CN118565661A - Temperature sensor calibration method based on true value regression - Google Patents
Temperature sensor calibration method based on true value regression Download PDFInfo
- Publication number
- CN118565661A CN118565661A CN202410796400.1A CN202410796400A CN118565661A CN 118565661 A CN118565661 A CN 118565661A CN 202410796400 A CN202410796400 A CN 202410796400A CN 118565661 A CN118565661 A CN 118565661A
- Authority
- CN
- China
- Prior art keywords
- temperature
- temperature sensor
- correction
- value
- inconsistency
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000012937 correction Methods 0.000 claims abstract description 85
- 238000005259 measurement Methods 0.000 claims abstract description 51
- 238000005457 optimization Methods 0.000 claims abstract description 28
- 230000006870 function Effects 0.000 claims description 78
- 239000012530 fluid Substances 0.000 claims description 24
- 238000009529 body temperature measurement Methods 0.000 claims description 21
- 238000012216 screening Methods 0.000 claims description 7
- 230000002068 genetic effect Effects 0.000 claims description 3
- 238000011478 gradient descent method Methods 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 39
- 239000000498 cooling water Substances 0.000 description 28
- 230000006872 improvement Effects 0.000 description 9
- 230000008859 change Effects 0.000 description 4
- 238000005057 refrigeration Methods 0.000 description 4
- 238000001816 cooling Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000004321 preservation Methods 0.000 description 3
- 238000004378 air conditioning Methods 0.000 description 2
- 238000005265 energy consumption Methods 0.000 description 2
- 238000007710 freezing Methods 0.000 description 2
- 230000008014 freezing Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 230000008020 evaporation Effects 0.000 description 1
- 238000001704 evaporation Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005293 physical law Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 239000000523 sample Substances 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
Landscapes
- Measuring Temperature Or Quantity Of Heat (AREA)
Abstract
The invention provides a temperature sensor calibration method based on true value regression, which comprises the following steps: step 10, constructing a temperature inconsistency objective function according to the temperature consistency logic association relation; step 20, constructing a correction function for each temperature sensor; step 30, collecting historical measurement data of all temperature sensors in a preset time period, and establishing a steady-state measurement data set; step 40, based on a steady-state measurement data set, taking a temperature inconsistency objective function as an optimization target, optimizing the correction function of each temperature sensor until the optimization cutoff condition is reached, and stopping optimization to obtain an optimized correction function; and step 50, taking the temperature correction value output by the correction function of each optimized temperature sensor as calibrated temperature data. The temperature sensor calibration method based on true value regression provided by the invention realizes the calibration of the temperature sensor, and is efficient and reliable.
Description
Technical Field
The invention belongs to the technical field of sensor calibration, and particularly relates to a temperature sensor calibration method based on true value regression.
Background
In most central air-conditioning cold source systems, the number of temperature sensors is relatively large, and the operation reliability is critical to ensuring the safety and energy-saving operation of the cold source. Due to the severe working environment of the cold source system, various faults with different degrees such as probe surface scaling, component aging, performance drift and the like of the temperature sensor can be caused, and the faults can influence temperature measurement data, so that the control effect of the cold source system is influenced, and the problems of increased energy consumption, reduced indoor thermal comfort or operation reliability and the like of the system are caused. The deviation fault is one of the most main fault types of the faults of the temperature sensor of the cold source system, and seriously affects the actual operation efficiency of the system, for example, when the freezing backwater temperature sensor generates negative deviation, the evaporation temperature of the water chilling unit can be directly reduced, the energy consumption of the unit can be increased, and the performance of the unit can be possibly deteriorated.
In practical application, the sensor is difficult to calibrate regularly, and over time, components and parts are aged to easily cause deviation faults, so that the calibration of the deviation faults is a precondition of energy-saving and reliable operation of a cold source system. However, there are certain limitations to either the traditional physical calibration method or the currently widely used sensor self-calibration method based on machine learning algorithm: (1) The physical calibration method is time-consuming and labor-consuming, can interrupt the normal operation of the system, and has difficulty in removal and installation; (2) Self-calibration methods based on machine learning algorithms either require accurate mathematical models or are highly dependent on data acquired by different types of sensors. These problems make calibration of the temperature sensor a challenge and difficulty in practical applications.
