Method and device for detecting critical humidity in housing of active module
By monitoring temperature and humidity changes during the cold start phase of the active module, predicting the dew point temperature using a dew point model, and detecting humidity changes in real time using the dew point model, the technical challenges that the patent application sought to address, which were difficult to solve in the prior art, were resolved. This enabled humidity monitoring within the enclosed housing of the active module, preventing condensation and potential damage to the components.
Patent Information
- Application Number
- CN202480037464.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2023-06-07
- Filing Date
- 2024-06-04
- Publication Date
- 2026-01-02
AI Technical Summary
Existing technologies struggle to detect excessive humidity levels in the enclosed housing of active modules in a timely manner, especially before it leads to condensation or functional failure of the components.
By monitoring the temperature and humidity changes inside the casing during the cold start phase, the maximum dew point temperature is predicted using a data-based dew point model and compared with a set threshold, thus detecting humidity exceeding the limit in real time.
This technology enables timely detection of excessive humidity before the active module is cold-started, preventing condensation and potential damage to components and improving safety and reliability.
Smart Images
Figure CN121263671A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The invention relates to a method for identifying a critical humidity within a closed housing of an active module. The invention further relates to a method for avoiding dew formation within components of an active module. BACKGROUND
[0002] Active modules, such as electronic modules (e.g. control devices), which are operated outside a protected environment, are usually encapsulated in a closed housing. The closed housing protects the components, in particular the electronic components, of the active module from dirt and humidity. The housing is usually moisture-tight closed, so that additional humidity cannot penetrate the housing under normal circumstances.
[0003] However, due to damage, cracks in the housing material due to aging, etc., it can happen that water or humidity enters the interior of the housing. This increase in the amount of humidity within the housing is not immediately and easily detected, in particular when the additional humidity that penetrates is only gradually increasing and does not immediately lead to water condensation on individual components, thus not directly causing a malfunction. However, for safety reasons, it is required to determine the unsealing of the housing and the penetration of humidity beyond a critical value as quickly as possible, in particular before a functional failure occurs. This is particularly necessary for such active modules (e.g. in the form of control devices) used in steering and braking systems in vehicles. SUMMARY
[0004] It is the task of the invention to provide a method for monitoring the humidity within a closed housing of an active module, by which a critical value of humidity can be detected before actual dew formation on individual components or water condensation within the housing occurs.
[0005] This task is solved by a method for monitoring the humidity within a housing of an active module according to claim 1 and an active module within a closed housing according to another independent claim.
[0006] Further design solutions are given in the dependent claims.
[0007] According to a first aspect, a method for monitoring the humidity inside a closed housing of an active module is provided, having the following steps: - detecting a time profile of the temperature and humidity of the gas inside the housing after a cold start of the active module; - providing an ambient temperature at the moment of the cold start of the active module; - indicating a too high humidity inside the housing by means of a provided data-based dew point model from the detected temperature and humidity profile, wherein the data-based dew point model is trained to output a maximum dew point temperature from the temperature and humidity profile or further to output an indication that the maximum dew point temperature can be undershot depending on the ambient temperature at the beginning of the cold start phase.
[0008] Furthermore, signaling an excessive humidity inside the housing can also comprise the following steps: - determining a modeled maximum dew point temperature during the cold start phase based on a temperature and humidity profile by means of a trained data-based dew point model; - signaling an excessive humidity inside the housing when the ambient temperature is below a threshold determined depending on the maximum dew point temperature.
[0009] Upon cold start of an encapsulated active module, e.g. a control device, usually a local warming of individual components inside the housing will occur first. The local warming leads to a release of stored humidity at the component material, thereby increasing the absolute humidity inside the housing. However, the inner walls of the housing and e.g. the unpowered components are initially at ambient temperature, so that depending on the degree of humidity increase and the temperature of the not yet warmed components, dew or water condensation can occur on the surface of the not yet warmed components. This can lead to impaired and damaged functionality of the active module upon contact to e.g. current-carrying elements.
