An air path control method, system, intelligent terminal and storage medium

Abstract

The application relates to the technical field of gas circuit control, and particularly discloses a gas circuit control method, a gas circuit control system, an intelligent terminal and a storage medium. The PID control algorithm is used to calculate a pressure control amount according to a current error, so that the system gradually reaches a stable state close to a target control value. Dynamic gain compensation and machine learning feedforward compensation are introduced to perform nonlinear compensation on disturbance caused by environmental temperature change, and rapid response is performed based on the rate of temperature change to compensate for the lag problem in the dynamic change process.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of gas path control, and more particularly, to a gas path control method and system, an intelligent terminal and a storage medium. BACKGROUND

[0002] In precision analytical instruments such as mass spectrometers, high stability and high precision gas path pressure control is the key to ensuring the accuracy and reproducibility of the analysis results. The gas path system is usually composed of a proportional valve, a pressure sensor and a controller. The controller collects the feedback signal of the pressure sensor and determines the opening of the proportional valve using a PID (Proportional-Integral-Derivative) control algorithm, so as to stabilize the gas path pressure around the set target value.

[0003] The sensor and the gas medium are affected by temperature changes, resulting in the stability of the gas path pressure being affected at different temperatures.

[0004] In the prior art, a simple PID controller is essentially a feedback mechanism for "after-the-fact adjustment". For problems caused by slow and continuous interference factors such as temperature changes, the traditional PID control has inherent defects: the controller only starts to act after the temperature change has caused a pressure deviation, and there is a lag in the control.

[0005] This results in different stability of the gas path system at different ambient temperatures, and even control instability at extreme temperatures, which cannot meet the stringent requirements of high-end precision instruments for pressure control robustness.

[0006] Therefore, there is a need for a gas path control method that can compensate for the effects of temperature changes to overcome the shortcomings of the prior art. SUMMARY

[0007] The technical problem to be solved by the present application is to provide a gas path control method, system, intelligent terminal and storage medium, which solves the problem of serious lag in single PID control in the prior art.

[0008] The technical problem to be solved by the present application is solved by the following technical solution:

[0009] In a first aspect, the present application provides a gas path control method applied to a mass spectrometer gas path module comprising a pressure sensor, a temperature sensor, a proportional valve and a controller. The gas path control method comprises:

[0010] obtaining a target pressure value and a pressure value sequence collected by the pressure sensor within a predetermined time period, wherein the last data of the pressure value sequence is a current pressure value;

[0011] calculating the difference between each data point in the pressure value sequence and the target pressure value to obtain an error pressure value sequence;

[0012] inputting the error pressure value sequence into a PID control algorithm to obtain a pressure control value;

[0013] obtaining an environment temperature value sequence collected by the temperature sensor within a predetermined time period;

[0014] determining a temperature calibration control value based on a trend correlation coefficient between the environment temperature value sequence and the pressure value sequence;

[0015] generating an opening control instruction for adjusting the proportional valve based on the pressure control value and the temperature calibration control value;

[0016] outputting the opening control instruction to the proportional valve, so that the proportional valve is adjusted according to the opening control instruction.

[0017] Further, the environment temperature value sequence and the pressure value sequence correspond to each other in the time dimension.

[0018] Further, the temperature calibration control value is determined based on a trend correlation coefficient between the environment temperature value sequence and the pressure value sequence, including:

[0019] for each data point of the environment temperature value sequence, calculating a difference value between the data point and a mean value of the sequence to obtain an environment temperature difference value sequence;

[0020] for each data point of the pressure value sequence, calculating a difference value between the data point and a mean value of the sequence to obtain a pressure difference value sequence;

[0021] calculating a product of each corresponding data point of the environment temperature difference value sequence and the pressure difference value sequence, and summing all product results to obtain an absolute value result as a change trend item;

[0022] calculating a product of a square root of a square sum of the environment temperature difference value sequence and a square root of a square sum of the pressure difference value sequence as a standard item;

[0023] calculating a trend correlation coefficient by dividing the change trend item by the standard item.

