A method for dynamically monitoring pressure rise rate of vacuum furnace and locating leakage point

Abstract

The application discloses a kind of vacuum furnace pressure rise rate dynamic monitoring and leak point partition positioning method, specifically related to the technical field of using pressure change to carry out tightness test, for solving the problem that in the existing partition isolation pressure rise rate test, explosive desorption gas pollution hearth background baseline when pipeline access, resulting in the problem of inaccurate judgment of leak point attribution.This method eliminates the interference of additional air release on baseline by individually pre-extracting the pipeline to be tested and using the morphological turning point of pressure decay curve as the termination criterion;By identifying the turning point of the pressure recovery rate in the integrated space from decay to constant stability and mapping it to the normalized time axis to calculate the phase offset, the influence of individual differences of the hearth on the comparison result is eliminated;By separating the trend and residual from the phase offset sequence obtained by multiple tests and identifying the constant positive component, the difference in timing behavior between pipeline true leakage and residual air release is distinguished, and accurate determination of the leakage state is realized.

Description

Technical Field

[0001] This invention relates to the technical field of sealing tests using pressure changes, and more specifically, to a method for dynamic monitoring of pressure rise rate and location of leaks in a vacuum furnace. Background Technology

[0002] In industrial applications of vacuum heat treatment furnaces, to reduce equipment downtime and quickly identify leaks affecting sealing, a common method is to conduct pressure rise rate tests using a zoned isolation approach through the furnace's built-in vacuum gauge and valve system. The general process is as follows: First, the furnace chamber is evacuated to a certain vacuum level, then the main valve is closed, and the pressure rise rate of the furnace chamber is measured as a baseline. Subsequently, valves connecting different bypass pipes to the furnace chamber are opened sequentially, connecting the corresponding pipes to the vacuum system and evacuating again. The total pressure rise rate of the space integrating the furnace chamber and that section of pipe is measured. By comparing the increase in the total pressure rise rate after connecting a particular pipe with the pressure rise rate of the furnace chamber itself, a leak in that pipe is determined. This method eliminates the need for external leak detection equipment; the system's own instruments can achieve preliminary zoned location of leaks.

[0003] However, during the implementation of the aforementioned step-by-step zoned testing, it was found that when the valves were opened and closed to connect each bypass pipeline to the vacuum-equipped furnace, the inner walls of those previously isolated pipelines experienced a sharp drop in pressure for the first time. This caused an explosive desorption of the adsorbed gas adhering to their surfaces. This gas, released at the moment the valves were opened, was difficult to fully expel during the subsequent short evacuation cycle. A significant proportion of it was re-adsorbed and trapped by the adjacent, negatively pressured furnace walls and insulation layer. This additional venting effect introduced by the testing process alters the baseline of the furnace's venting state during testing of different pipeline sections. This causes a non-uniform shift in the pressure rise rate comparison benchmark used for leak location during the testing process, thus interfering with the accuracy of determining the attribution of the leaking pipeline. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a method for dynamic monitoring of vacuum furnace pressure rise rate and leakage point location to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for dynamic monitoring of pressure rise rate and zoning of leak points in a vacuum furnace includes the following steps: S1: After evacuating the furnace, close the main valve and collect the dynamic trend of the furnace pressure rise rate; S2: Close the connecting valve between the pipeline under test and the furnace, and evacuate the pipeline under test separately until the pressure decay curve of the pipeline under test passes the morphological inflection point and tends to stabilize. S3: Open the connecting valve and evacuate the integrated space formed by the furnace and the pipeline under test. After closing the main valve, identify the turning point when the pressure recovery rate in the integrated space changes from a decaying trend to a constant trend. S4: Extract the furnace venting decay time constant from the dynamic trend of furnace pressure rise rate, map the turning point to the normalized time axis based on the furnace venting decay time constant, and calculate the phase offset of the turning point relative to the expected equilibrium position on the normalized time axis. S5: Analyze the behavior of phase offset over time and identify whether the phase offset contains a constant positive component that does not decay over time. S6: Determine the leakage status of the pipeline under test based on whether a constant positive component is identified.

[0006] Furthermore, after evacuating the furnace and closing the main valve, the dynamic trend of the furnace pressure rise rate is collected, including: recording the furnace pressure value multiple times during the observation period after closing the main valve, dividing the difference between each recorded pressure value and the adjacent previous recorded pressure value by the recording interval to obtain the pressure rise rate sequence, identifying the segment boundary points where the change characteristics in the pressure rise rate sequence change, dividing the observation period into the first segment and the second segment using the segment boundary points, and taking the pressure rise rate sequence in the first segment and the pressure rise rate sequence in the second segment together to constitute the dynamic trend of the furnace pressure rise rate.

[0007] Furthermore, identifying segmentation boundaries includes: successively subtracting adjacent pressure rise rate values ​​in the pressure rise rate sequence to obtain a first-order difference sequence; successively subtracting adjacent first-order difference values ​​in the first-order difference sequence to obtain a second-order difference sequence; and taking the recording time corresponding to the first second-order difference value in the second-order difference sequence that changes from a biased zero-value fluctuation to a positive fluctuation range exceeding zero-value fluctuation as the segmentation boundary.

[0008] Furthermore, the connecting valve between the pipeline under test and the furnace is closed, and the pipeline under test is evacuated separately until the pressure decay curve of the pipeline under test passes the morphological inflection point and tends to stabilize. This includes: continuously recording the pressure value inside the pipeline under test during the process of evacuating the pipeline under test separately, calculating the pressure decay rate between adjacent recording times, identifying the transition zone where the pressure decay rate changes from rapid decay to gradual decay, determining the moment when the rate of change of pressure decay rate in the transition zone first falls below the preset rate of change threshold as the morphological inflection point, and determining that the pressure decay rate fluctuation amplitude of the continuously recorded pressure decay rate after the morphological inflection point is maintained within the specified range.

[0009] Furthermore, the moment when the rate of change of pressure decay rate in the transition zone first falls below the preset rate of change threshold is determined as the morphological inflection point, including: taking the average rate of change of pressure decay rate between the moment when the pipeline under test starts to be evacuated and the moment when the rate of change of pressure decay rate enters the rapid decay stage as the reference rate of change, multiplying the reference rate of change by the proportional coefficient as the preset rate of change threshold, and taking the moment when the rate of change of pressure decay rate in the transition zone first falls below the preset rate of change threshold as the morphological inflection point.

