Non-contact fluid temperature measurement system and compensation method

Abstract

The application provides a non-direct contact fluid temperature measurement system and a compensation method, and relates to the technical field of temperature detection. The temperature measurement system comprises a dynamic compensation module, a steady-state deviation compensation module and a steady-state judgment module. A temperature sensor collects the temperature value of a non-direct contact fluid, and the temperature value is subjected to hysteresis compensation by the dynamic compensation module. The steady-state judgment module judges whether the temperature value after hysteresis compensation is in a steady state, and if not, the temperature value after hysteresis compensation is output. If yes, the temperature value after hysteresis compensation is subjected to steady-state deviation compensation by the steady-state deviation compensation module and then output. The temperature measurement system can effectively improve the accuracy and response characteristics of non-contact fluid temperature measurement by combining the double mechanisms of dynamic compensation and steady-state deviation compensation.

Description

Technical Field

[0001] This application relates to the field of temperature detection technology, specifically to a temperature measurement system and compensation method that does not directly contact the fluid. Background Technology

[0002] Temperature measurement is mainly divided into non-contact and contact temperature measurement. Non-contact temperature measurement primarily uses radiation, such as infrared thermometers. Non-contact methods have low thermal inertia and can effectively measure rapidly changing temperatures. However, they are affected by environmental radiation and distance, resulting in lower repeatability, typically ±2℃, and cannot effectively measure core temperatures (such as the true temperature of fluid inside a pipe). Contact temperature measurement, on the other hand, involves direct contact with the object being measured. After thermal equilibrium is reached through heat conduction, the temperature of the object is measured. Contact temperature measurement offers advantages such as high accuracy, reliable results, and less susceptibility to interference from environmental factors.

[0003] However, traditional contact temperature measurement methods have several drawbacks when measuring the temperature of corrosive fluids in pipelines. For example, the design and planning of pipelines requires the installation of temperature sensing holes into which the temperature sensor is inserted, increasing design costs. Furthermore, if the temperature sensor needs to be inserted into the fluid through a thermally conductive protective sleeve, the sleeve needs to be inspected and replaced regularly, increasing maintenance costs. Therefore, in existing technologies, the temperature of corrosive fluids in pipelines is often measured using non-contact temperature measurement. However, because the temperature sensor in non-contact methods needs to be installed on the outer wall of the pipeline, there is thermal resistance between the sensor and the corrosive fluid due to factors such as the casing, curing agent, and pipeline wall. This results in a hysteresis in the actual measurement of the fluid temperature. After a period of time, once thermal equilibrium is reached, the temperature difference between the medium and the external environment causes a steady-state deviation between the temperature measured by the sensor and the medium temperature. Therefore, existing non-contact temperature measurement solutions suffer from low measurement accuracy. Summary of the Invention

[0004] This application provides a hysteresis and deviation compensation system and method for a temperature measurement system that does not directly contact the fluid. By combining dynamic compensation and steady-state deviation compensation, the accuracy and response characteristics of non-contact fluid temperature measurement can be effectively improved.

[0005] In a first aspect, embodiments of this application provide a temperature measurement system for non-direct contact fluids, comprising: a dynamic compensation module, a steady-state deviation compensation module, and a steady-state judgment module; a temperature sensor acquires the temperature value of the non-directly contacting fluid, and the temperature value is hysteresis compensated by the dynamic compensation module; the steady-state judgment module determines whether the temperature value after hysteresis compensation is in a steady state; if not, it outputs the hysteresis-compensated temperature value; if so, the steady-state deviation compensation module performs steady-state deviation compensation on the hysteresis-compensated temperature value and outputs it.

[0006] In some embodiments, the steady-state determination module includes a unit delay unit; the steady-state determination module receives the temperature value after hysteresis compensation, and the unit delay unit records the temperature value after hysteresis compensation at the previous moment; the temperature change is calculated based on the temperature value after hysteresis compensation at the current moment and the temperature value after hysteresis compensation at the previous moment; the temperature change is compared with a set threshold, and if the temperature change is less than or equal to the set threshold, the fluid is in a steady state; if the temperature change is greater than the set threshold, the fluid is not in a steady state.

[0007] In some embodiments, the dynamic compensation module includes a dynamic prediction compensation model, which is established based on the dynamic response characteristics of the temperature sensor.

[0008] Based on the known equivalent fixed thermal resistance R2 and variable thermal resistance R1, the dynamic response characteristics of the temperature sensor for the multilayer heat transfer medium are described as a nonlinear temperature measurement response system, which is then decomposed into a linear dynamic subsystem and a memoryless nonlinear subsystem.

[0009] A( x(k)= B( )u(k)

[0010] The formula y(k) = f[x(k)] + ε(k) <3>

[0011] Mode <3> In this context, x(k) represents the internal state or intermediate output of the system.

[0012] u(k) is the system input;

[0013] A ( ) and B ( ) is a polynomial, representing the transfer function of the input and output;

[0014] d is the delay order, which represents the delay effect of the input on the output;

[0015] y(k) is the final output of the system;

[0016] f[x(k)] is a nonlinear function that describes the internal state and the mapping from heat flow Q at temperature T to the final output;

[0017] ξ(k) is the noise or interference term.

[0018] Assuming it's white noise, according to the whitening filter 1 / A( The dynamic prediction and compensation model is determined as follows:

[0019] A( )y_dc(k)=B( )y(k)+ξ(k)

[0020] ξ(k)= A( ) ε(k) <4>

[0021] Mode <4> In this context, ε(k) represents white noise;

[0022] The transfer function H(Z) corresponding to the dynamic prediction compensation model -1 )for:

[0023] Mode <5>

[0024] Mode <5> In this context, w represents the order of the input transfer function, and v represents the order of the output transfer function.

[0025] At time j, the output y_dc(j) of the dynamic prediction compensation model is:

[0026] Mode <6> .

[0027] In some embodiments, the steady-state deviation compensation module includes a temperature steady-state deviation compensation model; the temperature steady-state deviation compensation model is established based on the time-domain response characteristics of the temperature sensor.

[0028] The time-domain response characteristics of the temperature sensor are:

[0029] Mode <1>

[0030] In the formula , representing the convective heat transfer capacity and thermal resistance between the fluid and the shell; C1 represents the thermal resistance related to the fluid type; C2 represents the heat capacity related to the fluid type.

[0031] In the formula R2 represents the equivalent fixed thermal resistance, including the thermal resistance of the housing, the thermal resistance of the curing agent, and the thermal resistance of the surface encapsulation layer, which is a constant value; C2 is the equivalent heat capacity, which is a constant value.

[0032] Cm x is the current mass specific heat capacity of the fluid, K is the structure-related heat flux density mapping coefficient, which is a constant, and Qmass represents the current mass flow rate of the fluid. This indicates the measured value from the temperature sensor; This represents the fluid temperature after reaching a steady state.

[0033] When the time is long enough, i.e. t approaches infinity, the temperature steady-state deviation compensation model is obtained as follows:

[0034] Mode <2> .

[0035] Secondly, embodiments of this application provide a method for hysteresis and deviation compensation in a temperature measurement system that does not directly contact the fluid, implemented using the aforementioned temperature sensor. The method includes: establishing a temperature steady-state deviation compensation model for steady-state deviation compensation.

