High precision oven temperature monitoring device and method for coil drying
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- ZHANGJIAGANG SHUANGCHENG ELECTRICIAN EQUIP CO LTD
- Filing Date
- 2026-05-22
- Publication Date
- 2026-06-19
Smart Images
Figure CN122237792A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of coil drying technology, and specifically to a high-precision temperature monitoring device and method for coil drying ovens. Background Technology
[0002] In the manufacturing process of power equipment such as transformers and reactors, insulation drying is a crucial step in ensuring the electrical performance of the products. Currently, the industry commonly uses vapor phase drying or vacuum transformer drying to remove moisture from inside the coils. As the voltage levels of power equipment increase, coils typically employ thick-walled insulation structures, which leads to significant thermal inertia characteristics during the drying process.
[0003] In the later stages of the drying process, the surface insulation material of the coil has often reached the basic drying standard, but trace amounts of moisture may still remain in the deeper insulation. For this stage, traditional methods mainly rely on moisture content monitoring or dew point monitoring to determine the drying endpoint. However, under steady-state conditions, the evaporation of deep moisture is extremely slow, making it difficult to form a significant moisture content signal, which makes it difficult to guarantee the accuracy of judging the true drying state inside the coil. Summary of the Invention
[0004] To address the technical problem of accurately determining the true drying state inside coils, this application aims to provide a high-precision temperature monitoring device and method for coil drying ovens. The specific technical solution adopted is as follows: Firstly, a high-precision temperature monitoring method for an oven used for coil drying is provided. This method includes: maintaining the coil at a constant temperature within the oven; reducing the air pressure within the oven to induce a phase change and endothermic reaction in residual moisture inside the coil; and acquiring the actual temperature change curve of the coil during the air pressure adjustment process within the oven. Based on the actual temperature change curve and a reference temperature change curve, a waveform distortion value is determined, and a temperature deviation value is determined based on the same curve. The waveform distortion value characterizes the degree of morphological distortion of the actual temperature change curve relative to the reference temperature change curve, and the temperature deviation value characterizes the overall difference between the first time-series temperature data corresponding to the actual temperature change curve and the second time-series temperature data corresponding to the reference temperature change curve. Based on the waveform distortion value and the temperature deviation value, the evaporation characteristic value of the coil is determined, and the drying state of the coil is determined based on the evaporation characteristic value.
[0005] In one possible design, the process of determining the waveform distortion value includes: determining a two-dimensional distance matrix based on first and second time-series temperature data. The distance matrix characterizes the local distance between target point pairs, which are data corresponding to the same time point in both the first and second time-series temperature data. Based on the two-dimensional distance matrix, the minimum cumulative path for aligning the actual temperature change curve with the reference temperature change curve is determined. The total path distance and path length corresponding to the minimum cumulative path are determined. The waveform distortion value is determined based on the total path distance and path length.
[0006] In one possible design, the process of determining the temperature deviation value includes: determining the root mean square error of the first time series temperature data and the second time series temperature data. The root mean square error is then determined as the temperature deviation value.
[0007] In one possible design, acquiring the actual temperature change curve of the coil during the air pressure regulation process within the oven includes: acquiring time-series air pressure data of the coil during the air pressure regulation process within the oven; performing a first-order difference operation on the first time-series temperature data to obtain temperature change rate sequence data, and performing a first-order difference operation on the time-series air pressure data to obtain air pressure change rate sequence data; determining the target correlation between the coil temperature and the air pressure change within the oven based on the temperature change rate sequence data and the air pressure change rate sequence data; and acquiring the actual temperature change curve if the target correlation exceeds a preset correlation threshold.
[0008] In one possible design, the target correlation between the coil temperature and the oven pressure change is determined based on temperature change rate sequence data and air pressure change rate sequence data. This includes: adjusting the temperature change rate sequence data according to a preset lag step to obtain adjusted temperature change rate sequence data; calculating candidate correlation degrees between the adjusted temperature change rate sequence data and the air pressure change rate sequence data; and determining the candidate correlation degree as the target correlation degree if the candidate correlation degree meets the correlation conditions.
[0009] In one possible design, the above method further includes: determining the temperature delay response time based on temperature change rate sequence data and air pressure change rate sequence data. The temperature delay response time is the delay time in which the coil's temperature follows the change in air pressure inside the oven. The temperature data between the air pressure adjustment time and the temperature delay response time is determined as the second time-series temperature data; the air pressure adjustment time is the time when the rate of change of air pressure inside the oven is the largest or the time when the air pressure inside the oven drops to the target air pressure. A reference temperature change curve is determined based on the second time-series temperature data.
[0010] In one possible design, the dryness state of the coil is determined based on evaporation characteristic values, including: If the evaporation characteristic value does not exceed the preset evaporation characteristic threshold, the coil is determined to be in a dry state.
[0011] If the evaporation characteristic value exceeds the preset evaporation characteristic threshold, the coil is determined to be in a state to be dried.
[0012] In one possible design, the above method further includes: determining a drying strategy for the coil based on the temperature delay response time when the coil is in a state to be dried.
[0013] In one possible design, the drying strategy for the coil is determined based on the temperature delay response time, including: if the temperature delay response time does not exceed a preset time threshold, determining that the distribution depth of moisture in the coil does not exceed a preset depth threshold, and executing a first drying strategy; the first drying strategy is used to instruct the oven to dry the coil according to the current drying instruction.
[0014] If the temperature delay response time exceeds a preset time threshold, it is determined that the distribution depth of moisture in the coil exceeds a preset depth threshold, and a second drying strategy is executed; the second drying strategy is used to instruct the oven to extend the drying time of the coil and / or periodically adjust the air pressure in the oven.