[0010] Furthermore, the connecting valve is opened and the integrated space formed by the furnace and the pipeline under test is evacuated. After the main valve is closed, the turning point when the pressure recovery rate in the integrated space changes from a decreasing trend to a constant trend is identified. This includes: continuously recording the pressure value in the integrated space during the observation period after the main valve is closed, calculating the pressure recovery rate between adjacent recording times to form a pressure recovery rate sequence, identifying the transition moment in the pressure recovery rate sequence from continuous decrease to fluctuation within a fixed amplitude range, and taking the transition moment as the turning point.

[0011] Furthermore, identifying the transition moment in the pressure recovery rate sequence from a continuous decrease to fluctuation within a fixed amplitude range includes: calculating the sliding standard deviation of several consecutive pressure recovery rate values ​​in the pressure recovery rate sequence; when the starting time of the time window corresponding to the first sliding standard deviation in the sliding standard deviation sequence that is lower than the standard deviation threshold, and the pressure recovery rate sequence after this time window no longer shows a continuous decreasing trend, the starting time of this time window is taken as the transition moment.

[0012] Furthermore, the furnace venting attenuation time constant is extracted from the dynamic trend of the furnace pressure rise rate. The inflection point is mapped to a normalized time axis based on the furnace venting attenuation time constant. The phase offset of the inflection point relative to the expected equilibrium position on the normalized time axis is calculated. This includes: performing attenuation fitting on the pressure rise rate sequence in the second segment of the furnace pressure rise rate dynamic trend to obtain a second attenuation curve and extracting the furnace venting attenuation time constant; constructing a normalized time axis with the furnace venting attenuation time constant as the unit scale; performing attenuation fitting on the pressure rise rate sequence in the first segment of the furnace pressure rise rate dynamic trend to obtain a first attenuation curve; taking the intersection of the asymptotes of the first attenuation curve and the second attenuation curve on the normalized time axis as the expected equilibrium position; mapping the inflection point to the normalized time axis to obtain the normalized time position; and taking the difference between the normalized time position and the expected equilibrium position as the phase offset.

[0013] Furthermore, the behavior of the phase offset over time is analyzed to identify whether the phase offset contains a constant positive component that does not decay over time. This includes: arranging multiple phase offsets obtained from executing S1 to S4 multiple times according to the execution sequence to form a phase offset sequence; splitting the phase offset sequence into a trend sequence and a residual sequence; determining whether the sign of the trend sequence is consistently positive and the amplitude is stable or increasing; and simultaneously determining whether the fluctuation characteristics of the residual sequence are random distributions around zero. When the trend sequence is consistently positive and the amplitude is stable or increasing, and the residual sequence is a random distribution around zero, it is identified that the phase offset contains a constant positive component that does not decay over time.

[0014] Furthermore, the leakage status of the pipeline under test is determined based on whether a constant positive component is identified, including: when a constant positive component is identified, the pipeline under test is determined to have a leak; when a constant positive component is not identified, the pipeline under test is determined to have no leak.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. This method performs an independent pre-vacuuming operation on the pipeline under test before it is connected to the furnace, and uses the inflection point of the pressure decay curve as the criterion for the pre-vacuuming endpoint. This ensures that the explosive desorbed gas generated by the initial exposure of the pipeline inner wall to vacuum is fully removed in an isolated state, fundamentally preventing the transfer and re-adsorption of additional venting into the furnace. This ensures the consistency of the furnace background venting baseline during subsequent integrated space testing and eliminates the interference of baseline offset introduced by the testing process itself on leak attribution. The determination of the pre-vacuuming endpoint is based on the physical state transition of the pipeline's own venting process rather than a fixed duration, which can automatically adapt to differences in the volume and inner wall adsorption of different pipelines, balancing the adequacy of pre-vacuuming with testing efficiency.

[0016] 2. In the integrated space testing phase, by identifying the turning point where the pressure recovery rate changes from a decaying trend to a constant trend, and constructing a normalized time axis based on the furnace venting decay time constant, the turning point is mapped as a phase offset. This unifies the dynamic responses of different furnaces and under different testing conditions onto the same scale with the furnace's own venting rhythm as a reference, eliminating the influence of individual furnace differences on the comparison results. By separating the trend and residual of the phase offset sequence obtained from multiple tests, it is identified whether it contains a constant positive component that does not decay over time. This distinguishes the two physical processes, true pipeline leakage and residual venting, which are easily confused in terms of pressure rise rate values, in terms of temporal behavior, thereby achieving an accurate determination of the pipeline leakage status. Attached Figure Description

[0017] Figure 1 This is a flowchart of a method for dynamic monitoring of pressure rise rate and leak location in a vacuum furnace according to the present invention; Figure 2This is a schematic diagram of the connection structure of the vacuum furnace chamber and the pipeline to be tested in this invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] Example: Figure 1 This invention presents a method for dynamic monitoring of pressure rise rate and leak location in a vacuum furnace, comprising the following steps: S1: After evacuating the furnace, close the main valve and collect the dynamic trend of the furnace pressure rise rate; S2: Close the connecting valve between the pipeline under test and the furnace, and evacuate the pipeline under test separately until the pressure decay curve of the pipeline under test passes the morphological inflection point and tends to stabilize. S3: Open the connecting valve and evacuate the integrated space formed by the furnace and the pipeline under test. After closing the main valve, identify the turning point when the pressure recovery rate in the integrated space changes from a decaying trend to a constant trend. S4: Extract the furnace venting decay time constant from the dynamic trend of furnace pressure rise rate, map the turning point to the normalized time axis based on the furnace venting decay time constant, and calculate the phase offset of the turning point relative to the expected equilibrium position on the normalized time axis. S5: Analyze the behavior of phase offset over time and identify whether the phase offset contains a constant positive component that does not decay over time. S6: Determine the leakage status of the pipeline under test based on whether a constant positive component is identified.

[0020] Figure 2 A schematic diagram of the connection structure of the vacuum furnace chamber and the pipeline under test according to the present invention is provided. The furnace chamber is connected to one end of the main valve via a pipe, and the other end of the main valve is connected to a vacuum pump unit via a pipe. The vacuum pump unit is used to evacuate the furnace chamber. The furnace chamber is connected to one end of a connecting valve via a pipe, and the other end of the connecting valve is connected to the pipeline under test via a pipe. The other end of the pipeline under test is either closed or connected to a closing valve. When the connecting valve is open, the furnace chamber and the pipeline under test form an integrated space; when the connecting valve is closed, the pipeline under test is in an isolated state. A pressure sensor is installed on the pipe between the connecting valve and the pipeline under test to record the pressure value inside the pipeline under test when the pipeline under test is evacuated separately. A pressure sensor is installed on the furnace chamber to record the furnace chamber pressure value and the pressure value within the integrated space after the main valve is closed.