Disclosure of Invention
The invention aims to solve the technical problems that: the temperature sensor calibration method based on true value regression is high-efficiency and reliable in calibration of the temperature sensor.
In order to solve the technical problems, the embodiment of the invention adopts the following technical scheme:
the embodiment of the invention provides a temperature sensor calibration method based on true value regression, which comprises the following steps:
step 10, constructing a temperature inconsistency objective function according to the temperature consistency logic association relation;
step 20, constructing a correction function for each temperature sensor;
step 30, collecting historical measurement data of all temperature sensors in a preset time period, and establishing a steady-state measurement data set;
Step 40, based on a steady-state measurement data set, taking a temperature inconsistency objective function as an optimization target, optimizing the correction function of each temperature sensor until the optimization cutoff condition is reached, and stopping optimization to obtain an optimized correction function;
And step 50, taking the temperature correction value output by the correction function of each optimized temperature sensor as calibrated temperature data.
As a further improvement of the embodiment of the invention, the temperature consistency logic association relationship means that the fluid temperature at all positions in the same pipeline is equal or/and the fluid temperature in the main pipeline is equal to the fluid temperature in all branch pipelines or/and the fluid temperature in all branch pipelines under the condition that the pipeline fluid is fully developed and the pipeline fluid is stable.
As a further improvement of the embodiment of the present invention, the temperature inconsistency objective function is formula (1):
Where U T′ denotes a temperature inconsistency value, T i denotes a temperature measurement value of an ith temperature sensor, The average value of the temperature correction values of all the temperature sensors is represented, and n represents the total number of the temperature sensors.
As a further improvement of the embodiment of the present invention, the correction function is formula (2):
t i′＝aiTi+bi type (2)
Where T i' represents the temperature correction value of the i-th temperature sensor, a i represents the linear drift correction parameter of the i-th temperature sensor, T i represents the temperature measurement value of the i-th temperature sensor, and b i represents the fixed deviation correction parameter of the i-th temperature sensor.
As a further improvement of the embodiment of the present invention, the step 40 specifically includes:
step 401, calculating a temperature inconsistency value of each moment measured value in a steady-state data set by using a temperature inconsistency objective function, and screening typical measured data from the steady-state measured data set;
Step 402, initializing a correction function of each temperature sensor, and calculating a temperature correction value of each temperature sensor at each moment by using the correction function of each temperature sensor based on typical measurement data; calculating to obtain a temperature inconsistency value of the temperature correction value at each moment by using a temperature inconsistency objective function;
And step 403, taking the temperature inconsistency objective function as an optimization target, optimizing the correction function of each temperature sensor until the optimization cut-off condition is reached, and stopping optimizing to obtain the optimized correction function.
As a further improvement of the embodiment of the present invention, the step 401 specifically includes:
Calculating a measurement inconsistency value of all temperature measurement values at each moment in the steady-state measurement data set by using the formula (3):
Where U T denotes a measurement inconsistency value, T i denotes a temperature measurement value of an ith temperature sensor, Representing the average value of the temperature measurement values of all the temperature sensors, and n represents the total number of the temperature sensors;
Drawing an inconsistency histogram, determining a distribution model according to the distribution shape of the histogram, and fitting a probability density function;
and performing simulated sampling by using the fitted probability density function, calculating a 95% confidence interval of the data, and screening out the data falling in the 95% confidence interval from the steady-state measurement data set to form typical measurement data.
As a further improvement of the embodiment of the present invention, in the step 403, the temperature inconsistency objective function is used as an optimization target, and a gradient descent method is adopted to iteratively update the correction parameters of the correction function of each temperature sensor, so as to obtain the optimized correction function.
As a further improvement of the embodiment of the present invention, in the step 403, the temperature inconsistency objective function is used as an optimization target, and a genetic algorithm is used to iteratively update the correction parameters of the correction function of each temperature sensor, so as to obtain the optimized correction function.
As a further improvement of the embodiment of the invention, the optimized cut-off condition comprises a temperature inconsistency value being less than or equal to a preset threshold value.
As a further improvement of the embodiment of the invention, the preset threshold value is 0.05-0.1.