[0024] Further, the temperature calibration control value is determined based on a trend correlation coefficient between the environment temperature value sequence and the pressure value sequence, including:

[0025] when the trend correlation coefficient is less than a preset threshold value, taking the pressure control value as the temperature calibration control value;

[0026] when the trend correlation coefficient is greater than or equal to the preset threshold value, determining the temperature calibration control value based on a dynamic change rule of the environment temperature value sequence.

[0027] Further, when the trend correlation coefficient is greater than or equal to the preset threshold value, the temperature calibration control value is determined based on a dynamic change rule of the environment temperature value sequence, including:

[0028] determining a temperature change rate according to the sequence of ambient temperature values;

[0029] determining a dynamic gain compensation term based on the temperature change rate;

[0030] inputting the sequence of ambient temperature values into the trained machine learning based temperature-opening degree compensation model to obtain a feedforward compensation term;

[0031] calculating a mean value of the dynamic gain compensation term and the feedforward compensation term to determine a temperature calibration control value.

[0032] Further, the determining of the temperature change rate according to the sequence of ambient temperature values comprises:

[0033] extracting a sequence of temperature change rates from the sequence of ambient temperature values, a value in the sequence of temperature change rates being a temperature difference value between two adjacent ambient temperature values divided by a corresponding time interval;

[0034] calculating a mean value of the sequence of temperature change rates as the temperature change rate.

[0035] Further, the generating of the opening degree control instruction for adjusting the proportional valve based on the pressure control value and the temperature calibration value comprises:

[0036] calculating a temperature calibration confidence weight through a nonlinear mapping function based on the trend correlation coefficient;

[0037] calculating a pressure control confidence weight based on the temperature calibration confidence weight, a sum of the temperature calibration confidence weight and the pressure control confidence weight being one;

[0038] calculating a sum of a product of the pressure control value and the pressure control confidence weight and a product of the temperature calibration value and the temperature calibration confidence weight as the opening degree control value;

[0039] converting the opening degree control value into an electrical signal to obtain the opening degree control instruction of the proportional valve.

[0040] In a second aspect, the application further provides a gas path control system, which operates according to the above gas path control method.

[0041] In a third aspect, the application further provides an intelligent terminal, which comprises a memory and a processor coupled with the memory, the processor being configured to read and execute instructions stored in the memory, and the instructions, when executed by the processor, cause the processor to execute the above gas path control method.

[0042] In a fourth aspect, the application further provides a storage medium, which stores a computer program, and the computer program, when executed, implements the above gas path control method.

[0043] In summary, the application has the following beneficial technical effects:

[0044] The PID control algorithm calculates a pressure control amount according to a current error to gradually make the system reach a stable state close to a target control value, introduces dynamic gain compensation and machine learning feedforward compensation, performs nonlinear compensation on disturbances caused by changes in ambient temperature, and performs rapid response based on the rate of temperature change to compensate for the lag in the dynamic change process. BRIEF DESCRIPTION OF DRAWINGS

[0045] Figure 1 FIG. 1 is a flowchart of a gas path control method;

[0046] Figure 2 FIG. 2 is a gas path diagram in an embodiment. DETAILED DESCRIPTION

[0047] In order to facilitate understanding of the technical means, creative features, purposes and effects of the present application, the present application will be further described below in combination with specific drawings.

[0048] In actual application, the existing PID control scheme exposes some limitations.

[0049] For example, due to the dynamic temperature change of the local microenvironment where the sensor is located, the physical properties (such as sensitivity) of the pressure sensor as a sensing element will drift with the ambient temperature, resulting in temperature-dependent errors in the measured values that do not match the true pressure, further causing deviations in PID control and fluctuations in pressure.

[0050] Secondly, the change in ambient temperature caused by the heat generation of the instrument itself and other practical reasons will also affect the physical properties of the controlled gas medium, thereby causing the gas flow through the valve body to change at the same proportional valve opening, causing fluctuations in pressure.