[0015] Secondly, a high-precision temperature monitoring device for an oven used for coil drying is provided, comprising: a pressure regulating unit for reducing the pressure inside the oven when the coil is in a constant-temperature holding stage, thereby stimulating the residual moisture inside the coil to undergo phase change and absorb heat; a data acquisition unit for acquiring the actual temperature change curve of the coil during the pressure regulating process inside the oven; a waveform and deviation determination unit for determining the waveform distortion value and the temperature deviation value based on the actual temperature change curve and the reference temperature change curve. The waveform distortion value characterizes the degree of morphological distortion of the actual temperature change curve relative to the reference temperature change curve, and the temperature deviation value characterizes the overall difference between the first time series temperature data corresponding to the actual temperature change curve and the second time series temperature data corresponding to the reference temperature change curve; an eigenvalue calculation unit for determining the evaporation characteristic value of the coil based on the waveform distortion value and the temperature deviation value; and a state determination unit for determining the drying state of the coil based on the evaporation characteristic value.
[0016] This application offers the following advantages: During the constant-temperature maintenance phase of the coil within the oven, the air pressure inside the oven is reduced to induce a phase change and endothermic reaction in the residual moisture inside the coil. This disrupts the gas-liquid equilibrium under steady-state conditions, transforming the difficult-to-detect static distribution of deep residual moisture in the coil into a measurable dynamic temperature response, thus solving the technical challenge of weak and difficult-to-capture deep moisture signals under steady-state conditions. Furthermore, this application compares the actual temperature change curve with a reference curve characterizing the drying state, extracting waveform distortion values reflecting the degree of curve distortion and temperature deviation values reflecting overall numerical differences. Based on the comprehensive relationship between these two factors, an evaporation characteristic value is determined as the basis for judging the drying state. This comprehensive evaluation method, considering both waveform morphology and numerical differences, effectively distinguishes between nonlinear temperature distortion caused by moisture evaporation and linear numerical shifts caused by model errors. It avoids misjudging model inaccuracies as moisture residue or submerging the true moisture signal in model errors, significantly improving the accuracy and reliability of judging the drying state inside the coil. Attached Figure Description
[0017] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0018] Figure 1 This is a flowchart illustrating a high-precision temperature monitoring method for an oven used in coil drying, provided as an embodiment of this application. Figure 2 This is a schematic diagram of a high-precision temperature monitoring device for an oven used for coil drying, provided as an embodiment of this application. Detailed Implementation
[0019] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a high-precision temperature monitoring device and method for coil drying ovens proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.
[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.
[0021] The following description, in conjunction with the accompanying drawings, details a specific scheme for a high-precision temperature monitoring method for coil drying ovens provided in this application.
[0022] Please see Figure 1 It shows a flowchart of a high-precision temperature monitoring method for an oven used for coil drying according to an embodiment of this application, such as... Figure 1 As shown, the method includes the following steps S101-S105.
[0023] S101. When the coil inside the oven is in the constant temperature holding stage, the air pressure inside the oven is reduced to stimulate the residual moisture inside the coil to generate phase change and absorb heat.
[0024] One possible implementation is to perform air pressure regulation while the coil in the oven is in a constant-temperature holding phase (during which the coil is relatively dry with low moisture content) to induce a phase change and endothermic reaction in the residual moisture inside the coil. Specifically, before active excitation, the monitoring system first confirms that the oven environment meets the test conditions. The controller collects real-time air pressure and coil surface temperature data, calculates the fluctuation rate of the air pressure data, and determines whether the fluctuation rate is below a preset steady-state threshold. Simultaneously, the controller confirms that the current drying process is in a constant-temperature holding phase and that the coil surface temperature is close to the target heating temperature set by the process. Only when the air pressure is stable and the temperature is maintained does the monitoring system determine that the oven environment meets the test conditions, thus triggering the subsequent air pressure regulation operation.
[0025] After confirming that the oven environment meets the test conditions, the controller sends a command to the vacuum unit to execute a complete vacuum pulse operation. This vacuum pulse operation is not a permanent depressurization of the oven, but a short-term pressure disturbance designed to disrupt the gas-liquid balance within the coil to induce a phase change in moisture, while simultaneously preventing excessive heat loss due to prolonged low pressure. The specific execution steps are as follows: First, the controller opens the vacuum valve, causing the absolute pressure inside the oven to drop rapidly. This depressurization operation ensures that the saturated vapor pressure curve of the coil's moisture at the current temperature is crossed, and the pressure reduction can be set according to a preset ratio, for example, reducing the current pressure inside the oven by 20%. Second, after reaching the target pressure, this pressure state is maintained for a preset duration of pulse width, causing deep moisture within the coil to evaporate and absorb heat, but not enough to cause a significant overall cooling of the coil. Finally, the controller allows the pressure inside the oven to naturally or controllably rise back to the initial pressure, thus completing one full vacuum pulse operation.
[0026] While performing air pressure regulation, a sampling interval is determined to sample the air pressure data and coil temperature within the oven, thereby synchronously recording the time-series air pressure data within the oven and the time-series temperature data of the coil. To capture weak thermal response signals, this embodiment employs a high-precision, low-thermal-inertia temperature sensor with a measurement resolution exceeding 0.01°C. The high-precision, low-thermal-inertia temperature sensor can be installed close to the coil surface or embedded in the air duct outlet between coil layers to ensure rapid response to changes in the local thermal environment. This embodiment does not specifically limit the installation location of the high-precision, low-thermal-inertia temperature sensor.
[0027] Additionally, the following data processing logic can be executed in this step to establish a unified time reference: First, monitor the rate of change during the pressure drop process, and mark the time when the absolute value of the pressure drop rate reaches its maximum, the time when the pressure drops to the target pressure, or the time when the pressure drops to 95% of the target pressure difference as the pressure adjustment time. This pressure adjustment time will then be used as the time origin. Second, use the pressure adjustment time... Using the time origin as the starting point, a complete target time period is extracted, which covers the period before the barometric pressure regulation time. Includes coil temperature data before pressure regulation and the time period after pressure regulation. Data containing the complete coil temperature thermal response process, including the time period after the pressure regulation time. The length covers the preset maximum heat conduction delay time (e.g., take...). ),right No specific limitations are imposed. Ultimately, the time-series air pressure data and the first time-series temperature data for the target time period are obtained, and the time-series air pressure data and the first time-series temperature data are time-aligned.