[0036] A dynamic predictive compensation model for hysteresis compensation is established. The dynamic predictive compensation model performs hysteresis compensation based on the measured values ​​of the fluid that is not in direct contact with the fluid, and obtains the hysteresis-compensated temperature value. Based on the hysteresis-compensated temperature value at the current moment and the hysteresis-compensated temperature value at the previous moment, it is determined whether the fluid is in a steady state. If so, the temperature steady-state deviation compensation model performs steady-state compensation on the hysteresis-compensated temperature value and outputs it; if not, the hysteresis-compensated temperature value is output.

[0037] In some embodiments, the establishment of the temperature steady-state deviation compensation model includes the following steps:

[0038] A temperature sensor is mounted on the outer wall of the housing; a surface encapsulation layer and a curing agent are provided between the temperature sensor and the housing, and fluid is introduced into the interior of the housing;

[0039] The variable thermal resistance caused by different fluid types is equivalent to the variable thermal resistance R1; the thermal resistance of the shell, the thermal resistance of the curing agent, and the thermal resistance of the surface encapsulation layer are equivalent to the fixed thermal resistance R2.

[0040] The time-domain response characteristics of the temperature sensor are:

[0041] Mode <1>

[0042] In the formula , representing the convective heat transfer capacity and thermal resistance between the fluid and the shell; C1 represents the thermal resistance related to the fluid type; C2 represents the heat capacity related to the fluid type.

[0043] In the formula R2 represents the equivalent fixed thermal resistance, including the thermal resistance of the housing, the thermal resistance of the curing agent, and the thermal resistance of the surface encapsulation layer, which is a constant value; C2 is the equivalent heat capacity, which is a constant value.

[0044] Cm x is the current mass specific heat capacity of the fluid, K is the structure-related heat flux density mapping coefficient, which is a constant, and Qmass represents the current mass flow rate of the fluid. This indicates the measured value from the temperature sensor; This represents the fluid temperature after reaching a steady state.

[0045] When the time is long enough, i.e. t approaches infinity, the temperature steady-state deviation compensation model is obtained as follows:

[0046] Mode <2> .

[0047] In some embodiments, the method further includes the following steps:

[0048] A calibration temperature probe is installed on the inner wall of the housing and sealed. Then, a calibration fluid is introduced, and the calibration temperature probe is immersed in the current calibration fluid environment. The calibration fluid is a non-toxic, non-corrosive gas.

[0049] Introduce the first set of calibration fluid at temperature T1, and record the measured values ​​of the calibration temperature probe, the temperature sensor, and the current mass flow rate.

[0050] A second set of calibration fluids at a temperature of T2 was introduced, and the measured values ​​of the calibration temperature probe, the temperature sensor, and the current mass flow rate were recorded. The second set of calibration fluids was of the same type as the first set of calibration fluids, but at a different temperature.

[0051] The measurements from the calibration temperature probe, temperature sensor, and current mass flow rate when the first set of calibration fluids is introduced, and the measurements from the calibration temperature probe, temperature sensor, and current mass flow rate when the second set of calibration fluids is introduced, along with the specific heat capacity parameter Cm of the calibration fluids, are recorded. x and the variable thermal resistance R1 substituting <2> In the process, the heat flux density mapping coefficient K and the equivalent fixed thermal resistance R2 associated with the structure are obtained by solving, and then the temperature steady-state deviation compensation model is updated.

[0052] In some embodiments, the establishment of the dynamic prediction compensation model includes the following steps:

[0053] Based on the equivalent fixed thermal resistance R2 and the variable thermal resistance R1, the dynamic response characteristics of the temperature sensor of the multilayer heat transfer medium are described as a nonlinear temperature measurement response system, and the system is decomposed into a linear dynamic subsystem and a memoryless nonlinear subsystem:

[0054] A( x(k)= B( )u(k)

[0055] The formula y(k) = f[x(k)] + ξ(k) <3>

[0056] Mode <3> In this context, x(k) represents the internal state or intermediate output of the system.

[0057] u(k) is the system input;

[0058] A ( ) and B ( ) is a polynomial, representing the transfer function of the input and output;

[0059] d is the delay order, which represents the delay effect of the input on the output;

[0060] y(k) is the final output of the system;

[0061] f[x(k)] is a nonlinear function that describes the internal state and the mapping from heat flow Q at temperature T to the final output;

[0062] ξ(k) is the noise or interference term.

[0063] Assuming it's white noise, according to the whitening filter 1 / A( The dynamic prediction and compensation model is designed as follows:

[0064] A( )y_dc(k)=B( )y(k)+ξ(k)

[0065] ξ(k)= A( ) ε(k) <4>

[0066] Mode <4> In this context, ε(k) represents white noise;

[0067] The transfer function H(Z) corresponding to the dynamic prediction compensation model -1 )for:

[0068] Mode <5>

[0069] Mode <5> In this context, w represents the order of the input transfer function, and v represents the order of the output transfer function.

[0070] At time j, the output y_dc(j) of the dynamic prediction compensation model is:

[0071] Mode <6> .

[0072] In some embodiments, the method further includes the following steps:

[0073] A performance evaluation function for the dynamic prediction and compensation model, including overshoot constraints, is constructed. The performance evaluation function is expressed as follows:

[0074] Mode <7>

[0075] For real-time dynamic error terms, This is the transient overshoot error term. This is the global overshoot error term. For smoothness constraints;

[0076] Mode <7> middle Let be the critical threshold of the overshoot control function, and be a constant; m and n are constants, where m>n and n>1 are even numbers. This is the maximum value of the current temperature series;

[0077] Err_Avg The average compensation error is expressed as follows:

[0078] Mode <8> .

[0079] In some embodiments, the method further includes the following steps:

[0080] A temperature sensor is installed on the outer wall of the housing; a surface encapsulation layer and a curing agent are provided between the temperature sensor and the housing; a calibration temperature probe is installed on the inner wall of the housing and sealed, then a calibration fluid is introduced, and the calibration temperature probe is immersed in the current calibration fluid environment; a first group of calibration fluids, a second group of calibration fluids, ..., a Wth group of calibration fluids with different temperatures are sequentially introduced into the housing; the first group of calibration fluids, the second group of calibration fluids, ..., the Wth group of calibration fluids have different temperatures but the same fluid type, and the measured values ​​of the calibration temperature probe and the temperature sensor for the first group to the Wth group of calibration fluids are recorded respectively; a sequence of measured values ​​of the calibration temperature probe and the temperature sensor changing over time is formed; the measured value of the calibration temperature probe is taken as the expected temperature value, the matrix parameters of the dynamic prediction compensation model are obtained, and the dynamic prediction compensation model is updated; a non-calibrated fluid is introduced, the measured value is obtained through the temperature sensor, and the first compensation value of the non-calibrated fluid is calculated based on the dynamic prediction compensation model; the non-calibrated fluid is a toxic and highly corrosive fluid.

[0081] Compared with existing technologies, the advantages of this application are: the dynamic compensation module reduces hysteresis bias in transient processes, improving the system's tracking capability during temperature changes; while the steady-state bias compensation module specifically corrects for hysteresis effects, reducing measurement errors in steady state. By combining the dual mechanisms of dynamic compensation and steady-state bias compensation, the accuracy and response characteristics of non-contact fluid temperature measurement can be effectively improved, thereby enhancing the adaptability and reliability of the temperature sensor under non-direct contact conditions.