Compared with the prior art, the technical scheme of the invention has the following beneficial effects:
According to the temperature sensor calibration method based on true value regression, firstly, according to a temperature consistency logic association relationship, temperature inconsistency objective functions of all temperature sensors to be calibrated are constructed; then, according to the deviation relation between the measured value and the temperature true value, establishing a correction function of each temperature sensor to be calibrated; finally, based on the measurement data of all the temperature sensors to be calibrated, a steady-state measurement data set is established, a temperature inconsistency objective function is used as an optimization target, and the correction function of each temperature sensor is optimized to obtain an optimized correction function; and taking the temperature correction value output by the correction function of each optimized temperature sensor as calibrated temperature data. Therefore, the calibration of the temperature sensor is realized, and the influence of sensor faults on fault diagnosis and operation optimization strategies of the thermal equipment is avoided.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments of the present invention will be briefly described below, and it is obvious that the drawings described below are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort to a person of ordinary skill in the art.
FIG. 1 is a flow chart of a method of an embodiment of the present invention;
FIG. 2 is a schematic diagram of a calibration application scenario common to the method of the present invention;
Fig. 3 is a comparison of the temperature sensor calibration before and after the calibration in an embodiment of the present invention.
Detailed Description
The technical scheme of the invention is described in detail below with reference to the accompanying drawings.
The embodiment of the invention provides a temperature sensor calibration method based on true value regression, which is shown in fig. 1 and comprises the following steps:
And step 10, constructing a temperature inconsistency objective function according to the temperature consistency logic association relation.
A correction function is built for each temperature sensor, step 20.
And step 30, collecting historical measurement data of all temperature sensors in a preset time period, and establishing a steady-state measurement data set.
And step 40, based on the steady-state measurement data set, taking the temperature inconsistency objective function as an optimization target, optimizing the correction function of each temperature sensor until the optimization cut-off condition is reached, and stopping optimizing to obtain the optimized correction function.
And step 50, taking the temperature correction value output by the correction function of each optimized temperature sensor as calibrated temperature data.
Specifically, in step 10, the temperature consistency logic association relationship means that the fluid temperatures at all positions in the same pipeline are equal under the condition that the pipeline fluid is fully developed and the pipeline fluid is stable, or the fluid temperatures in the main pipeline and the fluid temperatures in all branch pipelines are equal, or the fluid temperatures in all branch pipelines are equal. The temperature consistency logic association relations of different calibration application scenes are different.
The application scene is that a section of fluid main pipeline with good heat preservation and no branch pipeline is provided with a plurality of temperature sensors at positions. The plurality of temperature sensors are respectively Rt 1、Rt2、Rt3、……、Rtn, and the measured value of the temperature measured at the same time is T 1、T2、T3、……、Tn. Since the fluid in the same pipeline is fully mixed, the temperature is not changed, and in theory, when the temperature sensors have no errors, the temperature values of all the temperature sensors should be equal. When the temperature sensors are calibrated, the fluid temperature at each position in the fluid main pipeline is equal according to the objectively existing temperature consistency logic relationship, namely the true values of the n temperature sensors are completely equal.
The built temperature inconsistency objective function is (1):
Where U T′ denotes a temperature inconsistency value, T i denotes a temperature measurement value of an ith temperature sensor, The average value of the temperature correction values of all the temperature sensors is represented, and n represents the total number of the temperature sensors. Theoretically, when the temperature sensors are error-free, the temperature inconsistency value of all the temperature sensors is 0.
Another application scenario is shown in fig. 2, which is a section of fluid main pipeline with good heat preservation, and has m branch pipelines. The temperature sensors on the fluid main pipeline are Rt, and the temperature sensors on the m branch pipelines are Rt 1、Rt2、Rt3、……、Rtm respectively.
Fig. 2 may be a schematic layout diagram of a cooling water return water temperature sensor in a high-efficiency refrigeration machine room, and the method of the present invention may be used for calibration of the cooling water return water temperature sensor. The temperature sensor of the cooling water return water main pipe is Rt, the temperature sensors of all the cooling water return water branch pipes are Rt 1、Rt2、Rt3、……、Rtm, m+1=n respectively, and the temperature measured value measured at the same time is T n、T1、T2、……、Tm. Fig. 2 can also be a schematic layout diagram of a chilled water return temperature sensor of a high-efficiency refrigeration machine room, and the method can be used for calibrating the chilled water return temperature sensor. The temperature sensor of the chilled water return main pipe is Rt, the temperature measured value measured at the same time is T m+1, the temperature sensors of all chilled water return branch pipes are Rt 1、Rt2、Rt3、……、Rtm respectively, and the temperature measured value measured at the same time is T 1、T2、……、Tm.