[0028] The above The range of values and The range of values is pre-configured and not specifically limited, for example... The range of values for can be ( ), The range of values for can be ( ).
[0029] In this embodiment, the first time-series temperature data can be filtered data. Specifically, after acquiring the original time-series temperature data, a moving average filter or Gaussian smoothing is performed on the time-series temperature data to obtain the first time-series temperature data. Taking moving average filtering as an example, the filter window width is set to W (for example, the number of sampling points corresponding to 5 seconds). For the temperature data of each sampling point i, the arithmetic mean of the data within W / 2 before and after sampling point i is taken as the filtered temperature. In this way, since the acquired original signal is often mixed with high-frequency electromagnetic noise and environmental thermal fluctuations, this step can filter out the high-frequency random noise of the sensor itself, retain the low-frequency trend characteristics of temperature changes, and prevent noise peaks from being erroneously amplified in subsequent calculations.
[0030] In this embodiment, the time-series pressure data represents the oven pressure at different times within the target time period, and the first time-series temperature data represents the coil temperature at different time points within the target time period. The time-series pressure data and the first time-series temperature data can be collected using the same sampling interval. The maximum heat conduction delay time can be a pre-configured time, and this embodiment does not specifically limit it. This embodiment does not specifically limit the target pressure and target pressure difference; they can be flexibly configured according to the specific pressure inside the oven. For example, the target pressure can be 70% to 90% of the pressure before depressurization (i.e., a depressurization range of 10% to 30%). If the current pressure is maintained at 1000 Pa, the target pressure can be set to 700 Pa. 900 Pa, the target differential pressure can be 100 Pa or 200 Pa.
[0031] Understandably, in practical engineering, the pressure drop is not an ideal step signal, but rather a curve with near-exponential decay. At the beginning of vacuuming, the pressure drops rapidly; the closer to the target pressure, the slower the drop. The point where the pressure drops to 95% of the target pressure difference is located on the steep section of the pressure curve, minimally affected by small fluctuations at the end, and easily captured stably by the algorithm. Therefore, the time it takes for the pressure to drop to 95% of the target pressure difference is marked as the pressure regulation time.
[0032] The aforementioned vacuum pulse operation can cross the saturated vapor pressure curve corresponding to the current temperature inside the oven, thereby enabling the residual moisture inside the excitation coil to generate phase change endothermic heat.
[0033] S102. Obtain the actual temperature change curve of the coil during the air pressure regulation process in the oven.
[0034] As one possible approach, given the first time-series temperature data, an actual temperature change curve is generated based on that data.
[0035] As another possible implementation, time-series air pressure data of the coil during the air pressure regulation process inside the oven is acquired. A first-order difference operation is performed on the first time-series temperature data to obtain a temperature change rate sequence, and a first-order difference operation is performed on the time-series air pressure data to obtain a pressure change rate sequence. Based on the temperature and pressure change rate sequence data, the target correlation between the coil temperature and the air pressure change inside the oven is determined. If the target correlation exceeds a preset correlation threshold, the actual temperature change curve is acquired. This step is detailed in the following embodiment and will not be repeated here.
[0036] S103. Based on the actual temperature change curve and the reference temperature change curve, determine the waveform distortion value and the temperature deviation value respectively.
[0037] Among them, the reference temperature change curve is used to characterize the temperature change of the coil in the drying state during the air pressure regulation process in the oven, the waveform distortion value is used to characterize the degree of distortion of the shape of the actual temperature change curve relative to the reference temperature change curve, and the temperature deviation value is used to characterize the overall difference between the first time series temperature data corresponding to the actual temperature change curve and the second time series temperature data corresponding to the reference temperature change curve.
[0038] One possible approach is to determine a two-dimensional distance matrix based on first and second time-series temperature data. This distance matrix characterizes the local distance between target point pairs, which are data points corresponding to the same time point in both the first and second time-series temperature data. Based on the two-dimensional distance matrix, the minimum cumulative path for aligning the actual temperature change curve with the reference temperature change curve is determined. The total path distance and path length corresponding to the minimum cumulative path are then determined. Finally, the waveform distortion value is determined based on the total path distance and path length.
[0039] For example, first construct a two-dimensional distance matrix. For instance, establish a matrix of size... The two-dimensional distance matrix, where and , respectively, represent the sequence lengths of the first and second time-series temperature data. The second time-series temperature data in the two-dimensional distance matrix... Line number Column elements This indicates the first time series temperature data. The point and the second time series temperature data The distance between points is typically calculated using Euclidean distance. This two-dimensional distance matrix records the degree of difference between all possible pairs of points, providing a basis for subsequent pathfinding.
[0040] Secondly, search for the minimum cumulative path. For example, a dynamic programming approach can be used, starting from the beginning of the matrix. Start by calculating the distance to each location step by step. Minimum cumulative distance The recursive formula can be: in, Corresponding to the upward movement distance, Corresponding to the distance moved to the left, The corresponding distance to move to the upper left, Elements in a two-dimensional distance matrix The distance between the points. In the above recursive formula, nonlinear scaling can be performed on the time axis to achieve the best alignment of the waveform characteristics of the two curves, thereby obtaining the search for the minimum cumulative path, where min() represents the minimum value function.
[0041] Subsequently, given the minimum cumulative path, the total distance corresponding to the minimum cumulative path is determined as the total path distance, and the number of matrix cells actually traversed by the minimum cumulative path is determined as the path length. Then, based on the total path distance and the path length, the waveform distortion value is determined. For example, the total path distance and the path length are input into the waveform distortion cost formula to obtain the waveform distortion value.