[0082] This invention can establish corresponding dynamic supplementation models and steady-state deviation compensation models through calibrated safe fluids, and use indirect measurement methods to accurately measure the temperature of uncalibrated toxic and highly corrosive gases. Attached Figure Description

[0083] Figure 1 This is a schematic diagram of the architecture of a temperature measurement system that does not directly contact the fluid, as provided in an embodiment of this application.

[0084] Figure 2This is a schematic diagram of the installation structure of a temperature sensor that does not directly contact the fluid, as provided in an embodiment of this application.

[0085] Figure 3 This is a schematic diagram of the equivalent temperature measurement model provided in the embodiments of this application.

[0086] Figure 4 This is a schematic diagram of the calibration group temperature measurement structure provided in the embodiments of this application.

[0087] Figure 5 This is a schematic diagram of the structure for compensating the temperature measurement system that does not directly contact the fluid provided in the embodiments of this application.

[0088] Figure 6 This is a schematic diagram of the results of the high-purity nitrogen gas flow test group at 27°C provided in the embodiments of this application.

[0089] Figure 7 This is a schematic diagram of the results of the high-purity nitrogen gas test group at 42°C provided in the embodiments of this application.

[0090] Figure 8 This is a schematic diagram of the results of the high-purity carbon dioxide 32°C test group provided in the embodiments of this application. Detailed Implementation

[0091] The present application will now be described in further detail with reference to experimental examples and specific embodiments. However, this should not be construed as limiting the scope of the subject matter of the present application to the following embodiments. All technologies implemented based on the content of the present application fall within the scope of protection of the present application.

[0092] Unless otherwise specified, the terms "upper," "lower," "left," "right," "center," "inner," "outer," and "side" used in the description of specific embodiments of this application to indicate orientation or positional relationships are based on the orientation or positional relationships shown in the accompanying drawings, or the orientation or positional relationship in which the product / equipment / device is usually placed during use. These terms are merely for the purpose of facilitating the description of the solution in this application or simplifying the description in specific embodiments, so as to enable those skilled in the art to quickly understand the solution, and do not indicate or imply that a particular device / component / element must have a specific orientation, or be constructed and operated in a specific positional relationship. Therefore, they should not be construed as limitations on this application.

[0093] In the description of the embodiments of this application, technical terms such as "first" and "second" only distinguish one entity or operation from another, and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary or secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0094] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0095] During the research process, the applicant discovered that there are two main existing methods for reducing hysteresis in temperature measurement: one is to use temperature sensors with smaller time constants, but this still cannot solve the problem of large time constants in the heat transfer medium (shell, curing agent thermal capacity and thermal resistance); the other is to use genetic algorithms, particle swarm optimization, etc., to train dynamic compensation models, with the measured medium being the same as the training medium, i.e., when the system characteristics remain unchanged, the temperature measurement value is dynamically compensated in real time. In dynamic fluid temperature detection and control applications, the existence of temperature hysteresis and steady-state deviation leads to large errors in fluid temperature measurement and control. Existing dynamic measurement methods require compensation of the system based on the known dynamic characteristics of the system. In schemes using genetic algorithms and particle swarm optimization, the overshoot problem of dynamic compensation methods is not considered, i.e., the instantaneous temperature compensation calculation output value is greater than the steady-state fluid temperature. When the calculated value of the above dynamic compensation method is applied to the temperature alarm monitoring system and related control of fluid heaters, it will cause false alarms and shutdowns and erroneous heater responses. Moreover, when the system characteristics change, such as when the fluid type changes, the thermal capacity and thermal resistance characteristics of the system will also change due to the different convective heat transfer coefficients of the fluids, resulting in even greater errors in the measurement results. In practical industrial applications, it is often necessary to measure the temperature of fluids other than nitrogen and air that are toxic or highly corrosive. Therefore, it is not possible to use the fluid that actually needs to be measured as a training group to collect data. Existing methods do not have the broad applicability of switching fluids.

[0096] Based on this, and addressing the limitations of existing technologies, this application provides a temperature measurement system that does not directly contact the fluid. Please refer to... Figure 1 , Figure 1 This is a schematic diagram of the architecture of a temperature measurement system that does not directly contact the fluid, provided in an embodiment of this application. The temperature measurement system 10 that does not directly contact the fluid may include: a dynamic compensation module 11, a steady-state deviation compensation module 12, and a steady-state judgment module 13.

[0097] A temperature sensor acquires the temperature value of a fluid that is not in direct contact with the sensor. The temperature value is then hysteresis-compensated by the dynamic compensation module. The dynamic compensation module 11 performs compensation calculations using a dynamic compensation algorithm to dynamically compensate for the hysteresis of the original measured value from the temperature sensor. "Dynamic" means that it can adjust the compensation value over time to track the transient response of the temperature sensor. The steady-state deviation compensation module 12 performs steady-state compensation on the fluid's measured value using a temperature steady-state deviation compensation model. The steady-state judgment module 13 determines whether the temperature value after hysteresis compensation is in a steady state. If not, it outputs the hysteresis-compensated temperature value; if so, the steady-state deviation compensation module performs steady-state deviation compensation on the hysteresis-compensated temperature value before outputting it.

[0098] Among them, non-direct contact with fluid refers to the installation method in which the temperature sensor is isolated from the fluid through structural media such as housing and curing agent. Although this installation protects the sensor, it introduces an additional thermal resistance layer, which leads to an extension of the heat transfer path and an increase in heat capacity, thereby causing measurement deviation.

[0099] The process of establishing the temperature steady-state deviation compensation model and the dynamic prediction compensation model in this embodiment includes:

[0100] A temperature steady-state deviation compensation model is established based on the fixed structural characteristics of the temperature sensor. Please refer to [link / reference]. Figure 2 , Figure 2 This is a schematic diagram of the installation structure of a temperature sensor that does not directly contact the fluid, as provided in this application embodiment. The temperature sensor is encapsulated and fixed by a curing agent, which can be epoxy resin. The entire sensor and curing agent are encapsulated within a housing. The housing can be made of stainless steel or other metal materials, primarily used to protect the fragile internal sensor and circuitry from mechanical damage, corrosion, and high-pressure environments. The outermost layer of the housing in contact with the fluid is a surface encapsulation layer made of polytetrafluoroethylene (PTFE). The arrows in the diagram indicate the heat flow path. The medium temperature (T_flow_X) is the starting point of the heat transfer process, representing the true temperature of the fluid within the pipe or cavity, and is the target value that the measurement system ultimately needs to obtain. Heat is transferred from the fluid core to the sensor housing surface, overcoming the resistance of the convective heat transfer boundary layer. The convective heat transfer boundary layer temperature T_shell can be considered a transition range from T_flow_X to the housing surface temperature. While heat continuously flows towards the sensor, it also dissipates to the external ambient temperature (T_ambient) through the housing. Ultimately, the actual temperature value measured by the temperature sensor (T_sensors) comprehensively reflects the equilibrium state of the above thermal processes.