If the bypass pipe is not arranged between the water supply main pipe and the water return main pipe of the chilled water at the cooling water/freezing side of the refrigerating machine room, or the bypass pipe is arranged, but the valve is not opened, if the heat loss in the pipeline is neglected, the cooling water/chilled water of each branch pipeline is the same and fully mixed main pipeline, so that theoretically, when the temperature sensor has no error, the temperature value of the cooling water/chilled water return of the refrigerating machine is equal. When the temperature sensors on the main cooling water/chilled water return pipeline and the branch pipelines are calibrated, according to the objectively existing temperature consistency logic relationship, the temperature of the cooling water/chilled water return in the main pipeline is equal to that of the cooling water/chilled water return in all branch pipelines, namely the true values of Rt and Rt 1、Rt2、Rt3、……、Rtm are completely equal.
The built temperature inconsistency objective function is (1):
Where U T′ denotes a temperature inconsistency value, T i denotes a temperature measurement value of an ith temperature sensor, The average value of the temperature correction values of all the temperature sensors is represented, and n represents the total number of temperature sensors, where n=m+1. Theoretically, when the temperature sensor has no error, the temperature inconsistency value between the cooling water backwater temperature sensor in the main pipeline and the cooling water backwater temperature sensors in all branch pipelines is 0.
If the bypass valve is opened, the cooling water/chilled water in the main water supply pipe at the cooling side/chilled side is mixed in the main water return pipe to cause the change of the water return temperature in the main water return pipe, and if the heat loss in the pipeline is neglected, the cooling water/chilled water of each branch pipeline is the same and fully mixed main pipeline, so that theoretically, when the temperature sensor has no error, the water return temperature value of the cooling water/chilled water branch pipe of the cooling machine is equal and the water return temperature of the cooling water/chilled water return main pipe is unequal. When the temperature sensors on the cooling water/chilled water return branch pipes are calibrated, the temperatures of the cooling water/chilled water return in all branch pipes are equal according to the objectively existing temperature consistency logic relationship, namely the true values of Rt 1、Rt2、Rt3、……、Rtm are completely equal.
The built temperature inconsistency objective function is (1):
Where U T′ denotes a temperature inconsistency value, T i denotes a temperature measurement value of an ith temperature sensor, The average value of the temperature correction values of all the temperature sensors is represented, and n represents the total number of temperature sensors, where n=m. In theory, when the temperature sensor has no error, the temperature inconsistency value of the chilled water return water temperature sensors in all branch pipelines is 0.
In step 20, under the steady-state measurement condition, the measured value is generated by the influence of the systematic error due to the true value, and the two have a linear relationship, so that a correction function can be established for each temperature sensor, and the correction function is expressed as (2):
t i′＝aiTi+bi type (2)
Where T i' represents the temperature correction value of the i-th temperature sensor, a i represents the linear drift correction parameter of the i-th temperature sensor, T i represents the temperature measurement value of the i-th temperature sensor, and b i represents the fixed deviation correction parameter of the i-th temperature sensor.
In step 30, collecting the historical measurement data of all the temperature sensors in a preset time period, performing preprocessing operations such as outlier rejection and null value interpolation on the collected historical measurement data, and then constructing a plurality of steady-state measurement working condition sub-data sets to form a steady-state measurement data set. Wherein the temperature measurement value does not change by more than + -0.5deg.C every 30 minutes in each steady state measurement condition subset of data sets.
In the process of changing working conditions, the measured data are irregular, the temperature consistency logic relation is not met, and the state of balance and unchanged is achieved only under the condition that no external disturbance influence such as load change, environment temperature change and the like is avoided, namely, the measured parameters of the system in steady-state working conditions meet the physical law. In the embodiment, screening is performed from historical measurement data, a plurality of steady-state measurement working condition sub-data sets are constructed, a steady-state measurement data set is formed, the steady-state measurement data set reflects the real performance of the system in a non-disturbance state, correspondingly, the data is more representative, the processing speed of a follow-up optimization correction function is increased, and the quality of the data after calibration is improved to a certain extent.