[0010] The ambient temperature can be provided as the temperature at the moment of cold start of the active module by a temperature sensor inside or outside the active module, or as the first detected temperature value inside the housing.
[0011] For monitoring the humidity inside the housing, a temperature / humidity sensor can be provided, in particular as a combined sensor, which can be integrated into the electronics of the active device. The temperature / humidity sensor makes it possible to detect the temperature and humidity of the air inside the housing and to determine therefrom the dew point temperature by means of a known dew point table, which relates the humidity and temperature of the air to the dew point temperature, or to determine the absolute humidity in the form of the amount of water vapor present inside the housing. If humidity or water enters the housing, the water vapor content inside the housing rises, so that water condensation on the not yet warmed components of the active device can result.
[0012] The dew point temperature corresponds to the temperature at which water condensation occurs in the air, or at which water condensation occurs on a surface whose temperature is equal to the dew point. The dew point temperature can be determined by means of a dew point table or according to known mathematical models.
[0013] Determining the critical state inside the housing is not trivial, since the plastics and other e.g. hygroscopic materials of the components of the active module can release stored or bound water into the air inside the housing by warming.
[0014] From the course of the humidity and the temperature of the gas (e.g. air) enclosed in the housing after a cold start (i.e. all components and the housing are essentially at ambient temperature), the dew point temperature can now be determined at which condensation of water can occur on components of the active module which have not yet been warmed up, or can occur.
[0015] In order to monitor the critical humidity level inside the housing, the development or time course of the humidity (i.e. of the enclosed gas) and the temperature (i.e. of the enclosed gas) inside the housing during the cold start phase can now be monitored by means of a data-based dew point model and the dew point temperature can be determined. By means of a threshold comparison of the ambient temperature, which is assumed to be the lowest possible temperature of the components in the active module, with a temperature threshold defined depending on the dew point temperature, an excessive ingress of moisture into the interior of the housing, for example due to a lack of tightness, can be determined.
[0016] It can be provided that the threshold is higher than the maximum dew point temperature by a preset temperature offset.
[0017] The threshold can be derived from the dew point temperature plus an offset (tolerance).
[0018] Thus, for example, it can be assumed that the last warmed-up components have a temperature corresponding to the ambient temperature. If the dew point temperature is estimated from the initial course of the humidity and the temperature inside the housing, which dew point temperature differs from the temperature of the components involved by less than a threshold amount, a corresponding fault signal or alarm can be signaled. By means of the data-based dew point model, the storage capacity of plastics, hygroscopic materials, etc. inside the housing can be taken into account.
[0019] It can be provided that the data-based dew point model is configured as a regression model, which is configured in the form of a deep neural network or a recurrent neural network.
[0020] According to an embodiment, the data-based dew point model can be trained with a training data set in order to associate one or more course features of the time course of the temperature and the humidity during an initial phase with the maximum dew point temperature to be expected at a time during the entire cold start phase, wherein the initial phase indicates a time period at the beginning of the cold start phase. The initial phase can be, for example, a time period of 1 to 5 minutes.
[0021] Alternatively, the data-based dew point model can be trained as a recurrent model with a training data set in order to determine a predicted course of the temperature and the humidity on the basis of the time course of the temperature and the humidity during the initial phase and to determine the maximum dew point temperature from the predicted course of the temperature and the humidity.
[0022] The data-based dew point model can be trained from measurement sequences in which the time course of humidity and temperature is recorded for a specific active module starting from the moment when all components inside the housing appear to be at ambient temperature (cold start). The training data set is generated here as follows: the time sequence of temperature and humidity during a predetermined initial phase after the cold start is associated with the maximum value of the time course of the dew point temperature for a subsequent time period of, for example, at most 30 minutes (preferably at most 15 minutes) after the start of the cold start phase, which corresponds to the remainder of the warm-up phase.