[0051] In summary, the sensor and the gas medium will be affected by the change in temperature, resulting in the stability of the gas path pressure being affected at different temperatures.

[0052] A simple PID controller is essentially a feedback mechanism for "after-regulation". The traditional PID control has inherent defects: the controller only starts to act after the temperature change has caused a pressure deviation, and there is a lag in the control.

[0053] Based on this, as shown in FIG. 1, the present application discloses a gas path control method, which specifically comprises: Figure 1

[0054] S1, obtaining a target pressure value and a pressure value sequence collected by a pressure sensor within a predetermined time period, wherein the last data of the pressure value sequence is a current pressure value.

[0055] ​Here, the target pressure value is set by a user or an upper control system, and is input to a data processing center (e.g., a controller) through a human-machine interface, a communication interface, or a configuration file.

[0056] The pressure sensor deployed in the gas path module of the mass spectrometer continuously collects pressure value data at a fixed sampling frequency, such as Figure 2 “Pressure sensor 1” in FIG. 1. Specifically, the data processing center reads the pressure sensor signal through an ADC (analog-to-digital converter) and converts it into a digital pressure value. The controller maintains a circular buffer (e.g., storing the last 120 seconds of data) in memory, forming a time series of pressure values, with the last data in the sequence being the currently collected pressure value.

[0057] S2, calculate the difference between each data point in the pressure value sequence and the target pressure value to obtain an error pressure value sequence. Specifically, for each data point Pi in the pressure value sequence, perform ei = P target -Pi, P target is the target pressure value, and ei is the error pressure value at each time, with all eis forming the error pressure value sequence.

[0058] S3, input the error pressure value sequence into a PID control algorithm to obtain a pressure control value.

[0059] Here, the PID control algorithm is a feedback control algorithm, and its core idea is to calculate the control amount through the combination of proportional (P), integral (I), and differential (D) operations according to the current error, i.e., the difference between the target value and the actual value, to eliminate the error and make the system stable.

[0060] In the technical solution of the present application, a discrete PID algorithm is used to calculate the pressure control value, which is expressed by the formula:

[0061]

[0062] where e k is the last data value in the error pressure value sequence, i.e., the current error value, e k-1 is the error value at the previous time, the proportional term parameter Kp, the cumulative term parameter Ki, and the difference term parameter Kd are preset PID parameters, and the output U pid is the pressure control value, which has a mapping relationship with the opening degree (e.g., 0-100%) of the proportional valve.

[0063] In this algorithm, the error cumulative term (corresponding to the integral term of the continuous algorithm) and the difference term (corresponding to the differential term of the continuous algorithm) are updated once every sampling period, so as to quickly respond to the gas path control requirements in actual application.

[0064] PID parameter adjustment uses manual or automatic adjustment methods. Manual methods such as: by trial and observation of system response, first adjust the proportional term parameter Kp, find the appropriate proportional control strength; Then add the error accumulation term to adjust the accumulation term parameter Ki, reduce the steady-state error; Finally, adjust the difference term parameter Kd to enhance stability. Automatic adjustment such as: use Ziegler-Nichols method to automatically determine PID parameters. For example, for the reaction curve method, first make the system in pure proportional control, find the critical gain and critical period, and then calculate Kp, Ki and Kd according to the empirical formula.

[0065] In the S3 step, the PID control algorithm is the main force of feedback response, responsible for handling fast and random pressure disturbances, and is the basis of gas path control.

[0066] S4, obtaining the ambient temperature value sequence in the predetermined time period collected by the temperature sensor.

[0067] Specifically, the temperature sensor (such as PT100, thermocouple or digital sensor DS18B20) collects the ambient temperature at the same or similar sampling frequency. The controller reads the temperature value through ADC or digital interface (such as I 2 C, SPI), and also maintains an ambient temperature value sequence that is time-aligned with the pressure value sequence.