[0021] In the specific implementation of S1, a vacuum operation is performed on the furnace. The vacuum pump unit is started and the main valve is opened to reduce the internal pressure of the furnace to the required background vacuum level. Then, the main valve is closed, and the observation period begins from the moment the main valve is closed. During the observation period, the furnace pressure value is recorded at equal intervals. The interval between two adjacent records is set to a fixed value, and the pressure value recorded each time is recorded as the furnace pressure value. The difference between each recorded furnace pressure value and the furnace pressure value of the previous record is divided by the recording interval to obtain the pressure rise rate corresponding to that recording moment. All pressure rise rates are arranged in the recording sequence to form a pressure rise rate sequence.

[0022] Identify the turning points in the pressure rise rate sequence where the change characteristics change. Subtract adjacent pressure rise rate values ​​sequentially to obtain a first-order difference sequence. Each first-order difference value in the first-order difference sequence reflects the direction and magnitude of the pressure rise rate change. Subtract adjacent first-order difference values ​​sequentially to obtain a second-order difference sequence. Each second-order difference value in the second-order difference sequence reflects the acceleration characteristics of the pressure rise rate change behavior. In the early part of the observation period, the venting process of the furnace inner wall and insulation material after the main valve is closed dominates the pressure recovery. At this time, the venting rate decreases with time, and the pressure rise rate sequence shows a continuous downward trend. The corresponding first-order difference sequence is negative, while the second-order difference sequence fluctuates slightly around zero, i.e., it is biased towards zero. As the observation period progresses, once the furnace venting decreases to a certain extent, if a true leak exists in the furnace, the constant intake volume contributed by the leak will cause the pressure rise rate sequence to stop decreasing and may even increase. The first-order difference sequence will turn from negative to positive, and a value exceeding the zero-value fluctuation range will appear in the second-order difference sequence. The zero-value fluctuation range is determined based on the statistical distribution of data in the second-order difference sequence that falls into the zero-value fluctuation stage. The boundary of the zero-value fluctuation range is calculated as follows: ; in, Indicates the boundary of the zero-value fluctuation range. Indicates the preset multiple. It represents the standard deviation of a series of consecutive second-order difference values ​​that have been confirmed to be in a phase of fluctuation towards zero in a second-order difference sequence. and All quantities are dimensionless. Preset multiple. The selection is based on the furnace chamber volume and the background vacuum level, for example, taking... The value is 3. The recording time corresponding to the first second-order difference value in the second-order difference sequence that changes from a biased zero-value fluctuation to a positive fluctuation range exceeding zero-value fluctuation is taken as the segmentation boundary.

[0023] The observation period is divided into a first segment and a second segment, with the segment boundary point serving as the dividing line. The first segment is the time interval from the moment the main valve is closed to the segment boundary point, and the second segment is the time interval from the segment boundary point to the end of the observation period. The pressure rise rate sequences within the first segment and the pressure rise rate sequences within the second segment together constitute the dynamic trend of the pressure rise rate in the furnace.

[0024] Using the fluctuation characteristics of the second-order difference sequence as the basis for segmentation, this study leverages the fundamental differences in pressure recovery behavior between the furnace venting process and the actual leakage process. The pressure recovery rate contributed by furnace venting decays over time, exhibiting relatively stable behavior and manifesting as random fluctuations near zero in the second-order difference sequence. In contrast, the pressure recovery rate contributed by actual leakage remains essentially constant. When venting decays to a level comparable to leakage, the steady contribution of leakage gradually dominates the direction of pressure rise, manifesting as a positive shift that breaks through the original fluctuation range in the second-order difference sequence. By capturing the first occurrence of this shift to segment the sequence, it is possible to separate the stage in the pressure rise sequence mainly dominated by venting from the stage where venting and leakage compete significantly. The pressure rise sequence in the first segment mainly reflects the decay characteristics of furnace venting, while the pressure rise sequence in the second segment contains more information about leakage contributions. Together, they constitute the dynamic trend of furnace pressure rise, providing a complete benchmark for subsequent extraction of the furnace venting decay time constant and comparison of the dynamic response of the integrated space.

[0025] In the specific implementation of S2, the connecting valve between the pipeline under test and the furnace is closed, and the pipeline under test is evacuated separately. The vacuum pump unit is started to evacuate the isolated pipeline under test. During the evacuation process, the pressure value inside the pipeline under test is continuously recorded at equal intervals. The interval between two adjacent records is uniformly set to a fixed value, and the pressure value recorded each time is recorded as the pressure value of the pipeline under test. The difference between each recorded pressure value and the pressure value of the pipeline under test recorded next to the previous recorded pressure value is divided by the recording interval, and the absolute value is taken to obtain the pressure decay rate corresponding to that recording time. All pressure decay rates are arranged in the recording time sequence to form a pressure decay rate sequence.

[0026] In the pressure decay rate sequence, a transition region from rapid decay to gradual decay is identified. Initially, during the initial vacuuming of the pipeline under test, the gas adsorbed on the pipeline wall desorbs rapidly upon first exposure to the vacuum environment. At this time, the gas is quickly removed, and the pressure decay rate is high and fluctuates dramatically; this stage is the rapid decay stage. As vacuuming continues, the adsorbed gas on the pipeline wall gradually decreases, the release of pressure energy slows down, and the pressure decay rate continuously decreases from a high level, gradually entering a stage of gradual change; this stage is the gradual decay stage. Between the rapid decay stage and the gradual decay stage, there exists a continuous recording interval where the pressure decay rate change characteristics show a clear turning point; this continuous recording interval is identified as the transition region. Specifically, the transition region is identified by calculating the difference between adjacent pressure decay rate values ​​in the pressure decay rate sequence to obtain the pressure decay rate change rate sequence. The segment in the pressure decay rate change rate sequence that transitions from large fluctuations to a more stable state is considered the transition region.