[0042] For example, the waveform distortion cost formula is as follows.
[0043] in, This represents the waveform distortion value. This represents the total path distance. This represents the path length. Since the actual temperature curve contains multiple sampling points, multiple correspondences are needed to fully describe the similarity between the two curves. Therefore, the path length is greater than 1 and will not be zero.
[0044] The above path length The number of matrix cells traversed by the path, typically between and Between them. Normalization was used to make sequences of different lengths comparable. The resulting... This refers to the waveform distortion value; a larger value indicates a more severe distortion of the measured curve relative to the reference curve. Because the heat absorption of moisture causes nonlinear distortions such as "concave" or "plateauing" in the temperature recovery curve, and these distortions cannot be completely eliminated by time axis scaling, therefore... It is highly sensitive to the presence of moisture; while simple model errors usually only cause an overall translation or scaling of the curve. The dynamic time warping algorithm can compensate for these differences well by scaling the time axis. It is robust to model errors.
[0045] The process of determining the temperature deviation value includes: determining the root mean square error of the first time series temperature data and the second time series temperature data, and determining the root mean square error as the temperature deviation value.
[0046] Specifically, the temperature deviation value is calculated using the root mean square error method, and the temperature deviation value is determined based on the second time series temperature data, the first time series temperature data, and the temperature deviation determination formula.
[0047] For example, the formula for determining the temperature deviation is as follows.
[0048] in, This indicates the first time series temperature data. A measured temperature, This indicates the first time series temperature data. A reference temperature, This represents the total number of points in the first time series temperature data. This indicates the numerical value of the temperature deviation.
[0049] The temperature deviation determination formula calculates the deviation at each sampling point (measured value minus the baseline), squares it, sums the results, divides by the total number of points to obtain the mean square, and finally takes the square root to obtain the root mean square error. The squaring operation eliminates the cancellation of positive and negative deviations, and the square root restores the dimension to temperature. This index integrates the differences at all points, including both systematic shifts or scaling caused by inaccurate model parameter estimations and local depressions or bulges caused by moisture endothermic reactions. The larger the temperature deviation value, the more severe the deviation between the measured value and the baseline.
[0050] Temperature deviation represents the overall degree of deviation between the measured value and the reference value, but it cannot distinguish whether this deviation originates from model error or moisture effect. The unit is °C, and the range is non-negative real numbers.
[0051] S104. Determine the evaporation characteristic value of the coil based on the waveform distortion value and temperature deviation value.
[0052] As one possible approach, given the waveform distortion value and temperature deviation value, the evaporation characteristic value of the coil can be obtained by determining the formula based on the waveform distortion value, temperature deviation value, and evaporation characteristic value.
[0053] For example, the formula for determining the evaporation characteristic value is as follows.
[0054] in, This is the characteristic value of evaporation. This represents the waveform distortion value. This represents the temperature deviation value. The preset system noise basis constant (e.g., it can take values of...) ), used to prevent when When the denominator approaches zero, it becomes invalid.
[0055] In this embodiment of the application, for The range of values for is not specifically limited, for example .
[0056] In the formula for determining the evaporation characteristic value, The waveform distortion value is expressed in temperature units (°C), representing the degree of distortion of the actual temperature change curve relative to the reference temperature change curve. The temperature deviation is a numerical value, and its unit is also a temperature unit (°C), representing the overall degree of difference between the two curves in terms of values. The preset noise floor constant, in temperature units (°C), is used to prevent noise from... The calculation is performed to address numerical instability caused by a zero denominator when the denominator approaches zero, while also suppressing background noise interference. The value is usually slightly higher than the sensor noise level, for example, it can be taken as... ℃.
[0057] The evaporation characteristic value obtained through this ratio calculation This is a dimensionless number. Its physical meaning is that when the coil is in a dry state, the difference between the measured curve and the reference curve is mainly due to linear errors caused by inaccurate estimation of model parameters; at this time, the waveform distortion value... The difference is relatively small (because the dynamic time warping algorithm can effectively compensate for linear differences through time axis scaling), while the temperature deviation value is relatively small. It may be relatively large, therefore Approaching zero; when residual moisture exists inside the coil, the moisture evaporation absorbs heat, causing nonlinear distortion of the measured curve, and the waveform distortion value... Significantly increased, while the temperature deviation value The increase was relatively small, making The value increased significantly. Therefore, the evaporation characteristic value... This becomes a criterion that is highly sensitive to moisture and robust to model errors, effectively amplifying the contribution of moisture signals to the comprehensive index and providing a reliable quantitative basis for accurately determining the dryness of the coil.
[0058] S105. Determine the dryness state of the coil based on the evaporation characteristic value.
[0059] One possible approach is to determine whether the evaporation characteristic value exceeds a preset evaporation characteristic threshold. If the evaporation characteristic value does not exceed the preset evaporation characteristic threshold, the coil is determined to be in a dried state. If the evaporation characteristic value exceeds the preset evaporation characteristic threshold, the coil is determined to be in a state requiring drying.
[0060] The aforementioned preset evaporation characteristic threshold can be the evaporation characteristic value of the sample coil. For example, before drying the coil in the oven, a coil of the same model as the coil in the oven, which has been confirmed to be completely dry, is selected as the sample coil. The calibration sample is placed in the oven, and under the same process temperature conditions as the actual drying process, a complete vacuum pulse operation is performed on the sample coil. The actual temperature change curve and reference temperature change curve of the sample coil are obtained according to the above steps, thereby calculating the waveform distortion value and temperature deviation value of the sample coil, and then solving for the evaporation characteristic value of the sample coil. The above process is repeated multiple times to obtain a set of sample data of the evaporation characteristic values of the sample coil.
[0061] Furthermore, the arithmetic mean of the sample data set for calculating the evaporation characteristic values is obtained. and standard deviation Among them, the arithmetic mean Reflects the central tendency and standard deviation of evaporation characteristic values under dry conditions. This reflects its fluctuation range. The preset evaporation characteristic threshold is used. Set as: .