[0068] Preferably, the pressure sensor and the temperature sensor collect the two signals at the same frequency, and mark the collected data points with timestamps to ensure that the ambient temperature value sequence and the pressure value sequence correspond one by one in the time dimension.

[0069] In this way, the ambient temperature value and the pressure value correspond to each other in the time dimension, have a high causal relationship, and the trend correlation coefficient calculated in the subsequent steps can more truly and accurately reflect the internal relationship and lag correlation between the ambient temperature change and the pressure change, ensuring the accuracy of system perception and decision-making.

[0070] S5, determining the temperature calibration control value based on the trend correlation coefficient between the ambient temperature value sequence and the pressure value sequence.

[0071] In the technical solution of the present application, in order to intelligently judge whether the current pressure fluctuation is affected by temperature and decide whether to enable the temperature compensation mechanism, the trend correlation coefficient is introduced to judge whether the ambient temperature change and the pressure change are similar or related in trend, so as to avoid overcompensation when the correlation between temperature change and pressure change is low.

[0072] Specifically, first, for each data point of the ambient temperature value sequence, the difference between it and the sequence mean is calculated to obtain an ambient temperature difference value sequence; at the same time, for each data point of the pressure value sequence, the difference between it and the sequence mean is calculated to obtain a pressure difference value sequence; then, the product of each corresponding data point of the ambient temperature difference value sequence and the pressure difference value sequence is calculated, and the absolute value of the sum of all product results is obtained as a change trend item, where Ti is the ambient temperature value at the i-th moment, is the mean of the ambient temperature value sequence, is the mean of the pressure value sequence; and the product of the square root of the sum of squares of the ambient temperature difference value sequence and the square root of the sum of squares of the pressure difference value sequence is calculated as a standard item; finally, the change trend item is divided by the standard item to obtain the trend correlation coefficient.

[0073] In this way, the similarity or correlation between the ambient temperature value sequence and the pressure value sequence is measured. When the trend correlation coefficient is close to 1, it indicates that the change trend correlation between the two is high, and the temperature compensation mechanism needs to be enabled; when the trend correlation coefficient is close to 0, it indicates that the change trend correlation between the two is low, and the temperature compensation mechanism does not need to be enabled to avoid false compensation or overcompensation.

[0074] Further, when the trend correlation coefficient is less than a preset threshold, the pressure control value is taken as the temperature calibration control value to avoid overcompensation; when the trend correlation coefficient is greater than or equal to the preset threshold, the temperature compensation mechanism is enabled, and the temperature calibration control value is determined based on the dynamic change law of the ambient temperature value sequence. The preset threshold is determined through statistical analysis of experimental data. Here, a suitable preset threshold can ensure system stability while avoiding overcompensation of control operations.

[0075] For example, various working scenarios of the mass spectrometer are simulated in the laboratory, a temperature disturbance factor is added, the ambient temperature and pressure change conditions are recorded, and samples are divided from the recorded data, the trend correlation coefficient is calculated, each sample is labeled according to the experimental record whether the disturbance source is temperature dominant or non-temperature dominant, the intersection point of the data distribution is analyzed, and thus the preset threshold is determined.

[0076] The preset threshold can also be obtained by the following method: collecting historical running data of the gas path module under various working conditions, where the working conditions include temperature dominant disturbance working conditions and non-temperature dominant disturbance working conditions; calculating a plurality of trend correlation coefficient samples based on the historical running data, and classifying the samples according to the working conditions; and analyzing and selecting the trend correlation coefficient value that can best distinguish the two types of samples as the preset threshold through a statistical method.

[0077] More specifically, when the trend correlation coefficient is greater than or equal to a preset threshold, first, a temperature change rate is determined according to the sequence of the ambient temperature values. Then, a dynamic gain compensation term is determined based on the temperature change rate. Then, the sequence of the ambient temperature values is input into the trained machine learning-based temperature-opening degree compensation model to obtain a feedforward compensation term. Then, the mean value of the dynamic gain compensation term and the feedforward compensation term is calculated to determine the temperature calibration control value Ucomp.