[0027] The moment when the rate of change of pressure decay in the transition zone first falls below a preset rate of change threshold is defined as the morphological inflection point. The preset rate of change threshold is set by averaging the rate of change of pressure decay from the moment the pipeline under test begins evacuation to the moment the pressure decay rate enters the rapid decay phase; this average is used as the baseline rate of change. The start time of the rapid decay phase is determined when the pressure decay rate sequence first exceeds a preset multiple of the baseline decay rate. The baseline decay rate is the average of the pressure decay rates at the first few recorded moments after the pipeline under test begins evacuation. The preset rate of change threshold is calculated by multiplying the baseline rate of change by a proportionality coefficient. ;in, This indicates the preset rate of change threshold. This represents the proportionality coefficient. Indicates the rate of change of the benchmark. and The units are all pressure units divided by the square of time. (Proportionality coefficient) The selection is based on the volume of the pipeline to be tested and the venting characteristics of the pipeline inner wall material, for example, taking... The value is between 0.2 and 0.5. Within the transition zone, the rate of change of pressure decay rate is compared with the preset rate of change threshold sequentially according to the recording time sequence. The recording time when the rate of change of pressure decay rate first falls below the preset rate of change threshold is taken as the morphological inflection point.

[0028] proportionality coefficient The selection of the pressure decay rate is based on the combination of the test pipeline volume and the outgassing characteristics of the pipeline inner wall material. The test pipeline volume determines the total surface area of ​​the pipeline inner wall, and the outgassing characteristics of the pipeline inner wall material determine the initial outgassing intensity per unit area when first exposed to vacuum. When the test pipeline volume is small and the outgassing characteristics of the pipeline inner wall material are weak, the transition between the rapid decay phase and the gradual decay phase of the pressure decay rate is relatively short, and the decrease in the rate of change of pressure decay rate near the morphological inflection point is significant, with a high proportionality coefficient. A larger value can be chosen, such as close to 0.5, to avoid misidentifying normal fluctuations during the rapid decay phase as morphological inflection points. When the volume of the pipeline under test is large or the gas release characteristics of the pipeline wall material are strong, the transition of the pressure decay rate between the rapid and gradual decay phases is relatively smooth, and the decrease in the rate of change of pressure decay rate near the morphological inflection point is relatively small, resulting in a smaller proportionality coefficient. A smaller value should be chosen, such as close to 0.2, to ensure that the morphological inflection point can be accurately identified within the transition zone without being missed. In practical applications, for common stainless steel pipelines, if the volume of the pipeline under test is less than one-fiftieth of the furnace volume, the proportionality coefficient should be... The value can be selected between 0.35 and 0.5. If the volume of the pipeline to be tested is greater than one-fiftieth of the furnace volume, the proportionality coefficient should be adjusted accordingly. The value can be selected between 0.2 and 0.35. Those skilled in the art can determine the appropriate value within the above range according to the specific specifications of the pipeline to be tested.

[0029] When the fluctuation amplitude of the pressure decay rate recorded continuously after the morphological inflection point remains within the specified range, the pressure decay curve of the pipeline under test is determined to be stable. The specified range is determined based on the statistical characteristics of the pressure decay rate during the gradual decay phase. The average of several consecutive pressure decay rate values ​​after the morphological inflection point is taken as the stable reference value, and a preset percentage fluctuation above and below the stable reference value is used as the upper and lower limits of the specified range, for example, a preset percentage of 10% to 20%. When all the pressure decay rate values ​​recorded continuously after the morphological inflection point fall within the specified range, the pressure decay curve of the pipeline under test is determined to be stable, and at this time, individual vacuuming of the pipeline under test is stopped.

[0030] The aforementioned method of individually evacuating the pipeline under test and using the morphological inflection point as the criterion for determining the operational endpoint, compared to the method using a fixed duration as the evacuation endpoint, can accurately correspond to the physical critical position where the explosive desorption process on the pipeline inner wall transitions from intense to gradual. The fixed duration method cannot adapt to differences in pipeline volume and inner wall adsorption state. If the evacuation time is insufficient, residual explosive desorption gas will rush into the furnace when the connecting valve is opened, and be re-adsorbed by the furnace inner wall and insulation layer, altering the background venting baseline of the furnace during integrated space testing. If the evacuation time is too long, it increases unnecessary equipment occupancy time. Using the first time the rate of change of pressure decay falls below a preset threshold as the criterion for the morphological inflection point can automatically adapt to the actual venting process of the pipeline under test. When the impact effect of explosive desorption tends to subside and the pressure decay enters a gradual phase, it is determined that the pre-evacuation is sufficient, eliminating the interference of additional venting effects on the subsequent integrated space testing baseline from the source.

[0031] In the specific implementation of S3, the connecting valve is opened and the integrated space formed by the furnace and the pipeline under test is evacuated. The vacuum pump group is started to evacuate the integrated space until the internal pressure is reduced to the background vacuum level required by the process. Then, the main valve is closed, and the observation period begins from the moment the main valve is closed. During the observation period, the pressure value in the integrated space is continuously recorded at equal intervals. The interval between two adjacent records is uniformly set to a fixed value, and the pressure value recorded each time is recorded as the integrated space pressure value. The difference between each recorded integrated space pressure value and the pressure value of the previous adjacent integrated space record is divided by the recording interval to obtain the pressure recovery rate corresponding to that recording moment. All pressure recovery rates are arranged in the recording sequence to form a pressure recovery rate sequence.

[0032] The transition point in the pressure recovery rate sequence, from a continuous decrease to fluctuation within a fixed amplitude range, is identified. Initially after the main valve is closed, the venting process of the integrated space wall and insulation material dominates the pressure recovery, with the venting rate decreasing over time, resulting in a continuously decreasing pressure recovery rate sequence. As the observation period progresses, the contribution of venting gradually weakens. If a true leak exists in the tested pipeline, the constant intake volume contributed by the leak will gradually become the dominant factor in pressure recovery, and the pressure recovery rate sequence will no longer continuously decrease but will fluctuate within a fixed amplitude range. The point of transition between the continuously decreasing trend and fluctuation within the fixed amplitude range is the transition point.

[0033] The transition moment is identified by calculating the moving standard deviation of several consecutive pressure rise rate values ​​in the pressure rise rate sequence. A fixed-length time window is set, the length of which is determined by the total number of records in the observation period, for example, one-tenth to one-fifth of the total number of records. Starting from the beginning of the pressure rise rate sequence, the time window is slid forward successively, each time for one recording interval. The standard deviation of several pressure rise rate values ​​within the time window after each slide is calculated. The calculated standard deviations are then arranged according to the starting time of the time window to form a moving standard deviation sequence.

[0034] The sliding standard deviation is calculated as follows: ;in, Indicates the sliding standard deviation. This indicates the number of pressure recovery rate values ​​within the time window. Indicating the first [number] in the pressure recovery rate sequence A pressure recovery rate value, Indicates the time window The average of the pressure recovery rate values For serial number, and The dimensions of all of them are dimensionless. The dimensions of the pressure recovery rate are the same. This indicates the time window from the 1st to the 1st. Sum the pressure recovery rate values.