[0062] The aforementioned preset evaporation characteristic threshold represents the dry background noise limit at a 99.7% confidence level. When the coil under test is subsequently monitored, if the calculated evaporation characteristic value does not exceed this threshold, there is sufficient confidence that the difference between the measured curve and the reference curve does not exceed the normal fluctuation range of the dry background noise, and the coil can be determined to be in a dry state. If the evaporation characteristic value exceeds this threshold, it indicates a significant difference beyond the background noise, which usually means that there is still residual moisture evaporating inside the coil. The dry / wet determination threshold determined through the above calibration process fully considers the background noise level under specific operating conditions, avoids the risk of misjudgment caused by subjectively setting thresholds, and improves the objectivity and reliability of dry state determination.
[0063] The aforementioned preset evaporation characteristic threshold can also be pre-configured, and this application embodiment does not specifically limit this.
[0064] As one possible approach, the drying strategy for the coil can be determined based on the temperature delay response time while the coil is in a state awaiting drying.
[0065] In some embodiments, if the temperature delay response time does not exceed a preset time threshold, it is determined that the distribution depth of moisture in the coil does not exceed a preset depth threshold, and a first drying strategy is executed; the first drying strategy is used to instruct the oven to dry the coil according to the current drying instruction.
[0066] If the temperature delay response time exceeds a preset time threshold, it is determined that the moisture distribution depth within the coil exceeds a preset depth threshold, and a second drying strategy is executed. The second drying strategy is used to instruct the oven to extend the drying time of the coil and / or periodically adjust the air pressure within the oven. Periodically adjusting the air pressure within the oven can be achieved by periodically executing the aforementioned vacuum pulse operation.
[0067] The aforementioned preset time threshold is a pre-configured time, such as 120 seconds. This application embodiment does not specifically limit this value. The preset depth threshold can be 4 mm or 5 mm; this application embodiment does not specifically limit this value.
[0068] Understandably, if the temperature delay response time does not exceed the preset time threshold, the coil in the oven is determined to be in a shallowly moistened state. In this state, the moisture is mainly distributed in the shallower part of the coil's insulation layer, and the evaporation channel is relatively unobstructed. Maintaining the current vacuum level and heating power, the current drying process can continue. If the temperature delay response time exceeds the preset time threshold, the coil in the oven is determined to be in a deeply moistened state. In this state, the moisture is located deep within the insulation layer, resulting in high resistance to drainage and a risk of "false drying." In this case, it is necessary to extend the coil's drying time and / or periodically adjust the air pressure inside the oven to enhance the escape of deep moisture.
[0069] In one possible design, S102 above includes: S1021-S1024.
[0070] S1021. Obtain time-series air pressure data of the coil during the air pressure regulation process in the oven.
[0071] S1022. Perform a first-order difference operation on the first time series temperature data to obtain the temperature change rate sequence data, and perform a first-order difference operation on the time series air pressure data to obtain the air pressure change rate sequence data.
[0072] One possible approach is to input the first time-series temperature data into the temperature change rate formula to obtain the temperature change rate series data.
[0073] For example, the formula for the rate of temperature change is as follows.
[0074] in, This represents the nth measured temperature in the first time series temperature data. This represents the (n-1)th measured temperature in the first time series temperature data. The sampling interval is any two adjacent measured temperatures in the first time series temperature data. The sampling interval is a positive number greater than 0. This represents the rate of change of the nth measured temperature in the first time series temperature data.
[0075] This formula for the rate of temperature change is the simplest forward (or backward) difference approximation in numerical differentiation, used to calculate the rate of temperature change over time. Its logic is: divide the difference between the current moment and the previous moment by the time interval to obtain the average rate of change over that time interval, and use this as an approximation of the instantaneous rate of change at the current moment. If the temperature increases with time, then... If the temperature decreases over time, then Differential operations can eliminate the DC component (i.e., absolute value) of the original signal, highlighting dynamic change characteristics and facilitating subsequent analysis of the correlation between excitation and response.
[0076] Temperature change rate series data characterizes the speed and direction of temperature changes, forming the basis for subsequent cross-correlation analysis. It transforms temperature signals into rate-of-change signals, allowing analysis to focus on dynamic responses rather than static values, thus adapting to monitoring needs at different process temperatures.
[0077] The time-series air pressure data is input into the air pressure change rate formula to obtain the air pressure change rate series data. The time-series air pressure data can be the data obtained after applying a moving average filter to the collected raw air pressure data.
[0078] For example, the formula for the rate of change of air pressure is as follows.
[0079] in, This represents the nth measured air pressure value in the time series air pressure data. This represents the (n-1)th measured air pressure value in the time series air pressure data. The sampling interval is any two adjacent measured temperatures in the first time series temperature data. This represents the rate of temperature change of the nth measured air pressure value in the time series air pressure data.
[0080] This formula for the rate of change of air pressure is the simplest forward (or backward) difference approximation in numerical differentiation, used to calculate the rate of change of air pressure over time. Its logic is as follows: the difference between the current moment and the previous moment is divided by the time interval to obtain the rate of change of air pressure over that time interval, which is then approximated as the instantaneous rate of change at the current moment. Through differential operations, the DC component (i.e., absolute value) of the original signal can be eliminated, highlighting dynamic change characteristics and facilitating subsequent analysis of the correlation between excitation and response.
[0081] Pressure change rate sequence data characterizes the speed and direction of temperature changes and forms the basis for subsequent cross-correlation analysis. It transforms pressure signals into pressure change rate signals, allowing the analysis to focus on dynamic responses rather than static values, thus adapting to monitoring needs under different process pressures.
[0082] S1023. Based on the temperature change rate sequence data and the air pressure change rate sequence data, determine the target correlation between the coil temperature and the air pressure change in the oven.