Step 40 specifically includes:
Step 401, calculating a measurement inconsistency value of all temperature measurement values at each time in the steady-state measurement data set by using equation (3), namely, a temperature inconsistency value before calibration at each time:
Where U T denotes a measurement inconsistency value, T i denotes a temperature measurement value of an ith temperature sensor, Represents the average of the temperature measurements of all temperature sensors, and n represents the total number of temperature sensors.
According to the temperature inconsistency value before each moment calibration, an inconsistency histogram is drawn, a distribution model is determined according to the distribution shape of the histogram, and a probability density function is fitted.
And performing simulation sampling by using the fitted model, calculating a 95% confidence interval of the data, and screening out measurement data falling in the 95% confidence interval from the steady-state measurement data set to form typical measurement data. The measured data in the 95% confidence interval can represent the main trend and distribution characteristic of the data, and meanwhile, the follow-up correction function parameter optimizing speed can be increased, and the data quality of the temperature correction value is improved.
Step 402, initializing a correction function of each temperature sensor, and calculating a temperature correction value of each temperature sensor at each time based on the typical measurement data by using the correction function of each temperature sensor. And calculating to obtain a temperature inconsistency value of the temperature correction value at each moment by using a temperature inconsistency objective function.
And step 403, taking the temperature inconsistency objective function as an optimization target, optimizing the correction function of each temperature sensor until the optimization cut-off condition is reached, and stopping optimizing to obtain the optimized correction function.
In the step 403, the temperature inconsistency objective function is used as an optimization target, and the correction parameters of the correction functions of the temperature sensors are iteratively updated by using a gradient descent method or a genetic algorithm, so as to obtain the optimized correction functions.
The optimization cutoff condition comprises that the temperature inconsistency value is smaller than or equal to a preset threshold value. Preferably, the preset threshold is 0.05-0.1. When the temperature inconsistency value of the temperature correction value at each moment is smaller than or equal to a preset threshold value or reaches the iteration cut-off condition of the optimization algorithm, the optimization is finished, and the optimized correction function is obtained.
According to the temperature sensor calibration method based on true value regression, a regression model between true values of the temperature sensor is established by utilizing temperature consistency logic association, and the consistency degree of the association is measured by using temperature inconsistency. The method provided by the embodiment of the invention is applied to the calibration of the cooling water backwater temperature sensor in the refrigerating machine room or the calibration of the chilled water backwater temperature sensor, and according to the first law of energy conservation and thermodynamics in physics, if the heat loss caused by the heat preservation aspect of a pipeline is ignored, the chilled backwater or the cooling backwater separated from a backwater main pipe can be approximately considered to have unchanged temperature at the moment (same moment) of entering each branch pipe, namely backwater temperature consistency correlation. Unlike the conventional temperature sensor calibration method, which relies on the deviation between the measured value and the true value to verify the validity, the method of the embodiment evaluates the credibility of the measured value of each backwater temperature sensor at a certain moment by establishing a temperature inconsistency objective function. Assuming that the measured values of all the temperature sensors in the main backwater pipe and the backwater branch pipes of the cold machine are true values in an ideal state at a certain moment, the temperature inconsistency of the main backwater pipe and the backwater branch pipes of the cold machine is 0.00 at the moment, namely the lower the inconsistency of the measured values of all backwater temperature sensors at the same moment is, the more reliable the measured values are. According to the method provided by the embodiment of the invention, the redundant correlation among the temperature parameters is constructed through the physical characteristics of the water system of the refrigerating machine room, the problems of small number of sensors, low precision and the like of the water system are solved, and high-quality data support is provided for the optimized operation of the high-efficiency machine room.
A specific example is provided below.