[0078] In one example, determining the temperature change rate according to the sequence of the ambient temperature values comprises: extracting a temperature change rate sequence from the sequence of the ambient temperature values, the value in the temperature change rate sequence being the temperature difference value of two adjacent ambient temperature values divided by the corresponding time interval; and calculating the mean value of the temperature change rate sequence as the temperature change rate.

[0079] In one example, the dynamic gain compensation term Ucompdynamic=K dy ×Kt, K dy is a dynamic gain coefficient, which represents the sensitivity of the entire system to the speed of temperature change, and Kt is the temperature change rate. The dynamic gain compensation term here is to consider the dynamic effect brought by the temperature change trend.

[0080] In one example, the trained machine learning-based temperature-opening degree compensation model is a temperature-opening degree compensation model based on a recurrent neural network. The recurrent neural network (RNN) has a recurrent connection structure and has a timing memory capability. The neuron of the RNN receives not only the input at the current time but also the hidden state of itself at the last time as input. This allows information to be continuously transmitted within the sequence and form a "memory". In the technical solution of the present application, the RNN is used to process the sequence of the ambient temperature values, and can capture the dynamic change relationship of the temperature in the data.

[0081] More specifically, the sequence of the ambient temperature values is input into the trained temperature-opening degree compensation model based on the recurrent neural network to obtain a predicted value from the output layer of the model, i.e., the feedforward compensation term. The predicted value is the compensation amount for the opening degree of the proportional valve that should be applied at present, which is predicted by the temperature-opening degree compensation model based on the historical temperature law. In this way, the machine learning model is used to internalize the complex and nonlinear compensation relationship between temperature and opening degree into a "function", so that the system has intelligent decision-making ability based on data driving.

[0082] In addition, the process of training the temperature-opening degree compensation model based on the recurrent neural network comprises the following steps: obtaining a training environment temperature value sequence in a historical time period; dividing the training environment temperature value sequence to obtain X training environment temperature value subsequences, and extracting an actual feedforward compensation term corresponding to a last training environment temperature value in each training environment temperature value subsequence; constructing the temperature-opening degree compensation model based on the recurrent neural network; inputting the X training environment temperature value subsequences and the corresponding actual feedforward compensation terms as X training data pairs into the temperature-opening degree compensation model for iterative optimization training, wherein, in the training process, a mean square error or a mean absolute error between a training feedforward compensation term output by an output layer of the temperature-opening degree compensation model and the corresponding actual feedforward compensation term is taken as a loss function.

[0083] In particular, the actual feedforward compensation term cannot be directly measured at runtime, and is not a reading of a physical sensor. The actual feedforward compensation term represents a compensation amount that should be given to the system in advance (excluding the dynamic gain compensation term) in order to offset the impact of temperature on the system under the current conditions (environmental temperature). In the training phase, the value is obtained by back calculation from historical data of system operation, and then used as a training target value to train the recurrent neural network.

[0084] In an example, the entire control system is run in an actual device or in a high-precision simulation environment, and key variables such as temperature sensor readings, target pressure values, actual pressure values, actual opening degree control instructions, actual pressure control values, and actual dynamic gain compensation terms are recorded at a predetermined frequency. High-quality data points are selected from the recorded data, and for the selected data points, the actual feedforward compensation term = total control value corresponding to the actual opening degree control instruction (pressure control value + temperature calibration value) - actual pressure control value (PID output) - actual dynamic gain compensation term is calculated.

[0085] Here, the high-quality data points refer to the system being in or close to a stable state, i.e., the actual pressure is very close to the target pressure, and data in abnormal time periods such as sensor failure and communication terminal are excluded.