[0083] As one possible approach, the temperature change rate sequence data is adjusted according to a preset lag step to obtain the adjusted temperature change rate sequence data; the candidate correlation degree between the adjusted temperature change rate sequence data and the air pressure change rate sequence data is calculated, and if the candidate correlation degree meets the correlation conditions, the candidate correlation degree is determined as the target correlation degree.
[0084] The aforementioned association condition can be either the highest association degree among multiple candidate association degrees, or the candidate association degree exceeding a preset association degree value. The preset association degree value is pre-configured, and this embodiment does not specifically limit its value.
[0085] For example, the association condition mentioned above can be that the candidate association degree is the highest among multiple candidate association degrees. A preset lag step size is set as a variable. The range of values is from up to the maximum preset number of steps The corresponding time lag range is to For each lag step Shift the temperature change rate series data to the left One unit (i.e., assuming the coil's temperature response lags behind the gas pressure excitation) (Sampling intervals), then multiply point-by-point with the pressure change rate sequence data and sum them to obtain the candidate correlation degree. : in, This represents the first [number] in the sequence data of air pressure change rate. The rate of change of air pressure at each sampling point This represents the temperature change rate sequence data after shifting. Temperature change rate values at each sampling point, summation index Cover the overlapping portion of the two sequences. In this way, the similarity between the two sequences at different relative offsets is calculated using a sliding dot product, with the translation amount... When the overlap between the two waveforms is exactly equal to the actual physical lag time, the sum of their products reaches its extreme value.
[0086] Traverse all The value yields the absolute number of the correlation between multiple candidates. This allows us to determine the target correlation degree as the one with the largest absolute value among multiple candidate correlation degrees, and to determine the lag step size corresponding to the target correlation degree as the optimal lag step size. The reason for taking the absolute number is that, regardless of whether the temperature response and the pressure excitation are positively correlated (pressure decrease leads to temperature decrease) or negatively correlated (pressure decrease leads to temperature increase), as long as there is a significant causal relationship, the absolute number of the candidate correlation degree can reflect the correlation strength.
[0087] Additionally, multiply the optimal hysteresis step size by the sampling interval. The temperature delay response time is obtained. The temperature delay response time characterizes the time difference from the start of air pressure regulation to the point where the coil surface temperature reaches its maximum rate of change. Its magnitude directly reflects the radial depth of the moisture: the deeper the moisture, the longer the heat conduction path, and the greater the response delay.
[0088] S1024. If the target correlation degree exceeds the preset correlation degree threshold, obtain the actual temperature change curve.
[0089] In some embodiments, it is determined whether the target correlation degree exceeds a preset correlation degree threshold. If the target correlation degree exceeds the preset correlation degree threshold, the actual temperature change curve is acquired. If the target correlation degree does not exceed the preset correlation degree threshold, it is determined that there is no significant statistical correlation between temperature change and pressure change, i.e., the pressure pulse does not cause an effective thermal response. In this case, the current state can be marked as a no-response characteristic state, and further determined as "ineffective excitation of the coil" or "the coil is completely dry", and the drying operation is directly terminated.
[0090] In other embodiments, when the target correlation degree exceeds a preset correlation degree threshold, the temperature drop of the coil after the air pressure in the oven decreases is obtained, and when the temperature drop exceeds a preset threshold, the actual temperature change curve is obtained.
[0091] The aforementioned preset amplitude threshold is pre-configured, for example, it can be 0.05℃, and this embodiment of the application does not specifically limit it. The aforementioned preset correlation degree threshold is pre-configured, for example, it can be 0.6, and this embodiment of the application does not specifically limit it.
[0092] Understandably, if the temperature drop does not exceed a preset threshold, the pressure drop inside the oven is considered an invalid excitation. This indicates that the current physical environment (such as excessively low temperature or insufficient vacuum) cannot support the phase change of moisture, or that the temperature sensor has failed to detect the temperature change signal of the coil. In this case, the coil is reheated, the vacuuming is paused, the heating system is started to replenish the coil's enthalpy, and the temperature drop is re-detected after a period of time. If the temperature drop exceeds the preset threshold, the drying status of the coil continues to be monitored.
[0093] The aforementioned preset correlation threshold is pre-configured, and this application embodiment does not impose specific limitations on it.
[0094] This step involves a preliminary screening of the coil's dryness, comparing the target correlation degree with a preset correlation degree threshold. If the target correlation degree does not exceed the preset correlation degree threshold, it indicates that there is no significant statistical correlation between temperature and pressure changes, meaning the pressure pulse did not elicit a valid thermal response; in this case, the current state is marked as no response. If the target correlation degree exceeds the preset correlation degree threshold, it indicates that a significant thermal response signal has been captured, and the response delay time can be calculated. This is used for subsequent coil water level depth positioning analysis.
[0095] In one possible design, in order to obtain a more accurate reference temperature change curve, the high-precision temperature monitoring method for the drying oven provided in this application embodiment further includes: S106-S107.
[0096] S106. Obtain the benchmark fitting sample.
[0097] As one possible implementation, given a fixed pressure regulation time and temperature delay response time, temperature response data is acquired during the delay period between the pressure regulation time and the temperature delay response time, and this acquired temperature response data is used as a benchmark fitting sample. Subsequently, it is determined whether the delay duration corresponding to the delay period exceeds the minimum fitting duration. If the delay duration does not exceed the minimum fitting duration, the coil is directly determined to be in a surface wet state, and the coil is then dried. If the delay duration exceeds the minimum fitting duration, step S107 is executed to obtain a reference temperature change curve.
[0098] The minimum fitting time can be determined based on the properties of the coil, and can be 10s or 13s. This application does not make specific limitations on this.