[0035] The standard deviation threshold is determined based on the statistical characteristics of the moving standard deviation sequence. The mean of several consecutive moving standard deviations in the early part of the sequence that show a decreasing trend is taken as the standard deviation threshold. When the first moving standard deviation below the threshold appears in the sequence, the starting time of the time window corresponding to this first instance is taken as the candidate transition time. A trend test is performed on the pressure recovery rate sequence after the candidate transition time. If the pressure recovery rate sequence after the candidate transition time no longer shows a continuously decreasing trend, the candidate transition time is confirmed as a transition time, and this transition time is taken as the turning point. If the pressure recovery rate sequence after the candidate transition time still shows a continuously decreasing trend, the time window is continued to slide forward to find the next moving standard deviation below the threshold, and the trend test is repeated.

[0036] The method for defining a series of consecutive moving standard deviations (MSDs) in a moving standard deviation sequence that are at the beginning and exhibit a downward trend is as follows: Starting from the beginning of the moving standard deviation sequence, examine the relationship between each moving standard deviation and its subsequent adjacent moving standard deviations along the sequence direction. Identify segments where subsequent moving standard deviations are consistently less than the current moving standard deviation as downward segments. Take all moving standard deviations within the first downward segment of the moving standard deviation sequence as the series of consecutive moving standard deviations at the beginning and exhibiting a downward trend. If several moving standard deviations at the beginning of the moving standard deviation sequence are already at a low level and show no obvious downward trend, then take the first few moving standard deviations of the sequence as the definition object. The number of the first few is between one-fifth and one-quarter of the total length of the moving standard deviation sequence. The standard deviation threshold is calculated as the arithmetic mean of the defined series of consecutive moving standard deviations. The above definition method utilizes the changing pattern of pressure recovery rate fluctuation characteristics after the main valve is closed. In the early part of the observation period, the venting process dominates, and the fluctuation amplitude of the pressure recovery rate gradually narrows as the venting decays. The sliding standard deviation shows a downward trend accordingly. By automatically identifying the downward segment, the data range used to calculate the standard deviation threshold is defined, which can adapt to the morphological differences of the sliding standard deviation sequence under different furnace volumes and vacuuming process conditions.

[0037] The transition point is identified by using the sliding standard deviation, leveraging the fundamental difference in pressure recovery rate fluctuations between the venting decay process and the leakage steady-state process. During the venting decay phase, the pressure recovery rate continuously decreases, with significant fluctuations in the pressure recovery rate values ​​between adjacent time periods, resulting in a relatively high and decreasing sliding standard deviation. During the leakage steady-state phase, the pressure recovery rate fluctuates within a fixed amplitude range, tending to stabilize between adjacent time periods, and the corresponding sliding standard deviation drops to a low level and remains stable. Capturing this transition point by identifying the first moment in the sliding standard deviation sequence that falls below the standard deviation threshold accurately identifies the critical point of equilibrium between venting and leakage, providing a clear time reference for subsequent phase shift calculations.

[0038] In the specific implementation of S4, the furnace venting attenuation time constant is extracted from the dynamic trend of the furnace pressure rise rate. The dynamic trend of the furnace pressure rise rate is composed of the pressure rise rate sequence in the first segment and the pressure rise rate sequence in the second segment, where the pressure rise rate sequence in the second segment mainly reflects the furnace venting attenuation characteristics. Attenuation fitting is performed on the pressure rise rate sequence in the second segment, using an exponential decay function. The fitting function is: ;in, This indicates the pressure rise rate sequence within the second segment at time [time value missing]. The fitted value, Indicates the initial amplitude. Indicates time, This represents the furnace venting decay time constant. The dimensions of are the same as those of the pressure rise rate. and All dimensions are time units. During the fitting process, the actual pressure rise rate at each recorded moment in the pressure rise rate sequence within the second segment is used as the fitting target. The value of the furnace venting decay time constant τ is determined by minimizing the deviation between the fitted value and the actual value. The fitting deviation is calculated as the sum of the squares of the differences between the fitted value and the actual value at each recorded moment.

[0039] A normalized time axis is constructed using the extracted furnace venting decay time constant τ as the unit scale. The normalized time axis converts the actual time into a dimensionless time based on the furnace venting decay time constant. The scale value on the normalized time axis is calculated as: tnorm = tactual / τ; where tnorm represents the normalized time position, tactual represents the actual time, τ represents the furnace venting decay time constant, and tnorm is dimensionless.

[0040] A decay fitting was performed on the pressure rise rate sequence within the first segment of the dynamic trend of the furnace pressure rise rate. This first segment primarily reflects the transitional characteristics of the competition between venting and leakage within the furnace. The decay fitting also adopted an exponential decay function, and the fitting function is as follows: ;in, This indicates the pressure rise rate sequence within the first segment at time [time value missing]. The fitted value, Indicates the initial amplitude. This represents the decay time constant of the first section. The dimensions of are the same as those of the pressure rise rate. and All dimensions are time-based. During the fitting process, the actual pressure rise rate at each recorded moment in the pressure rise rate sequence within the first segment is used as the fitting target, and the target value is determined by minimizing the deviation between the fitted value and the actual value. and The value of is used to obtain the first attenuation curve.

[0041] The first decay curve is mapped to a normalized time axis, and its asymptote position is determined on the normalized time axis. As the normalized time position approaches infinity on the normalized time axis, the function value of the first decay curve approaches a constant value; this constant value is the asymptote position of the first decay curve, and its dimension is the same as the pressure rise rate. The second decay curve is mapped to the normalized time axis; the second decay curve is the representation of the decay fitting function of the pressure rise rate sequence within the second segment on the normalized time axis. The asymptote position of the second decay curve is determined on the normalized time axis, and as the normalized time position approaches infinity on the normalized time axis, the function value of the second decay curve approaches a constant value; this constant value is the asymptote position of the second decay curve. The difference between the asymptote positions of the first and second decay curves is taken as the asymptote difference value, and the normalized time position corresponding to this asymptote difference value is taken as the expected equilibrium position on the normalized time axis. The expected equilibrium position is determined by dividing the asymptote difference by the absolute value of the slope of the decay fitting function of the pressure rise rate sequence in the second segment at the origin of the normalized time axis, thus obtaining the normalized time position corresponding to the expected equilibrium position.

[0042] The asymptote of the first decay curve is the constant value that Q(t) approaches when tnorm approaches infinity, denoted as Qasym1; the asymptote of the second decay curve is the constant value that P(t) approaches when tnorm approaches infinity, denoted as Pasym2.

[0043] The formula for calculating the asymptote difference Δasym is: Δasym = Qasym1 - Pasym2; where Δasym represents the asymptote difference, Qasym1 represents the asymptote position of the first decay curve, and Pasym2 represents the asymptote position of the second decay curve. The dimensions of Δasym, Qasym1, and Pasym2 are the same as the dimensions of the pressure rise rate.

[0044] The absolute slope of the decay fitting function P(t) of the pressure rise rate sequence in the second segment at the origin of the normalized time axis is the absolute value of the first derivative of P(t) with respect to tnorm when tnorm equals zero, denoted as |P'(0)|. Since P(t) adopts the form of an exponential decay function P(t) = A × e (-t / τ) And tnorm=t / τ, the absolute value of the slope of the derivative of P(t) with respect to tnorm at the origin of the normalized time axis is denoted as SE, and the dimension of SE is the same as that of the pressure rise rate.

[0045] The expected equilibrium position tbalance is calculated as: tbalance = Δasym / |P'(0)|; where tbalance represents the expected equilibrium position, Δasym represents the asymptotic difference, and |P'(0)| represents the absolute value of the slope of the decay fitting function of the pressure rise rate sequence in the second segment at the origin of the normalized time axis. The dimensions of Δasym and |P'(0)| are both based on the pressure rise rate. After dividing the two, the dimension of tbalance is dimensionless, which is consistent with the dimension of the normalized time position on the normalized time axis. Therefore, tbalance can be directly used as the expected equilibrium position on the normalized time axis.

[0046] The physical basis of the above calculation method lies in the fact that the asymptote difference between the first and second decay curves reflects the difference in pressure rise rate between the first and second sections when the venting decay tends to stabilize. This difference mainly stems from the additional contribution of the leakage and venting competition process in the first section relative to the pure venting decay process in the second section. The absolute value of the slope of the decay fitting function of the pressure rise rate sequence in the second section at the origin of the normalized time axis characterizes the decay rate of furnace venting decay in the initial stage. Dividing the asymptote difference by this decay rate is equivalent to converting the difference in pressure rise rate level into the time offset required to produce this difference on the normalized time axis, thereby obtaining a predicted equilibrium position with a clear temporal meaning.

[0047] The inflection point identified in S3 is mapped to the normalized time axis to obtain the normalized time position. The mapping method is to divide the inflection point by the furnace venting decay time constant. The difference between the normalized time position and the expected equilibrium position is taken as the phase offset, which is calculated as: Δφ = ttransnorm - tbalance; where Δφ represents the phase offset, ttransnorm represents the normalized time position obtained by mapping the inflection point to the normalized time axis, and tbalance represents the expected equilibrium position. The dimensions of Δφ, ttransnorm, and tbalance are all dimensionless.

[0048] By unifying the dynamic behavior of pressure rise rates under different capacity vacuum furnaces and test conditions onto the same scale based on the furnace's own venting rhythm, the furnace venting decay time constant reflects the inherent decay rate of the venting process of the furnace inner wall and insulation material. Using it as a normalization benchmark can eliminate the influence of individual furnace differences on the test results. By determining the expected equilibrium position through the difference in the asymptote positions of the first and second decay curves, combined with the slope of the decay fitting function of the pressure rise rate sequence in the second segment at the origin of the normalized time axis, the expected position of the leakage and venting competition balance reflected by the change in the decay trend of furnace venting at different stages can be captured. The phase shift obtained by comparing the turning point with this expected equilibrium position characterizes the temporal shift of the actual venting and leakage competition balance relative to the expected equilibrium position, providing a unified quantitative index for subsequent analysis of the existence of the constant positive component in the phase shift.

[0049] In the specific implementation of S5, S1 to S4 are executed multiple times for the same vacuum furnace in different furnace batches or different test cycles, and a phase offset is obtained each time. The multiple phase offsets obtained from the multiple executions are arranged according to the execution sequence to form a phase offset sequence. The execution sequence is the chronological order of the executions of S1 to S4, for example, arranged in ascending order of date or furnace number.

[0050] The phase offset sequence is split into a trend sequence and a residual sequence. The trend sequence reflects the long-term trend of the phase offset over execution time, while the residual sequence reflects the short-term fluctuations of the phase offset around the trend sequence. The splitting method involves performing a moving average on the phase offset sequence. The window length of the moving average is determined based on the total length of the phase offset sequence, for example, one-third to one-half of the total length. The smoothed sequence obtained after the moving average is used as the trend sequence. The residual sequence is calculated as: r k =φ k -φtrend k ; where r k φ represents the k-th residual value in the residual sequence. k φtrend represents the k-th phase offset in the phase offset sequence. k This represents the k-th trend value in the trend sequence, where k is the index and r is the trend value. k φ k and φtrend k All of them are dimensionless.

[0051] This function determines whether the sign of a trend sequence remains consistently positive and its amplitude is stable or increasing over time. A consistently positive sign means all trend values ​​in the sequence are greater than zero. Stable amplitude means the absolute value of the difference between adjacent trend values ​​is less than a preset fluctuation tolerance. This tolerance is set based on the statistical characteristics of the trend sequence, using the standard deviation of the trend sequence multiplied by a fluctuation coefficient (e.g., 0.5 to 1.0). Increasing amplitude means the mean of the later trend values ​​is significantly greater than the mean of the earlier trend values. Specifically, the trend sequence is divided into two equal segments based on the execution time. The mean of the earlier and later trends is calculated separately. An increasing amplitude is defined as the difference between the later and earlier trend mean values ​​being positive and exceeding a preset significance threshold. This threshold is set by multiplying the absolute value of the earlier trend mean by a scaling factor (e.g., 0.1 to 0.3).

[0052] The preset fluctuation tolerance is used to determine whether the fluctuation of the difference between adjacent trend values ​​in a trend sequence is within the normal range of random fluctuations. The selection of the fluctuation coefficient within the range of 0.5 to 1.0 is based on the relationship between the stationarity characteristics of the trend sequence and the physical properties of the leakage process. When the contribution of true leakage to the phase shift is dominant, the trend sequence exhibits high stability in time, and the difference between adjacent trend values ​​is mainly caused by random disturbances in the measurement conditions and fitting process in each test. At this time, the standard deviation of the trend sequence is small, and the fluctuation coefficient can be taken as a larger value, such as close to 1.0, so that the preset fluctuation tolerance covers the normal range of random fluctuations. When the contribution of true leakage and the contribution of residual venting are in competition, the trend sequence may be mixed with the decaying trend residue of residual venting, and the standard deviation is relatively large. The fluctuation coefficient should be taken as a smaller value, such as close to 0.5, to improve the sensitivity to abnormal amplitude fluctuations. In practical applications, the absolute value sequence of the difference between each adjacent trend value in the trend sequence can be calculated first. If the mean of the absolute value sequence is less than 0.3 times the standard deviation of the trend sequence, it indicates that the fluctuation of the trend sequence is mainly random disturbance, and the fluctuation coefficient can be selected between 0.7 and 1.0. If the mean of the absolute value sequence is greater than 0.3 times the standard deviation of the trend sequence, the fluctuation coefficient can be selected between 0.5 and 0.7.

[0053] Simultaneously, it is determined whether the fluctuation characteristics of the residual sequence are a random distribution around zero. A random distribution around zero means that the signs of the residual values ​​in the residual sequence are evenly distributed and there are no continuous segments with the same sign. The determination method is as follows: calculate the mean of the residual sequence. The mean of the residual sequence should be close to zero. The criterion for being close to zero is that the absolute value of the residual sequence mean is less than a preset proportion of the absolute value of the phase offset sequence mean, for example, a preset proportion of 0.05 to 0.15. The consecutive lengths of adjacent residual values ​​with the same sign in the residual sequence are statistically analyzed. If the maximum length of consecutive identical signs does not exceed a preset length threshold, it is determined that the residual sequence does not have continuous segments with the same sign. The preset length threshold is determined based on the total length of the residual sequence, for example, one-quarter to one-third of the total length.

[0054] A preset length threshold is used to determine whether the maximum length of consecutive identical-sign segments in the residual sequence exceeds the reasonable range of random dispersion. The basis for setting the preset length threshold to one-quarter to one-third of the total length of the residual sequence is that, for a completely random sequence of symbols, the expected maximum length of consecutive identical-sign segments increases slowly with the total sequence length, and its theoretical expected value is approximately the logarithm of the total sequence length to base 2. When the total sequence length is small, the ratio of the theoretical expected value to the total sequence length is relatively large, and the preset length threshold should be a higher proportion, such as close to one-third, to avoid misclassifying normal random identical-sign segments as persistent structures. When the total sequence length is large, the ratio of the theoretical expected value to the total sequence length tends to decrease, and the preset length threshold should be a lower proportion, such as close to one-quarter, to maintain the sensitivity to persistent structures. In practical applications, if the total length of the residual sequence is less than 20, the preset length threshold can be one-third of its total length; if the total length of the residual sequence is greater than or equal to 20, the preset length threshold can be one-quarter of its total length. The value of the preset length threshold can be determined within the above range based on the actual length of the residual sequence.

[0055] When the trend sequence is consistently positive with a stable or increasing amplitude and the residual sequence is randomly distributed around zero, the phase offset is identified as containing a consistently positive component that does not decay over time. "Does not decay over time" means the trend sequence does not exhibit a monotonically decreasing trend over time, and "constantly positive" means the sign of the trend sequence remains consistently positive. If the trend sequence shows negative values ​​or a monotonically decreasing trend, or the mean of the residual sequence deviates significantly from zero, or the residual sequence contains continuous segments with the same sign, the phase offset is identified as not containing a consistently positive component.

[0056] This method of identifying the constant positive component by splitting the phase shift sequence into a trend sequence and a residual sequence and judging their statistical characteristics separately leverages the essential difference in temporal behavior between true leaks and residual venting. The phase shift contribution from a true leak is a continuous and stable positive shift, appearing as consistently positive and undiminished in the trend sequence, while in the residual sequence it shows a dispersion around zero due to random errors in measurement and fitting. The phase shift contribution from residual venting, on the other hand, decays as the furnace venting gradually depletes, showing a monotonically decreasing trend sequence that eventually approaches zero. By independently judging the sign and amplitude trends of the trend sequence and the dispersion characteristics of the residual sequence, the constant positive characteristic of true leaks and the decaying characteristic of residual venting can be effectively distinguished, enabling accurate identification of the leak state.

[0057] In the specific implementation of S6, if a constant positive component that does not decay over time is identified in the phase offset in S5, it is determined that there is a leak in the pipeline under test. If a constant positive component is not identified in the phase offset in S5, it is determined that there is no leak in the pipeline under test.

[0058] The basis for determining whether a pipeline under test has a leak or not lies in the presence or absence of a constant positive component, which directly reflects the continuous and stable positive contribution of a true leak to the phase shift. The intake flow rate generated by a true leak remains essentially constant over the test timescales S1 to S4, not decreasing as the furnace venting decays. The portion of the phase shift contributed by the true leak exhibits a consistently positive sign and a stable or gradually increasing amplitude in the trend sequence of the phase shift sequence. In the residual sequence, it shows a random distribution around zero due to random disturbances in the measurement conditions and fitting process during each test. This is precisely the statistical manifestation of a constant positive component that does not decay over time. When the pipeline under test only has residual venting without a true leak, the contribution of residual venting to the phase shift will decrease synchronously with the furnace venting decay until it approaches zero. In the trend sequence of the phase shift sequence, it shows a monotonically decreasing trend that eventually approaches zero, failing to meet the identification condition that the constant positive component is continuously positive in the trend sequence and its amplitude does not decay. By directly mapping the identification results of the constant positive component of S5 to the presence or absence of leakage in the pipeline under test, a clear determination of the leakage status of the pipeline under test is achieved, providing a direct basis for subsequent instrument leak detection or targeted repair of pipelines under test that are determined to have leaks.

[0059] All calculations involved in the embodiments are dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to the actual situation.

[0060] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0061] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and inventive constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

Claims

1. A method for dynamic monitoring of pressure rise rate and zoned location of leaks in a vacuum furnace, characterized in that, Includes the following steps: S1: After evacuating the furnace, close the main valve and collect the dynamic trend of the furnace pressure rise rate; S2: Close the connecting valve between the pipeline under test and the furnace, and evacuate the pipeline under test separately until the pressure decay curve of the pipeline under test passes the morphological inflection point and tends to stabilize. S3: Open the connecting valve and evacuate the integrated space formed by the furnace and the pipeline under test. After closing the main valve, identify the turning point when the pressure recovery rate in the integrated space changes from a decaying trend to a constant trend. S4: Extract the furnace venting decay time constant from the dynamic trend of furnace pressure rise rate, map the turning point to the normalized time axis based on the furnace venting decay time constant, and calculate the phase offset of the turning point relative to the expected equilibrium position on the normalized time axis. S5: Analyze the behavior of phase offset over time and identify whether the phase offset contains a constant positive component that does not decay over time. S6: Determine the leakage status of the pipeline under test based on whether a constant positive component is identified.

2. The method for dynamic monitoring of pressure rise rate and leakage point location in a vacuum furnace according to claim 1, characterized in that, After evacuating the furnace and closing the main valve, the dynamic trend of the furnace pressure rise rate is collected, including: recording the furnace pressure value multiple times during the observation period after closing the main valve, dividing the difference between each recorded pressure value and the adjacent previous recorded pressure value by the recording interval to obtain the pressure rise rate sequence, identifying the segment boundary points where the change characteristics in the pressure rise rate sequence change, dividing the observation period into the first segment and the second segment based on the segment boundary points, and taking the pressure rise rate sequence in the first segment and the pressure rise rate sequence in the second segment together to constitute the dynamic trend of the furnace pressure rise rate.

3. The method for dynamic monitoring of pressure rise rate and leakage point location in a vacuum furnace according to claim 2, characterized in that, Identifying segmentation boundaries includes: successively subtracting adjacent pressure rise rate values ​​in the pressure rise rate sequence to obtain a first-order difference sequence; successively subtracting adjacent first-order difference values ​​in the first-order difference sequence to obtain a second-order difference sequence; and taking the recording time corresponding to the first second-order difference value in the second-order difference sequence that changes from a biased zero-value fluctuation to a positive fluctuation range exceeding zero-value fluctuation as the segmentation boundary.

4. The method for dynamic monitoring of pressure rise rate and leakage point location in a vacuum furnace according to claim 1, characterized in that, Close the connecting valve between the pipeline under test and the furnace, and evacuate the pipeline under test separately until the pressure decay curve of the pipeline under test passes the morphological inflection point and tends to stabilize. This includes: continuously recording the pressure value inside the pipeline under test during the process of evacuating the pipeline under test separately, calculating the pressure decay rate between adjacent recording times, identifying the transition zone where the pressure decay rate changes from rapid decay to gradual decay, determining the moment when the rate of change of pressure decay rate in the transition zone first falls below the preset rate of change threshold as the morphological inflection point, and determining that the pressure decay rate fluctuation amplitude of the continuously recorded pressure decay rate after the morphological inflection point is maintained within the specified range.

5. The method for dynamic monitoring of pressure rise rate and leak point location in a vacuum furnace according to claim 4, characterized in that, The moment when the rate of change of pressure decay rate in the transition zone first falls below the preset rate of change threshold is the morphological inflection point. This includes: taking the average rate of change of pressure decay rate between the moment when the pipeline under test starts to be evacuated and the moment when the rate of change of pressure decay rate enters the rapid decay stage as the reference rate of change; multiplying the reference rate of change by the proportional coefficient as the preset rate of change threshold; and taking the moment when the rate of change of pressure decay rate in the transition zone first falls below the preset rate of change threshold as the morphological inflection point.

6. The method for dynamic monitoring of pressure rise rate and leak location in a vacuum furnace according to claim 1, characterized in that, Open the connecting valve and evacuate the integrated space formed by the furnace and the pipeline under test. After closing the main valve, identify the turning point when the pressure recovery rate in the integrated space changes from a decreasing trend to a constant trend. This includes: continuously recording the pressure value in the integrated space during the observation period after closing the main valve, calculating the pressure recovery rate between adjacent recording times to form a pressure recovery rate sequence, identifying the transition moment in the pressure recovery rate sequence from continuous decrease to fluctuation within a fixed amplitude range, and taking the transition moment as the turning point.

7. The method for dynamic monitoring of pressure rise rate and leakage point location in a vacuum furnace according to claim 6, characterized in that, Identifying the transition moment in the pressure recovery rate sequence from a continuous decrease to fluctuation within a fixed amplitude range includes: calculating the moving standard deviation of several consecutive pressure recovery rate values ​​in the pressure recovery rate sequence; when the starting moment of the time window corresponding to the first moving standard deviation in the moving standard deviation sequence that is lower than the standard deviation threshold, and the pressure recovery rate sequence after this time window no longer shows a continuous decreasing trend, the starting moment of this time window is taken as the transition moment.

8. The method for dynamic monitoring of pressure rise rate and leakage point location in a vacuum furnace according to claim 1, characterized in that, The furnace venting attenuation time constant is extracted from the dynamic trend of the furnace pressure rise rate. The inflection point is mapped to a normalized time axis based on the furnace venting attenuation time constant. The phase offset of the inflection point relative to the expected equilibrium position on the normalized time axis is calculated. This includes: performing attenuation fitting on the pressure rise rate sequence in the second segment of the furnace pressure rise rate dynamic trend to obtain the second attenuation curve and extracting the furnace venting attenuation time constant; constructing a normalized time axis with the furnace venting attenuation time constant as the unit scale; performing attenuation fitting on the pressure rise rate sequence in the first segment of the furnace pressure rise rate dynamic trend to obtain the first attenuation curve; taking the intersection of the asymptotes of the first attenuation curve and the second attenuation curve on the normalized time axis as the expected equilibrium position; mapping the inflection point to the normalized time axis to obtain the normalized time position; and taking the difference between the normalized time position and the expected equilibrium position as the phase offset.

9. The method for dynamic monitoring of pressure rise rate and leakage point location in a vacuum furnace according to claim 1, characterized in that, Analyzing the behavior of phase offset over time and identifying whether the phase offset contains a constant positive component that does not decay over time includes: arranging multiple phase offsets obtained from multiple executions of S1 to S4 according to the execution sequence to form a phase offset sequence; splitting the phase offset sequence into a trend sequence and a residual sequence; determining whether the sign of the trend sequence is consistently positive and the amplitude is stable or increasing; and simultaneously determining whether the fluctuation characteristics of the residual sequence are random dispersion around zero. When the trend sequence is consistently positive and the amplitude is stable or increasing, and the residual sequence is random dispersion around zero, it is identified that the phase offset contains a constant positive component that does not decay over time.

10. The method for dynamic monitoring of pressure rise rate and leakage point location in a vacuum furnace according to claim 1, characterized in that, The leakage status of the pipeline under test is determined based on whether a constant positive component is identified, including: when a constant positive component is identified, the pipeline under test is determined to have a leak; when a constant positive component is not identified, the pipeline under test is determined to have no leak.

Images