And selecting half-year temperature measurement values of the temperature sensors on the cooling water backwater main pipe in the refrigeration machine room and the cooling water backwater branch pipes of the six coolers to calibrate, wherein the arrangement structure of the cooling water side temperature sensors in the refrigeration machine room is shown in figure 2. And collecting historical measurement data T 7 of a temperature sensor on a main cooling water return pipe within half a year and historical measurement data T 1～T6 of temperature sensors on cooling water return branch pipes corresponding to six cold machines, establishing a steady-state measurement data set, and screening out typical measurement data. And optimizing the correction functions of the seven cooling water backwater temperature sensors to finally obtain correction values corresponding to typical measurement data. Four measurement data of different working conditions at different moments in the typical measurement data are selected, as shown in table 1. Correction data of different working conditions at four moments after calibration are shown in table 2. The probability density distribution of the data before and after calibration is shown in fig. 3.
Table 1 measurement values and inconsistencies for different conditions at four times before calibration
TABLE 2 correction values and inconsistencies for different conditions at four times after calibration
As can be seen from table 1, the inconsistency of the seven cooling water return water temperature sensors before calibration is large, and the inconsistency of the seven cooling water return water temperature sensors after calibration is small. As can be seen from fig. 3, the method of the embodiment of the invention can be effectively applied to calibration of different working conditions of the return water temperature sensor of the air conditioning system, and compared with the measured value, the corrected value inconsistency is greatly reduced, and the data distribution is concentrated, so that the accuracy of the return water temperature measurement of the chilled water or the cooling water is ensured.
The foregoing has shown and described the basic principles, principal features and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the specific embodiments described above, and that the above specific embodiments and descriptions are provided for further illustration of the principles of the present invention, and that various changes and modifications may be made therein without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (10)
1. The temperature sensor calibration method based on true value regression is characterized by comprising the following steps of:
step 10, constructing a temperature inconsistency objective function according to the temperature consistency logic association relation;
step 20, constructing a correction function for each temperature sensor;
step 30, collecting historical measurement data of all temperature sensors in a preset time period, and establishing a steady-state measurement data set;
Step 40, based on a steady-state measurement data set, taking a temperature inconsistency objective function as an optimization target, optimizing the correction function of each temperature sensor until the optimization cutoff condition is reached, and stopping optimization to obtain an optimized correction function;
And step 50, taking the temperature correction value output by the correction function of each optimized temperature sensor as calibrated temperature data.
2. The method for calibrating a temperature sensor based on true value regression according to claim 1, wherein the logical association relationship of temperature consistency means that the fluid temperature at all positions in the same pipeline is equal or/and the fluid temperature in the main pipeline is equal to the fluid temperature in all branch pipelines or/and the fluid temperature in all branch pipelines under the condition that the pipeline fluid is fully developed and the pipeline fluid is stable.
3. The temperature sensor calibration method based on true value regression according to claim 1, wherein the temperature inconsistency objective function is of formula (1):
Where U T′ denotes a temperature inconsistency value, T i denotes a temperature measurement value of an ith temperature sensor, The average value of the temperature correction values of all the temperature sensors is represented, and n represents the total number of the temperature sensors.
4. The temperature sensor calibration method based on true value regression according to claim 1, wherein the correction function is formula (2):
In the formula (2) of T i′＝aiTi+bi, T' i represents a temperature correction value of the i-th temperature sensor, a i represents a linear drift correction parameter of the i-th temperature sensor, T i represents a temperature measurement value of the i-th temperature sensor, and b i represents a fixed deviation correction parameter of the i-th temperature sensor.
5. The method for calibrating a temperature sensor according to claim 1, wherein the step 40 specifically comprises:
step 401, calculating a temperature inconsistency value of each moment measured value in a steady-state data set by using a temperature inconsistency objective function, and screening typical measured data from the steady-state measured data set;
Step 402, initializing a correction function of each temperature sensor, and calculating a temperature correction value of each temperature sensor at each moment by using the correction function of each temperature sensor based on typical measurement data; calculating to obtain a temperature inconsistency value of the temperature correction value at each moment by using a temperature inconsistency objective function;
And step 403, taking the temperature inconsistency objective function as an optimization target, optimizing the correction function of each temperature sensor until the optimization cut-off condition is reached, and stopping optimizing to obtain the optimized correction function.