[0099] Understandably, given the premise that the surface insulation is already dry in the later stages of the drying process, the heat absorption effect of deep moisture needs a temperature delay response time to be transferred to the coil surface after the air pressure changes in the oven. Therefore, during the delay period between the air pressure adjustment time and the temperature delay response time, the change in the coil's surface temperature has not yet been affected by the heat absorption of deep moisture and is determined solely by the thermal conductivity characteristics of the insulation material itself. The data for this period is the dry-state response data.
[0100] S107. Obtain the reference temperature change curve based on the benchmark fitting sample.
[0101] As one possible approach, regression analysis is performed on the baseline fitted sample to obtain the reference temperature change curve.
[0102] For example, a pre-defined single-parameter exponential model is used to perform regression analysis on the benchmark fitted sample. This pre-defined single-parameter exponential model describes the law of temperature recovery to the target value of the coil after cooling, and its mathematical expression is: in, This indicates the reference temperature change curve at time [time missing]. Temperature value, To heat to the target temperature, a pre-configured constant is used. It can be the measured temperature at the time of pressure regulation or the arithmetic mean of the measured temperatures within a preset window (e.g., 3-5 seconds) before and after the pressure regulation time. The thermal recovery rate constant to be identified is given in units of . This characterizes how quickly the coil recovers its temperature after a disturbance. It is a natural constant, approximately 2.71828. It is an exponential decay function, representing the law of decay of the initial temperature difference over time; As a time variable, with time base zero The starting point is [value]. The model converges to its final value through mandatory constraints. To ensure that the fitted curve does not diverge and has a clear physical meaning: when hour, ;when hour, .
[0103] Subsequently, the least squares method was used to perform regression analysis on the benchmark fitted sample. Specifically, the regression analysis was performed on the benchmark fitted sample at each time step. Substituting the corresponding measured temperature values into the aforementioned preset single-parameter exponential model, an optimization algorithm is used to search for the heat recovery rate constant that minimizes the sum of squared errors between the calculated and measured values of the preset single-parameter exponential model. To obtain the optimal fitting parameters The objective function of the least squares method is: in, For a moment The measured temperature value, To heat to the target temperature, The measured temperature at the time of pressure regulation. Temperature delay response time, For the best fit parameters, To Perform a minimization optimization.
[0104] Finally, after obtaining the optimal fitting parameters Then, substitute this parameter into the aforementioned preset single-parameter exponential model, and add the time variable. Extending to the entire sampling period, that is, the period after the pressure regulation time. By calculating point by point, a reference temperature change curve covering the entire sampling period is generated.
[0105] The reference temperature change curve constructed through the above steps is based on measured data from the early stage of drying, before any moisture interference. This data is used to deduce the coil's thermal conductivity characteristics under current operating conditions, thereby calculating the ideal temperature recovery trajectory assuming no moisture evaporation throughout the entire monitoring period. This reference temperature change curve eliminates the interference of moisture evaporation on the temperature curve, providing a reliable reference standard for identifying moisture signals by comparing the actual temperature change curve with the reference temperature change curve.
[0106] The high-precision temperature monitoring method for coil drying ovens provided in this application offers at least the following advantages: When the coil is in a constant-temperature holding phase within the oven, the air pressure inside the oven is reduced to induce a phase change and endothermic reaction in the residual moisture inside the coil. This disrupts the gas-liquid equilibrium under steady-state conditions, transforming the difficult-to-detect static distribution of deep residual moisture in the coil into a measurable dynamic temperature response, thus solving the technical challenge of weak and difficult-to-capture deep moisture signals under steady-state conditions. Furthermore, this application compares the actual temperature change curve with a reference curve characterizing the drying state, extracting waveform distortion values reflecting the degree of curve distortion and temperature deviation values reflecting overall numerical differences. Based on the comprehensive relationship between these two factors, an evaporation characteristic value is determined as the basis for judging the drying state. This comprehensive evaluation method, considering both waveform morphology and numerical differences, effectively distinguishes between nonlinear temperature distortion caused by moisture evaporation and linear numerical shifts caused by model errors. It avoids misjudging model inaccuracies as moisture residue or submerging the true moisture signal in model errors, significantly improving the accuracy and reliability of judging the drying state inside the coil.
[0107] Please see Figure 2The illustration shows a schematic diagram of a high-precision temperature monitoring device for an oven used in coil drying, according to an embodiment of the present invention. This device includes: an air pressure regulation unit 201, a data acquisition unit 202, a waveform and deviation determination unit 203, an eigenvalue calculation unit 204, and a status determination unit 205. The units can communicate bidirectionally via a communication link, ensuring real-time interaction of collected data and analysis results. The communication link can employ wired or wireless transmission methods to meet the communication needs of different monitoring scenarios.
[0108] The air pressure regulating unit 201 is used to reduce the air pressure inside the oven when the coil is in the constant temperature holding stage, so as to stimulate the residual moisture inside the coil to generate phase change heat absorption.
[0109] The data acquisition unit 202 is used to acquire the actual temperature change curve of the coil during the air pressure regulation process in the oven.
[0110] The waveform and deviation determination unit 203 is used to determine the waveform distortion value and the temperature deviation value based on the actual temperature change curve and the reference temperature change curve. The reference temperature change curve is used to characterize the temperature change of the coil in the drying state during the air pressure adjustment process in the oven. The waveform distortion value is used to characterize the degree of distortion of the shape of the actual temperature change curve relative to the reference temperature change curve. The temperature deviation value is used to characterize the overall difference between the first time series temperature data corresponding to the actual temperature change curve and the second time series temperature data corresponding to the reference temperature change curve.
[0111] The eigenvalue calculation unit 204 is used to determine the evaporation characteristic value of the coil based on the waveform distortion value and the temperature deviation value.
[0112] The state determination unit 205 is used to determine the dry state of the coil based on the evaporation characteristic value.
[0113] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0114] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.
Claims
1. A method for high precision monitoring of oven temperature for coil drying, characterized by, The method includes: The coil inside the oven is in a constant temperature holding stage. The air pressure inside the oven is reduced to stimulate the residual moisture inside the coil to generate phase change endothermic reaction, and the actual temperature change curve of the coil during the air pressure adjustment process inside the oven is obtained. Based on the actual temperature change curve and the reference temperature change curve, waveform distortion value and temperature deviation value are determined respectively; the waveform distortion value is used to characterize the degree of shape distortion of the actual temperature change curve relative to the reference temperature change curve, and the temperature deviation value is used to characterize the overall difference between the first time series temperature data corresponding to the actual temperature change curve and the second time series temperature data corresponding to the reference temperature change curve. Based on the waveform distortion value and the temperature deviation value, the evaporation characteristic value of the coil is determined, and based on the evaporation characteristic value, the dryness state of the coil is determined.
2. The high-precision temperature monitoring method for an oven used for coil drying according to claim 1, characterized in that, The process of determining the waveform distortion value includes: A two-dimensional distance matrix is determined based on the first time series temperature data and the second time series temperature data; the distance matrix is used to characterize the local distance between target point pairs of data, and the target point pairs of data are the data corresponding to the same time point in the first time series temperature data and the second time series temperature data. Based on the two-dimensional distance matrix, determine the minimum cumulative path for aligning the actual temperature change curve with the reference temperature change curve; Determine the total path distance and path length corresponding to the minimum cumulative path; The waveform distortion value is determined based on the total path distance and the path length.
3. The high-precision temperature monitoring method for an oven used for coil drying according to claim 1, characterized in that, The process of determining the temperature deviation value includes: Determine the root mean square error of the first time series temperature data and the second time series temperature data; The root mean square error is determined as the temperature deviation value.
4. The high-precision temperature monitoring method for an oven used for coil drying according to claim 1, characterized in that, The process of obtaining the actual temperature change curve of the coil during the gas pressure regulation process in the oven includes: Acquire time-series air pressure data of the coil during the air pressure regulation process inside the oven; Perform a first-order difference operation on the first time series temperature data to obtain temperature change rate series data, and perform a first-order difference operation on the time series air pressure data to obtain air pressure change rate series data; Based on the temperature change rate sequence data and the air pressure change rate sequence data, determine the target correlation between the temperature of the coil and the air pressure change in the oven; If the target correlation degree exceeds a preset correlation degree threshold, the actual temperature change curve is obtained.
5. The high-precision temperature monitoring method for an oven used for coil drying according to claim 4, characterized in that, The step of determining the target correlation between the temperature of the coil and the air pressure change in the oven based on the temperature change rate sequence data and the air pressure change rate sequence data includes: The temperature change rate sequence data is adjusted according to a preset hysteresis step size to obtain the adjusted temperature change rate sequence data. Calculate the candidate correlation degree between the adjusted temperature change rate sequence data and the air pressure change rate sequence data, and determine the candidate correlation degree as the target correlation degree if the candidate correlation degree meets the correlation conditions.
6. The high-precision temperature monitoring method for an oven used for coil drying according to claim 4, characterized in that, The method further includes: Based on the temperature change rate sequence data and the air pressure change rate sequence data, the temperature delay response time is determined; the temperature delay response time is the delay time in which the temperature of the coil follows the change in air pressure inside the oven. The temperature data between the pressure adjustment time and the temperature delay response time is determined as the second time series temperature data; the pressure adjustment time is the time when the rate of change of the air pressure in the oven is the largest or the time when the air pressure in the oven drops to the target air pressure. The reference temperature change curve is determined based on the second time series temperature data.
7. The high-precision temperature monitoring method for an oven used for coil drying according to any one of claims 1-6, characterized in that, Determining the dryness state of the coil based on the evaporation characteristic value includes: If the evaporation characteristic value does not exceed a preset evaporation characteristic threshold, the coil is determined to be in a dried state. If the evaporation characteristic value exceeds a preset evaporation characteristic threshold, the coil is determined to be in a state to be dried.
8. The high-precision temperature monitoring method for an oven used for coil drying according to claim 7, characterized in that, The method further includes: When the coil is in a state to be dried, the drying strategy of the coil is determined based on the temperature delay response time.
9. The high-precision temperature monitoring method for an oven used for coil drying according to claim 8, characterized in that, The step of determining the drying strategy for the coil based on the temperature delay response time includes: If the temperature delay response time does not exceed a preset time threshold, it is determined that the distribution depth of moisture in the coil does not exceed a preset depth threshold, and a first drying strategy is executed; the first drying strategy is used to instruct the oven to dry the coil according to the current drying instruction; If the temperature delay response time exceeds the preset time threshold, it is determined that the distribution depth of moisture in the coil exceeds the preset depth threshold, and a second drying strategy is executed; the second drying strategy is used to instruct the oven to extend the drying time of the coil and / or periodically adjust the air pressure in the oven.
10. A high-precision temperature monitoring device for an oven used in coil drying, characterized in that, include: The air pressure regulating unit is used to reduce the air pressure inside the oven when the coil inside the oven is in the constant temperature holding stage, so as to stimulate the residual moisture inside the coil to generate phase change endothermic reaction. The data acquisition unit is used to acquire the actual temperature change curve of the coil during the air pressure adjustment process in the oven; The waveform and deviation determination unit is used to determine the waveform distortion value based on the actual temperature change curve and the reference temperature change curve, and to determine the temperature deviation value based on the actual temperature change curve and the reference temperature change curve; the waveform distortion value is used to characterize the degree of morphological distortion of the actual temperature change curve relative to the reference temperature change curve, and the temperature deviation value is used to characterize the overall difference between the first time series temperature data corresponding to the actual temperature change curve and the second time series temperature data corresponding to the reference temperature change curve; The eigenvalue calculation unit is used to determine the evaporation characteristic value of the coil based on the waveform distortion value and the temperature deviation value. A state determination unit is used to determine the dry state of the coil based on the evaporation characteristic value.