[0086] It is worth mentioning that in many cases, in order to simplify the problem, the actual feedforward compensation term can be temporarily regarded as 0 or considered to have a small contribution when collecting data and calculating the actual feedforward compensation term. The reason is that the main role of the dynamic gain compensation term is to handle rapid dynamic changes, and the main role of the feedforward compensation term is to handle steady-state or slowly varying disturbances caused by temperature. In order to simplify the problem, the confidence weights are all regarded as 1 in the above process, and the actual confidence weight values need to be considered in the actual operation process.

[0087] Through the above processing, the dynamic gain compensation + machine learning feedforward compensation mode combines the advantages of the two compensation ideas. The feedforward compensation term obtained by the machine learning model can predict how much opening degree the proportional valve needs to have at a certain temperature point to maintain the target pressure through learning a large amount of historical data, and it is better at processing nonlinear relationships. The dynamic gain compensation term directly responds to the rate of temperature change. The faster the temperature changes, the greater the compensation amount it gives. This way can make up for the lag of the system in the dynamic process. Taking the average of the two as the final temperature calibration value can include accurate judgments based on long-term experience and agility in responding to sudden changes, so as to perform well in various temperature change scenarios (slow drift or rapid fluctuation).

[0088] S6, generating an opening control instruction for adjusting the proportional valve based on the pressure control value and the temperature calibration value.

[0089] Further, based on the trend correlation coefficient, a temperature calibration confidence weight α is calculated through a nonlinear mapping function, for example, using a Sigmoid function; then a pressure control confidence weight is calculated based on the temperature calibration confidence weight, the sum of the temperature calibration confidence weight and the pressure control confidence weight is one; and the product of the pressure control value and the pressure control confidence weight (1-α)U pid The sum of the product of the temperature calibration value and the temperature calibration confidence weight αUcomp and (1-α)U pid +αUcomp as the opening control value; and converting the opening control value into an electrical signal to obtain the opening control instruction of the proportional valve.

[0090] S7, outputting the opening control instruction to the proportional valve, so that the proportional valve adjusts according to the opening control instruction. Wherein, the opening control instruction is sent to the driver of the proportional valve, the driver converts the electrical signal into electromagnetic force, drives the valve core to move, changes the valve opening, and adjusts the gas flow.

[0091] In the above technical solution, the temperature calibration compensation undertakes the main task of resisting temperature interference, so that the gas path control reduces the lag and speeds up the stability.

[0092] The application also provides a gas path control system which implements the method provided in any of the embodiments of the application during operation.

[0093] The application also provides a storage medium which stores a computer program, and the computer program is executed to implement the method provided in any of the embodiments of the application.

[0094] The application further discloses an intelligent terminal, which comprises a processor coupled with a memory and reading and executing instructions stored in the memory; when the processor executes the instructions, the processor is used to execute the gas path control method.

[0095] The memory can be an internal storage unit of the terminal in some embodiments, such as a hard disk or a memory of the terminal.

[0096] The memory can also be an external storage device of the terminal in some other embodiments, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc.

[0097] Further, the memory can include both an internal storage unit and an external storage device of the terminal.

[0098] The memory is used to store application software and various data installed on the smart terminal, such as program codes installed on the smart terminal, etc.

[0099] The memory can also be used to temporarily store data that has been output or will be output. In an embodiment, the memory stores instructions that can be executed by the processor to perform the air path control method.

[0100] The processor can be a central processing unit (CPU), a microprocessor or other data processing chip in some embodiments, used to run instructions or process data stored in the memory, such as to perform the air path control method, etc.

[0101] The above shows and describes the basic principles, main features and advantages of the present application. Those skilled in the art should understand that the present application is not limited to the above embodiments, and various changes and improvements can be made to the present application without departing from the spirit and scope of the present application. These changes and improvements all fall within the scope of the present application. The scope of the present application is defined by the appended claims and their equivalents.

Claims

1. An air path control method characterized by, The gas path control method is applied to a mass spectrometer gas path module comprising a pressure sensor, a temperature sensor, a proportional valve and a controller, and comprises the following steps: obtaining a target pressure value and a pressure value sequence collected by the pressure sensor within a predetermined time period, wherein the last data of the pressure value sequence is a current pressure value; calculating the difference between each data point in the pressure value sequence and the target pressure value to obtain an error pressure value sequence; inputting the error pressure value sequence into a PID control algorithm to obtain a pressure control value; obtaining an ambient temperature value sequence collected by the temperature sensor within the predetermined time period; determining a temperature calibration control value based on the trend correlation coefficient between the ambient temperature value sequence and the pressure value sequence; generating an opening degree control instruction for adjusting the proportional valve based on the pressure control value and the temperature calibration control value; outputting the opening degree control instruction to the proportional valve so that the proportional valve is adjusted according to the opening degree control instruction; determining a temperature calibration control value based on the trend correlation coefficient between the ambient temperature value sequence and the pressure value sequence, comprising: when the trend correlation coefficient is less than a preset threshold, taking the pressure control value as the temperature calibration control value; when the trend correlation coefficient is greater than or equal to the preset threshold, determining the temperature calibration control value based on the dynamic change law of the ambient temperature value sequence; when the trend correlation coefficient is greater than or equal to the preset threshold, determining the temperature calibration control value based on the dynamic change law of the ambient temperature value sequence, comprising: determining a temperature change rate according to the ambient temperature value sequence; determining a dynamic gain compensation term based on the temperature change rate; inputting the ambient temperature value sequence into a trained temperature-opening degree compensation model based on machine learning to obtain a feedforward compensation term; calculating the mean value of the dynamic gain compensation term and the feedforward compensation term to determine the temperature calibration control value.

2. The air path control method according to claim 1, characterized by, The ambient temperature value sequence and the pressure value sequence correspond to each other in the time dimension.

3. The air path control method according to claim 1, characterized by, determining a temperature calibration control value based on the trend correlation coefficient between the ambient temperature value sequence and the pressure value sequence, comprising: for each data point of the ambient temperature value sequence, calculating the difference between it and the sequence mean to obtain an ambient temperature difference value sequence; for each data point of the pressure value sequence, calculating the difference between it and the sequence mean to obtain a pressure difference value sequence; calculating the product of each corresponding data point of the ambient temperature difference value sequence and the pressure difference value sequence, and taking the absolute value of the sum of all product results as a change trend term; calculating the product of the square root of the sum of squares of the ambient temperature difference value sequence and the square root of the sum of squares of the pressure difference value sequence as a standard term; calculating the trend correlation coefficient by dividing the change trend term by the standard term.

4. The air path control method according to claim 1, characterized by, determining a temperature change rate according to the ambient temperature value sequence, comprising: extracting a temperature change rate sequence from the ambient temperature value sequence, wherein the values in the temperature change rate sequence are the temperature difference values of adjacent two ambient temperature values divided by the corresponding time interval; calculating the mean value of the temperature change rate sequence as the temperature change rate.

5. The air path control method according to claim 1, characterized by, generating an opening degree control instruction for adjusting the proportional valve based on the pressure control value and the temperature calibration value, comprising: calculating a temperature calibration confidence weight by a nonlinear mapping function based on the trend correlation coefficient; calculating a pressure control confidence weight based on the temperature calibration confidence weight, wherein the sum of the temperature calibration confidence weight and the pressure control confidence weight is one; calculating the sum of the product of the pressure control value and the pressure control confidence weight and the product of the temperature calibration value and the temperature calibration confidence weight as an opening degree control value; converting the opening degree control value into an electrical signal to obtain the opening degree control instruction of the proportional valve.

6. An air path control system characterized by comprising: operating the method of any one of claims 1-5.

7. A smart terminal, characterized in that comprising a memory and a processor coupled to the memory, the processor configured to read and execute instructions stored in the memory, and when the processor executes the instructions, the processor is configured to perform the method of any one of claims 1-5.

8. A storage medium, characterized by a storage medium having stored therein a computer program, which, when executed, implements the method of any one of claims 1-5.

Images