6. The method for calibrating a temperature sensor based on true value regression according to claim 5, wherein the step 401 comprises:
Calculating a measurement inconsistency value of all temperature measurement values at each moment in the steady-state measurement data set by using the formula (3):
Where U T denotes a measurement inconsistency value, T i denotes a temperature measurement value of an ith temperature sensor, Representing the average value of the temperature measurement values of all the temperature sensors, and n represents the total number of the temperature sensors;
Drawing an inconsistency histogram, determining a distribution model according to the distribution shape of the histogram, and fitting a probability density function;
and performing simulated sampling by using the fitted probability density function, calculating a 95% confidence interval of the data, and screening out the data falling in the 95% confidence interval from the steady-state measurement data set to form typical measurement data.
7. The method according to claim 5, wherein in step 403, the temperature inconsistency objective function is used as an optimization target, and the correction parameters of the correction functions of the temperature sensors are iteratively updated by using a gradient descent method to obtain the optimized correction functions.
8. The method according to claim 5, wherein in step 403, the temperature inconsistency objective function is used as an optimization target, and the correction parameters of the correction functions of the temperature sensors are iteratively updated by using a genetic algorithm to obtain the optimized correction functions.
9. The method of claim 5, wherein the optimal cutoff condition comprises a temperature inconsistency value that is less than or equal to a predetermined threshold.
10. The method for calibrating a temperature sensor according to claim 9, wherein the preset threshold is 0.05-0.1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410796400.1A CN118565661A (en) | 2024-06-20 | 2024-06-20 | Temperature sensor calibration method based on true value regression |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410796400.1A CN118565661A (en) | 2024-06-20 | 2024-06-20 | Temperature sensor calibration method based on true value regression |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118565661A true CN118565661A (en) | 2024-08-30 |
Family
ID=92478383
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410796400.1A Pending CN118565661A (en) | 2024-06-20 | 2024-06-20 | Temperature sensor calibration method based on true value regression |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118565661A (en) |
-
2024
- 2024-06-20 CN CN202410796400.1A patent/CN118565661A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Sun et al. | Building energy doctors: An SPC and Kalman filter-based method for system-level fault detection in HVAC systems | |
| CN108168030B (en) | Intelligent control method based on refrigeration performance curve | |
| CN107940679B (en) | Group control method based on performance curve of water chilling unit of data center | |
| CN103912966A (en) | Optimal control method for ground source heat pump refrigerating system | |
| CN114004374B (en) | Air conditioning unit sensor fault diagnosis method based on Boltzmann machine voting | |
| CN101813920B (en) | Virtual redundancy method for temperature sensor of power station turboset | |
| CN104697704A (en) | Method for detecting discharge pressure faults of water chilling unit | |
| Zhao et al. | Improvement of virtual in-situ calibration in air handling unit using data preprocessing based on Gaussian mixture model | |
| CN113822359B (en) | Heat pump system self-adaptive fault diagnosis method based on residual data scaling strategy | |
| CN110688740A (en) | Modelica combined simulation optimization-based cold water machine room model calibration method | |
| CN109711049B (en) | Hybrid subway air conditioner water chilling unit energy efficiency estimation method | |
| Zhang et al. | From laboratory to on-site operation: Reevaluation of empirically based electric water chiller models | |
| Sun et al. | A novel in-situ sensor calibration method for building thermal systems based on virtual samples and autoencoder | |
| CN118565661A (en) | Temperature sensor calibration method based on true value regression | |
| CN110826258B (en) | Space-time convolutional neural network-based data center machine room simulation environment construction method | |
| CN113188585B (en) | Freezing station sensor fault diagnosis method based on few redundant sensors | |
| Sun et al. | Research on fault detection method for air handling units system | |
| CN117366799A (en) | Control method of building cold source system | |
| CN111076372B (en) | Online identification and repair method for sensor error in primary return air conditioning system | |
| CN110779249B (en) | On-line diagnosis method for sensor concurrent faults in solar heat pump system | |
| Sun et al. | Partial Fault Detection of Cooling Tower in Building HVAC System | |
| CN116882264A (en) | Office building cold load data calibration and prediction method based on multi-feature fusion | |
| CN115930376B (en) | Air conditioner control algorithm evaluation method, device, equipment and storage medium | |
| CN117928052B (en) | Energy-saving control method and system for central air conditioner | |
| CN116857758A (en) | Cold source system data anomaly detection and fault investigation method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |