module is used to generate a gradient compensation command if the temperature gradient is greater than a preset threshold, and to superimpose the gradient compensation command onto the drive signal of the control valve. The closed-loop control module is used to generate closed-loop control commands to adjust the opening degree of the control valve based on the comparison result between the final mass flow rate signal and the set flow rate value.

Images