[0101] according to Figure 2Given the fixed structural characteristics of the temperature sensor, for an individual temperature sensor after installation, the thermal capacity and thermal resistance of its curing agent are constant. Therefore, the heat transfer model can be simplified to a fixed thermal resistance R2, including the housing thermal resistance, curing agent thermal resistance, and surface encapsulation layer, and a variable thermal resistance R1 due to different fluid types. Please refer to [link / reference]. Figure 3 , Figure 3 This is a schematic diagram of the equivalent temperature measurement model provided in the embodiments of this application. In the equivalent temperature measurement model, a typical housing is made of stainless steel, the curing agent is epoxy resin, and the surface encapsulation layer is polytetrafluoroethylene. Based on the thermal capacity, thermal resistance, and dynamic characteristics of the sensor, the time-domain response characteristics of the temperature sensor can be described by the fluid-varying thermal resistance R1 (due to different fluid types) and the equivalent fixed resistance R2 (including the housing thermal resistance, curing agent thermal resistance, and surface encapsulation layer), thus establishing a temperature steady-state deviation compensation model.

[0102] The time-domain response characteristics of the temperature sensor are as follows:

[0103]

[0104] Mode <1>

[0105] In the formula , representing the convective heat transfer capacity and thermal resistance between the fluid and the shell; R1 represents the thermal resistance related to the fluid type; C1 represents the heat capacity related to the fluid type.

[0106] In the formula R2 represents the equivalent fixed thermal resistance, including the thermal resistance of the housing, the thermal resistance of the curing agent, and the thermal resistance of the surface encapsulation layer, which is a constant value; C2 is the equivalent heat capacity, which is a constant value.

[0107] Cm x is the current mass specific heat capacity of the fluid, K is the structure-related heat flux density mapping coefficient, which is a constant, and Qmass represents the current mass flow rate of the fluid. This indicates the measured value from the temperature sensor; This represents the fluid temperature after reaching a steady state.

[0108] When the time is long enough, i.e. t approaches infinity, the temperature steady-state deviation compensation model is obtained as follows:

[0109] Mode <2> .

[0110] In the formula, The fluid temperature after steady state. The measured value of the temperature sensor is denoted by the heat flux density mapping coefficient associated with the K structure, which is a constant. The specific heat capacity of the fluid is . This represents the current mass flow rate of the fluid.

[0111] Then, calibration group tests were conducted to identify the heat flux density mapping coefficient K and the equivalent fixed thermal resistance R2.

[0112] according to Figure 2 The fluid temperature measurement module shown, through calibration testing, yielded the heat flux density mapping coefficient K and equivalent fixed thermal resistance R2 in the temperature steady-state deviation compensation model. Please refer to [link / reference]. Figure 4 , Figure 4 This is a schematic diagram of the calibration group temperature measurement structure provided in an embodiment of this application. By setting the inlet of the calibration temperature measurement probe below the fluid and sealing it, the temperature sensor is directly immersed in the current safe fluid environment (such as nitrogen, air, or water).

[0113] In the first set of tests, a safe fluid at temperature T1 was introduced. Since the calibrated temperature probe was completely immersed in the fluid, there was no thermal resistance from the housing, curing agent, or surface encapsulation layer. Only the fluid-related convective heat transfer resistance existed. Therefore, the temperature value T_flow_calibation measured by the calibrated temperature probe can be taken as the true temperature value of the current fluid, i.e., T_flow_X = T_flow_calibation. The steady-state fluid temperature T_flow_X_t1, the actual temperature value measured by the temperature sensor T_sensors_t1, and the current mass flow rate Qmass_t1 were recorded.

[0114] Similarly, in the second set of tests, the same safe fluid as in the first set was introduced at a temperature of T2. The steady-state fluid temperature T_flow_X_t2, the actual temperature value measured by the temperature sensor T_sensors_t2, and the current mass flow rate Qmass_t2 were recorded.

[0115] Substitute the parameters T_flow_X_t1, T_sensors_t1, and Qmass_t1 from the first group and the parameters T_flow_X_t2, T_sensors_t2, and Qmass_t2 from the second group into the temperature steady-state deviation compensation model, and input the specific heat capacity parameters of the fluid in the calibration group. By incorporating the convective heat transfer resistance R1 into the temperature steady-state deviation compensation model, the structurally associated heat flux density mapping coefficient K and the equivalent fixed thermal resistance R2 can be determined.

[0116] To improve the identification of the heat flux density mapping coefficient K and the equivalent fixed thermal resistance R2, similar to the first and second groups of tests mentioned above, several groups of steady-state temperature collection tests can be conducted, keeping the same medium fluid but with different fluid temperatures. The heat flux density mapping coefficient K and the equivalent fixed thermal resistance R2 can be calculated using the least squares method, thereby reducing random errors in the system.

[0117] Based on the dynamic response characteristics measured by the temperature sensor, a dynamic prediction and compensation model is established:

[0118] In a thermodynamic system, combining the equivalent thermal resistance R2 and the gas-driven thermal resistance mentioned above, the time-domain response characteristics of a temperature sensor with a multilayer heat transfer medium can be described in the form of Equation 1, or it can be regarded as a transfer function (the product of multiple first-order inertial elements and pure time delay). For example, the nonlinear temperature measurement response system can be described by the Wiener model, decomposing the system into a linear dynamic subsystem and a memoryless nonlinear static subsystem.

[0119] A( x(k)= B( )u(k)

[0120] The formula y(k) = f[x(k)] + ε(k) <3>

[0121] Mode <3> In this context, x(k) represents the internal state or intermediate output of the system.

[0122] u(k) is the system input;

[0123] A ( ) and B ( ) is a polynomial, representing the transfer function of the input and output;

[0124] d is the delay order, which represents the delay effect of the input on the output;

[0125] y(k) is the final output of the system;

[0126] f[x(k)] is a nonlinear function that describes the internal state and the mapping from heat flow Q at temperature T to the final output;

[0127] ξ(k) is the noise or interference term.

[0128] Assuming it's white noise, according to the whitening filter 1 / A( The dynamic prediction and compensation model is determined as follows:

[0129] A( )y_dc(k)=B( )y(k)+ξ(k)

[0130] ξ(k)= A( ) ε(k) <4>

[0131] Mode <4> In this context, ε(k) represents white noise; This is the output value of the dynamic compensator at time k, i.e., the temperature estimate after dynamic hysteresis correction. Indicates assumed observation noise The properties can be derived from the model's own dynamic polynomial A(Z). -1To describe it. Specifically, It acts as a whitening filter, removing relevant colored noise. Convert to uncorrelated white noise .

[0132] The transfer function H(Z) corresponding to the dynamic prediction compensation model -1 )for:

[0133] Mode <5>

[0134] Mode <5> In this context, w represents the order of the input transfer function, and v represents the order of the output transfer function. to for The coefficients in to for The coefficients in.

[0135] At time j, the output y_dc(j) of the dynamic prediction compensation model is:

[0136] Mode <6> .

[0137] in, to These are the autoregressive coefficients. to For the moving average term, to The output value of the dynamic compensator at each past moment. to These are the raw measurements from the temperature sensor at the current time and in the past w-1 time intervals.

[0138] The performance index function mainly consists of four parts: the first is the real-time dynamic error term, the second is the transient overshoot error term, the third is the global overshoot error term, and the fourth is the smoothness constraint term. Among them, the real-time dynamic error term is used to quantify the instantaneous deviation between the compensation output value and the desired temperature value, the transient overshoot error term is used to suppress the instantaneous overshoot during the temperature compensation process, the global overshoot error term is used to suppress the continuous accumulation of small overshoot, and the smoothness constraint term is used to ensure the temporal continuity of the compensation output.

[0139] For example, the performance evaluation function of the dynamic prediction compensation model can be expressed as:

[0140] Mode <7> .

[0141] in, Here, j is the discrete-time index, representing the j-th data point in the temperature sampling sequence, and N is the total number of sampling points. This represents the output value of the dynamic prediction compensation model at the j-th sampling time. This represents the expected temperature value at the j-th sampling time. The overshoot control threshold is a constant, where n and m are exponential constants, m>n, and n>1 is an even number. This is a record of the maximum values ​​during this round of temperature changes. This is to compensate for the average error.

[0142] Average compensation error for:

[0143]

[0144] Collect real-time fluid temperature response data from the calibration group, and calculate the matrix parameters X=[w1,w2,…,w] of the dynamic prediction compensation model using an optimization algorithm. v ,v1,v2,…,v w ].

[0145] Dynamic calibration test group 1: A safe fluid with a temperature of T1 is introduced. Since the calibration temperature probe is completely immersed in the fluid, there is no thermal resistance of the shell, curing agent, or surface encapsulation layer. Only the fluid-related convective heat transfer thermal resistance exists. Therefore, the temperature value T_flow_calibation measured by the calibration temperature probe can be taken as the true temperature value of the current fluid. That is, T_flow_X = T_flow_calibation. Record the temperature sequences of T_flow_calibation(j) and T_sensor(j) as a function of time t.

[0146] The tests for groups 2 to W of the dynamic calibration group refer to the test method of group 1 of the dynamic calibration group. The fluid is the same as that in group 1 of the dynamic calibration group, but the temperature is different. Record the temperature sequences T_flow_calibation(j) and T_sensor(j) as a function of time t.

[0147] The matrix parameters of the dynamic prediction compensation model are X=[w1,w2,…,w v ,v1,v2,…,v w The initial training values ​​are X0 = [0.5, 0, ..., 0, 0.5, 0, ..., 0], that is, the initial values ​​of w1 and v1 are 0.5, and the expected temperature value is y_dc(j) = T_flow_calibation(j).

[0148] The initial values ​​are processed as follows: When j=1, we take y_dc(j)= T_sensor(j), y_dc(j-1)= T_sensor(j),…,y_dc(jv)= T_sensor(j), y(j)= T_sensor(j), y(j-1)= T_sensor(j),…, y(jv)=T_sensor(j). When j>1, y_dc(j) is calculated by Equation 6, y(j)= T_sensor(j).

[0149] Combining the initial value processing methods described above and the performance index evaluation function defined in Equation 7, the matrix parameters X=[w1,w2,…,w] of the dynamic prediction compensation model that matches the current temperature dynamic compensation model can be obtained through constrained optimization algorithms, including but not limited to genetic training algorithms, neural network training algorithms, augmented multiplier methods, and penalty function methods. v ,v1,v2,…,v w ].

[0150] Please refer to Figure 5 , Figure 5 This is a schematic diagram illustrating the hysteresis and deviation compensation system of the temperature measurement system in non-direct contact with fluid provided in this application embodiment. The complete temperature sensor hysteresis and deviation compensation method is implemented as follows:

[0151] The dynamic compensation algorithm module, also known as the dynamic compensator in Equation 6, is used to correct the response hysteresis of the temperature sensor. The preceding Delay module, also known as the unit delay module, has an initial value of -273. When the dynamic compensation algorithm module is run for the first time, T_flag = -273, so refer to the initial value processing flow described above:

[0152] When T_flag=-273, i.e., j=1, in the first run, set T_dc_w(j)=T_sensor(j), T_dc_w(j-1)=T_sensor(j),…,T_dc_w(jv)=T_sensor(j), T_k_v(j)=T_sensor(j), T_k_v(j-1)=T_sensor(j),…,T_k_v(jv)=T_sensor(j), when… The output of the dynamic compensator at the previous moment is T_dc_now=y_dc(1), that is, the current output value of the dynamic compensator is placed in the first element of the T_dc_w array, and the output value of the dynamic compensator at v moments ago is placed in the vth element of the T_dc_w array, that is, at the end of the T_dc_w array. Similarly, the temperature sensor reading value at the current moment is placed in the first element of the T_k_v array, and the temperature sensor reading value at w moments ago, T_sensor(jw), is placed in the wth element, that is, at the end of the T_k_v array.

[0153] When j>1, i.e. T_flag≠-273, the current temperature sensor reading value array is updated in a rolling manner. T_k_v1(2:end)=T_k_v(1:end-1), T_k_v1(1)=T_sensor, where end represents the last element of the array, T_dc_w(1) is calculated based on the output of the dynamic compensator, and the dynamic compensator output value array T_dc_v(2:end)=T_k_v1(1:end-1), T_k_v(1)=T_dc_now is updated.

[0154] The steady-state judgment module primarily determines whether the temperature is close to steady state and whether the steady-state deviation compensation algorithm module should be activated. It obtains the dynamic compensation output value T_dc_now_1 from the previous moment using a unit delay circuit and calculates the temperature change. When the temperature change ΔT < , If the value is a very small constant (typically [0.1, 0.5]), then the temperature is considered to be approaching steady state, and the steady-state deviation compensation module is activated: T_compensate = T_dc_now + T_bias; when the temperature change ΔT > If the temperature has not approached steady state, then the steady-state deviation compensation value T_bias=0, that is, the temperature compensation output value of this method T_compensate=T_dc_now+0.

[0155] The output value T_bias of the steady-state deviation compensation algorithm is mainly calculated by the temperature steady-state deviation compensation model. When it is necessary to measure a fluid different from the calibration group, the specific heat capacity Cm of the current fluid is substituted. x The convective heat transfer resistance R1 (note: the reciprocal of the fluid convective heat transfer coefficient) is given by the formula, where Q is the thermal resistance. mass The fluid mass flow rate can be measured by mass flow meters such as Coriolis flow meters in the temperature measurement loop. The steady-state T_dc_now is taken as a value that approximates the true temperature of the fluid, i.e., T_flow_x = T_dc_now. Substituting this value into the equation, the output value T_bias of the steady-state deviation compensation module can be calculated. Thus, the temperature compensation output value of the fluid in the current condition, which is different from that of the calibration group, can be obtained by T_compensate = T_dc_now + T_bias.

[0156] The effects of the present invention will be further explained in detail below with reference to examples:

[0157] based on Figure 4 The calibration group temperature measurement structure shown introduces a high-purity nitrogen gas flow as the training fluid. The inlet is upstream, and a gas heater heats the high-purity nitrogen gas flow at an ambient temperature of 20°C. Nitrogen gas flows at 35°C and 50°C are introduced respectively. Based on Matheson gas data, the specific heat capacity of nitrogen is obtained. Based on the convective heat transfer coefficient, the structurally related heat flux density mapping coefficient K=0.67 and the equivalent fixed thermal resistance R2=3.81 were identified.

[0158] Real-time fluid temperature response data were collected from the calibration group of the dynamic compensation algorithm module. The temperature sensor used in the experiment had a data acquisition frequency of 10Hz. The temperature of the high-purity nitrogen fluid in the first calibration group was 27℃, and the temperature of the high-purity nitrogen fluid in the second group was 42℃. According to the performance index evaluation function of the dynamic prediction compensation model in step 4, the matrix parameter matrix X of the dynamic prediction compensation model was optimized using the hybrid penalty function method. The initial value of the matrix parameter of the dynamic prediction compensation model was taken as X0=[0.5,0,0,0,0,0.5,0,0,0,0], the length of the T_dc_w array was set to 5, and the length of the T_k_v array was set to 5. =0.5, n=2, m=3.

[0159] The parameter matrix calculated by the hybrid penalty function optimization algorithm is as follows:

[0160] X=[0.011032737721519,0.090533240680109,0.101358838742657,0.004057188692099,0.992123161948395, 4.530370208775528,3.041755645795530,0.226689986419032,-3.022179767492920,-4.772715480161359].

[0161] The parameters obtained from the above experiments were substituted into the temperature sensor hysteresis and deviation compensation method proposed in this invention, and two verification group tests were conducted. Figure 4 The calibration temperature measuring structure shown was tested by passing high-purity nitrogen streams at 27°C and 42°C respectively, and the results were as follows. Figure 6 and Figure 7 The results are shown. Among them, Figure 6 This is a schematic diagram of the results of the high-purity nitrogen gas flow test group at 27°C provided in the embodiments of this application. Figure 7 This is a schematic diagram of the results of the high-purity nitrogen gas test group at 42°C provided in the embodiments of this application.

[0162] The results from the high-purity nitrogen gas testing group show that the solution provided in this application can effectively solve the hysteresis problem in the temperature measurement process and effectively reduce the steady-state deviation value.

[0163] Based on the above testing methods, high-purity carbon dioxide fluid was tested, and the specific heat capacity of carbon dioxide was obtained based on Matheson gas data. The convective heat transfer coefficient was input into the steady-state deviation compensation algorithm module, and carbon dioxide was heated to 32℃. The results are as follows: Figure 8 As shown. Please refer to... Figure 8 , Figure 8 This is a schematic diagram of the results of the high-purity carbon dioxide 32°C test group provided in the embodiments of this application.

[0164] from Figure 8 It can be seen that after switching to a fluid different from the calibration group, the temperature measurement system based on the non-direct contact fluid temperature measurement system hysteresis and deviation compensation system compensates for the temperature measurement value, which can effectively compensate for the hysteresis, avoid overshoot, and reduce the temperature difference of the fluid.

[0165] In summary, in the hysteresis and deviation compensation system of the non-contact fluid temperature measurement system provided in this application embodiment, the dynamic compensation module reduces steady-state deviation during transient processes, improving the system's tracking capability during temperature changes; while the steady-state deviation compensation module specifically corrects for hysteresis effects, reducing measurement errors in steady state. By combining the dual mechanisms of dynamic compensation and steady-state deviation compensation, the accuracy and response characteristics of non-contact fluid temperature measurement can be effectively improved, thereby enhancing the adaptability and reliability of the temperature sensor under non-contact conditions.

[0166] Based on the same concept, this application also provides a method for compensating for hysteresis and deviation in a temperature measurement system that does not directly contact the fluid, which can be implemented using the temperature sensor described above. The method includes the following steps:

[0167] A temperature steady-state deviation compensation model is established for steady-state deviation compensation; a dynamic predictive compensation model is established for hysteresis compensation; the dynamic predictive compensation model performs hysteresis compensation based on the measured values ​​of the fluid in non-direct contact and obtains the hysteresis-compensated temperature value; based on the hysteresis-compensated temperature value at the current moment and the hysteresis-compensated temperature value at the previous moment, it is determined whether the fluid is in a steady state. If so, the temperature steady-state deviation compensation model outputs the hysteresis-compensated temperature value after steady-state compensation; otherwise, the hysteresis-compensated temperature value is output.

[0168] In some embodiments, the establishment of the temperature steady-state deviation compensation model includes the following steps:

[0169] A temperature sensor is mounted on the outer wall of the housing; a surface encapsulation layer and a curing agent are provided between the temperature sensor and the housing, and fluid is introduced into the interior of the housing;

[0170] The variable thermal resistance caused by different fluid types is equivalent to the variable thermal resistance R1; the thermal resistance of the shell, the thermal resistance of the curing agent, and the thermal resistance of the surface encapsulation layer are equivalent to the fixed thermal resistance R2.

[0171] The time-domain response characteristics of the temperature sensor are:

[0172] Mode <1>

[0173] In the formula , representing the convective heat transfer capacity and thermal resistance between the fluid and the shell; C1 represents the thermal resistance related to the fluid type; C2 represents the heat capacity related to the fluid type.

[0174] In the formula R2 represents the equivalent fixed thermal resistance, including the thermal resistance of the housing, the thermal resistance of the curing agent, and the thermal resistance of the surface encapsulation layer, which is a constant value; C2 is the equivalent heat capacity, which is a constant value.

[0175] Cm x is the current mass specific heat capacity of the fluid, K is the structure-related heat flux density mapping coefficient, which is a constant, and Qmass represents the current mass flow rate of the fluid. This indicates the measured value from the temperature sensor; This represents the fluid temperature after reaching a steady state.

[0176] When the time is long enough, i.e. t approaches infinity, the temperature steady-state deviation compensation model is obtained as follows:

[0177] Mode <2> .

[0178] In some embodiments, the method may further include:

[0179] A calibration temperature probe is installed on the inner wall of the housing and sealed. Then, a calibration fluid is introduced, and the calibration temperature probe is immersed in the current calibration fluid environment. The calibration fluid is a non-toxic, non-corrosive gas.

[0180] Introduce the first set of calibration fluid at temperature T1, and record the measured values ​​of the calibration temperature probe, the temperature sensor, and the current mass flow rate.

[0181] A second set of calibration fluids at a temperature of T2 was introduced, and the measured values ​​of the calibration temperature probe, the temperature sensor, and the current mass flow rate were recorded. The second set of calibration fluids was of the same type as the first set of calibration fluids, but at a different temperature.

[0182] The measurements from the calibration temperature probe, temperature sensor, and current mass flow rate when the first set of calibration fluids is introduced, and the measurements from the calibration temperature probe, temperature sensor, and current mass flow rate when the second set of calibration fluids is introduced, along with the specific heat capacity parameter Cm of the calibration fluids, are recorded. x And the convective heat transfer resistance R1, substituting the type <2> In the process, the heat flux density mapping coefficient K and the equivalent fixed thermal resistance R2 associated with the structure are obtained by solving, and then the temperature steady-state deviation compensation model is updated.

[0183] In some embodiments, the establishment of a dynamic prediction compensation model includes the following steps:

[0184] Based on the equivalent fixed thermal resistance R2 and the variable thermal resistance R1 of the calibration fluid, the dynamic response characteristics of the temperature sensor of the multilayer heat transfer medium are described as a nonlinear temperature measurement response system, and the system is decomposed into a linear dynamic subsystem and a memoryless nonlinear subsystem:

[0185] A( x(k)= B( )u(k)

[0186] The formula y(k) = f[x(k)] + ξ(k) <3>

[0187] Mode <3> In this context, x(k) represents the internal state or intermediate output of the system.

[0188] u(k) is the system input;

[0189] A ( ) and B ( ) is a polynomial, representing the transfer function of the input and output;

[0190] d is the delay order, which represents the delay effect of the input on the output;

[0191] y(k) is the final output of the system;

[0192] f[x(k)] is a nonlinear function that describes the internal state and the mapping from heat flow Q at temperature T to the final output;

[0193] ξ(k) is the noise or interference term.

[0194] Assuming it's white noise, according to the whitening filter 1 / A( The dynamic prediction and compensation model is designed as follows:

[0195] A( )y_dc(k)=B( )y(k)+ξ(k)

[0196] ξ(k)= A( ) ε(k) <4>

[0197] Mode <4> In this context, ε(k) represents white noise;

[0198] The transfer function H(Z) corresponding to the dynamic prediction compensation model -1 )for:

[0199] Mode <5>

[0200] Mode <5> In this context, w represents the order of the input transfer function, and v represents the order of the output transfer function.

[0201] At time j, the output y_dc(j) of the dynamic prediction compensation model is:

[0202] Mode <6> .

[0203] In some embodiments, the method may further include:

[0204] A performance evaluation function for the dynamic prediction and compensation model, including overshoot constraints, is constructed. The performance evaluation function is expressed as follows:

[0205] Mode <7>

[0206] For real-time dynamic error terms, This is the transient overshoot error term. This is the global overshoot error term. For smoothness constraints;

[0207] Mode <7> In this context, T_osc is the critical threshold of the overshoot control function and is a constant; m and n are constants, where m>n and n>1 are even numbers. d (j)_max is the maximum value of the temperature sequence in this round;

[0208] Err_Avg The average compensation error is expressed as follows:

[0209] Mode <8> .

[0210] In some embodiments, the method may further include:

[0211] A temperature sensor is installed on the outer wall of the housing; a surface encapsulation layer and a curing agent are provided between the temperature sensor and the housing; a calibration temperature probe is installed on the inner wall of the housing and sealed, then a calibration fluid is introduced, and the calibration temperature probe is immersed in the current calibration fluid environment; a first group of calibration fluids, a second group of calibration fluids, ..., a Wth group of calibration fluids with different temperatures are sequentially introduced into the housing; the first group of calibration fluids, the second group of calibration fluids, ..., the Wth group of calibration fluids have different temperatures but the same fluid type, and the measured values ​​of the calibration temperature probe and the temperature sensor for the first group to the Wth group of calibration fluids are recorded respectively; a sequence of measured values ​​of the calibration temperature probe and the temperature sensor changing over time is formed; the measured value of the calibration temperature probe is taken as the expected temperature value, the matrix parameters of the dynamic prediction compensation model are obtained, and the dynamic prediction compensation model is updated; a non-calibrated fluid is introduced, the measured value is obtained through the temperature sensor, and the first compensation value of the non-calibrated fluid is calculated based on the dynamic prediction compensation model; the non-calibrated fluid is a toxic and highly corrosive fluid.

[0212] The steady-state deviation compensation algorithm module in the temperature sensor hysteresis and deviation compensation method proposed in this application is based on the principle of heat transfer model. It takes into account the different convective heat transfer coefficients and thermal resistances of different fluids, effectively compensating for the steady-state deviation of temperature measurement. It solves the problem that it is impossible to effectively measure non-calibrated fluids in real time when switching medium fluids. The temperature sensor hysteresis and deviation compensation method proposed in this application has wide applicability when switching fluids.

[0213] It should be understood that when the various modules of the system provided in the above embodiments are working, the division of each functional module in the above description is only used as an example. In actual applications, the above functions can be assigned to different functional modules as needed. That is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0214] The functional modules in the above embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of the embodiments of this application.

[0215] Based on the same concept, embodiments of this application also provide a method for compensating for hysteresis and deviation in a temperature measurement system that does not directly contact the fluid, which may include:

[0216] The dynamic compensation module receives the temperature value of the non-directly contacting fluid collected by the temperature sensor and performs hysteresis compensation on the temperature value to obtain the hysteresis-compensated temperature value. The steady-state judgment module receives and stores the hysteresis-compensated temperature value, and then calculates the temperature change based on the current hysteresis-compensated temperature value and the previous hysteresis-compensated temperature value. It compares the temperature change with a set threshold. If the temperature change is less than or equal to the set threshold, the fluid is in a steady state. The steady-state deviation compensation module performs steady-state compensation on the current hysteresis-compensated temperature value and outputs it. If the temperature change is greater than the set threshold, the fluid is not in a steady state, and the current hysteresis-compensated temperature value is output.

[0217] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A temperature measurement system that does not directly contact a fluid, characterized in that, include: Dynamic compensation module, steady-state deviation compensation module, steady-state judgment module; The temperature sensor collects the temperature value of the fluid that is not in direct contact with the fluid. The temperature value is then hysteresis compensated by the dynamic compensation module. The steady-state judgment module determines whether the temperature value is in a steady state based on the hysteresis compensated temperature value. If not, the hysteresis compensated temperature value is output. If so, the steady-state deviation compensation module performs steady-state deviation compensation on the hysteresis compensated temperature value and outputs it. The steady-state deviation compensation module includes a temperature steady-state deviation compensation model; the temperature steady-state deviation compensation model is established based on the time-domain response characteristics of the temperature sensor. The time-domain response characteristics of the temperature sensor are: Mode <1> In the formula , representing the convective heat transfer capacity and thermal resistance between the fluid and the shell; C1 represents the thermal resistance related to the fluid type; C2 represents the heat capacity related to the fluid type. In the formula R2 represents the equivalent fixed thermal resistance, including the thermal resistance of the housing, the thermal resistance of the curing agent, and the thermal resistance of the surface encapsulation layer, which is a constant value; C2 is the equivalent heat capacity, which is a constant value. Cm x is the current mass specific heat capacity of the fluid, K is the structure-related heat flux density mapping coefficient, which is a constant, and Qmass represents the current mass flow rate of the fluid. This indicates the measured value from the temperature sensor; This represents the fluid temperature after reaching a steady state. When the time is long enough, i.e. t approaches infinity, the temperature steady-state deviation compensation model is obtained as follows: Mode <2> .

2. The temperature measurement system for non-direct contact with fluid according to claim 1, characterized in that, The steady-state judgment module includes a unit delay unit; the steady-state judgment module receives the temperature value after hysteresis compensation, and the unit delay unit records the temperature value after hysteresis compensation at the previous moment. The temperature change is calculated based on the hysteresis-compensated temperature value at the current moment and the hysteresis-compensated temperature value at the previous moment. The temperature change is compared with a set threshold. If the temperature change is less than or equal to the set threshold, the fluid is in a steady state. If the temperature change is greater than the set threshold, the fluid is not in a steady state.

3. The temperature measurement system for non-direct contact fluid as described in claim 1, characterized in that, The dynamic compensation module includes a dynamic prediction compensation model, which is established based on the dynamic response characteristics of the temperature sensor. Based on the known equivalent fixed thermal resistance R2 and variable thermal resistance R1, the dynamic response characteristics of the temperature sensor for the multilayer heat transfer medium are described as a nonlinear temperature measurement response system, which is then decomposed into a linear dynamic subsystem and a memoryless nonlinear subsystem. A( )x(k)= B( )u(k) The formula y(k) = f[x(k)] + ε(k) <3> Mode <3> In this context, x(k) represents the internal state or intermediate output of the system. u(k) is the system input; A ( ) and B ( ) is a polynomial, representing the transfer function of the input and output; d is the delay order, which represents the delay effect of the input on the output; y(k) is the final output of the system; f[x(k)] is a nonlinear function that describes the internal state and the mapping from heat flow Q at temperature T to the final output; ξ(k) is the noise or interference term. Assuming it's white noise, according to the whitening filter 1 / A( The dynamic prediction and compensation model is determined as follows: TO( )y_dc(k)=B( )y(k)+ξ(k) ξ(k)= A( ) ε(k)expression <4> Mode <4> In this context, ε(k) represents white noise; The transfer function H(Z) corresponding to the dynamic prediction compensation model -1 )for: Mode <5> Mode <5> In this context, w represents the order of the input transfer function, and v represents the order of the output transfer function. At time j, the output y_dc(j) of the dynamic prediction compensation model is: Mode <6> .

4. A compensation method for a temperature measurement system that does not directly contact a fluid, implemented using the temperature measurement system described in any one of claims 1-3, characterized in that, Includes the following steps: Establish a temperature steady-state deviation compensation model for steady-state deviation compensation; Establish a dynamic predictive compensation model for hysteresis compensation; The dynamic prediction compensation model performs hysteresis compensation based on the detected temperature value of the non-directly contacting fluid to obtain the hysteresis-compensated temperature value. Based on the current hysteresis-compensated temperature value and the previous hysteresis-compensated temperature value, it is determined whether the fluid is in a steady state. If so, the temperature steady-state deviation compensation model performs steady-state compensation on the hysteresis-compensated temperature value and outputs it. If not, output the temperature value after hysteresis compensation; The establishment of the temperature steady-state deviation compensation model includes the following steps: A temperature sensor is mounted on the outer wall of the housing; a surface encapsulation layer and a curing agent are provided between the temperature sensor and the housing, and fluid is introduced into the interior of the housing; The variable thermal resistance caused by different fluid types is equivalent to the variable thermal resistance R1; the thermal resistance of the shell, the thermal resistance of the curing agent, and the thermal resistance of the surface encapsulation layer are equivalent to the fixed thermal resistance R2. The time-domain response characteristics of the temperature sensor are: Mode <1> In the formula , representing the convective heat transfer capacity and thermal resistance between the fluid and the shell; C1 represents the thermal resistance related to the fluid type; C2 represents the heat capacity related to the fluid type. In the formula R2 represents the equivalent fixed thermal resistance, including the thermal resistance of the housing, the thermal resistance of the curing agent, and the thermal resistance of the surface encapsulation layer, which is a constant value; C2 is the equivalent heat capacity, which is a constant value. Cm x is the current mass specific heat capacity of the fluid, K is the structure-related heat flux density mapping coefficient, which is a constant, and Qmass represents the current mass flow rate of the fluid. This indicates the measured value from the temperature sensor; This represents the fluid temperature after reaching a steady state. When the time is long enough, i.e. t approaches infinity, the temperature steady-state deviation compensation model is obtained as follows: Mode <2> .

5. The temperature measurement system compensation method according to claim 4, characterized in that, It also includes the following steps: A calibration temperature probe is installed on the inner wall of the housing and sealed. Then, a calibration fluid is introduced, and the calibration temperature probe is immersed in the current calibration fluid environment. The calibration fluid is a non-toxic, non-corrosive gas. Introduce the first set of calibration fluid at temperature T1, and record the measured values ​​of the calibration temperature probe, the temperature sensor, and the current mass flow rate. A second set of calibration fluids at a temperature of T2 was introduced, and the measured values ​​of the calibration temperature probe, the temperature sensor, and the current mass flow rate were recorded. The second set of calibration fluids was of the same type as the first set of calibration fluids, but at a different temperature. The measurements from the calibration temperature probe, temperature sensor, and current mass flow rate when the first set of calibration fluids is introduced, and the measurements from the calibration temperature probe, temperature sensor, and current mass flow rate when the second set of calibration fluids is introduced, along with the specific heat capacity parameter Cm of the calibration fluids, are recorded. x And variable thermal resistance R1, substituting into the formula <2> In the process, the heat flux density mapping coefficient K and the equivalent fixed thermal resistance R2 associated with the structure are obtained by solving, and then the temperature steady-state deviation compensation model is updated.

6. The temperature measurement system compensation method according to claim 5, characterized in that, The establishment of the dynamic prediction and compensation model includes the following steps: Based on the equivalent fixed thermal resistance R2 and the variable thermal resistance R1, the dynamic response characteristics of the temperature sensor of the multilayer heat transfer medium are described as a nonlinear temperature measurement response system, and the system is decomposed into a linear dynamic subsystem and a memoryless nonlinear subsystem: A( )x(k)= B( )u(k) The formula y(k) = f[x(k)] + ξ(k) <3> Mode <3> In this context, x(k) represents the internal state or intermediate output of the system. u(k) is the system input; A ( ) and B ( ) is a polynomial, representing the transfer function of the input and output; d is the delay order, which represents the delay effect of the input on the output; y(k) is the final output of the system; f[x(k)] is a nonlinear function that describes the internal state and the mapping from heat flow Q at temperature T to the final output; ξ(k) is the noise or interference term. Assuming it's white noise, according to the whitening filter 1 / A( The dynamic prediction and compensation model is designed as follows: TO( )y_dc(k)=B( )y(k)+ξ(k) ξ(k)= A( ) ε(k)expression <4> Mode <4> In this context, ε(k) represents white noise; The transfer function H(Z) corresponding to the dynamic prediction compensation model -1 )for: Mode <5> Mode <5> In this context, w represents the order of the input transfer function, and v represents the order of the output transfer function. At time j, the output y_dc(j) of the dynamic prediction compensation model is: Mode <6> .

7. The temperature measurement system compensation method according to claim 6, characterized in that, It also includes the following steps: A performance evaluation function for the dynamic prediction and compensation model, including overshoot constraints, is constructed. The performance evaluation function is expressed as follows: Mode <7> For real-time dynamic error terms, This is the transient overshoot error term. This is the global overshoot error term. For smoothness constraints; Mode <7> In this context, T_osc is the critical threshold of the overshoot control function and is a constant; m and n are constants, where m>n and n>1 are even numbers. d (j)_max is the maximum value of the temperature sequence in this round; Err_Avg The average compensation error is expressed as follows: Mode <8> .

8. The temperature measurement system compensation method according to claim 7, characterized in that, It also includes the following steps: A temperature sensor is installed on the outer wall of the housing; a surface encapsulation layer and a curing agent are provided between the temperature sensor and the housing; a calibration temperature probe is installed on the inner wall of the housing and sealed, then a calibration fluid is introduced, and the calibration temperature probe is immersed in the current calibration fluid environment; Inside the housing, calibration fluids of different temperatures are sequentially introduced: a first group, a second group, ..., a Wth group of calibration fluids. The temperatures of the first group, the second group, ..., the Wth group of calibration fluids are different, but the fluid types are the same. The measured values ​​of the calibration temperature probe and the temperature sensor for the first to the Wth groups of calibration fluids are recorded respectively, forming a sequence of measured values ​​of the calibration temperature probe and the temperature sensor that change over time. The measured value of the calibrated temperature probe is taken as the expected temperature value, the matrix parameters of the dynamic prediction compensation model are obtained, and the dynamic prediction compensation model is updated. A non-calibrated fluid is introduced, and a temperature sensor is used to obtain a measurement value. A first compensation value for the non-calibrated fluid is calculated based on a dynamic prediction compensation model. The non-calibrated fluid is a toxic and highly corrosive fluid.

Images