[0023] In addition, the measurement sequences for determining the training data set can also take into account entries for different amounts of water entering the housing interior. Thus, when observing the initial time course of temperature and humidity inside the housing after the cold start, it is possible to deduce the maximum dew point temperature that will be reached during the warm-up phase of the active module.
[0024] By comparing the threshold value derived from the maximum dew point temperature, taking into account a safety tolerance, with the ambient temperature, it is possible to determine whether, in the warm-up phase of, for example, 1 minute to 30 minutes (in particular between 5 minutes and 15 minutes), there is a risk of dew formation (i.e. condensation of water) on the component that heats up most slowly inside the housing.
[0025] The data-based dew point model can output the maximum dew point temperature or, as a classification model, an indication that there is a risk of dew formation on a component due to a temperature below the dew point temperature. BRIEF DESCRIPTION OF DRAWINGS
[0026] Embodiments will be described in more detail below with reference to the drawings.
[0027] Figure 1 A schematic diagram of an active module inside an enclosed housing is shown; Figure 2 A flow chart for describing a method for operating an active module to monitor critical humidity quantities inside a housing is shown; and Figure 3 A flow chart for illustrating a method for training a data-based dew point model is shown. DETAILED DESCRIPTION
[0028] Figure 1 A schematic cross-sectional view of an active module 1 with electronic components is shown. The active module 1 can be a control device which has a control device circuit board 2 on which an electronic component group is implemented. The control device circuit board 2 comprises an electronic circuit with components 3 which are energized when activated. Different powers are converted in different components of the circuit board 2 when activated, which can lead to different heating.
[0029] The circuit board 2 is accommodated in a housing 4, which is moisture-tight closed. Thereby, the circuit board 2 and the components 3 of the electronic circuit located on the circuit board 2 are protected from moisture and water.
[0030] The circuit board 2 or the electronic components also comprise a temperature / humidity sensor 5 with which the temperature of the air or gas in the interior of the housing 4 and its humidity can be measured. The temperature / humidity sensor 5 can be configured as a combined sensor.
[0031] By means of an evaluation unit 6, which performs data processing implemented in software and / or hardware and which can likewise be configured on the circuit board 2 or separately therefrom, the time curves of temperature and humidity can be detected and evaluated, and a warning signal can be generated upon recognition of a critical state in the interior of the housing 4.
[0032] The housing also has an interface 7 via which the active module 1 can be connected to a power supply and via which it can communicate with external devices by means of electrical signals.
[0033] The warning signal can be communicated to the external devices via the interface, for example.
[0034] The evaluation unit 6 is used to evaluate the time curves of temperature and humidity after activation of the active module 1 during a cold start phase. For the cold start phase, it is initially assumed that all components of the module 1 have a temperature corresponding to the ambient temperature. As soon as the module 1 is activated, the evaluation unit 6 starts recording the time curves of temperature and humidity in the interior of the housing 4. These curves can be buffered or directly evaluated in order to determine a model output of a dew point model implemented in the evaluation unit 6, for example the maximum dew point temperature during the cold start phase, when the dew point model is implemented as a data-based regression model. Alternatively, in the case of a data-based dew point model implemented as a classification model, an indication about a critical state in the interior of the active module 1 can be determined.
[0035] The evaluation can be carried out by means of a data-based dew point model, for example. The data-based dew point model can extract curve features from the time curves of temperature and humidity and map them to a maximum dew point by means of a data-based dew point model configured as a neural network, for example.
[0036] Alternatively, the data-based dew point model can also be configured as an LSTM (Long Short term memory) or other data-based recurrent model, which enables the prediction of a curve of the dew point in the interior of the housing during a warm-up phase and stores the maximum value of the predicted dew point temperature curve.
[0037] The evaluation unit 6 can also provide that it is determined by means of a threshold comparison whether a component, in particular the component that warms up most slowly, is in danger of dew formation.
[0038] Figure 2 The method performed in the evaluation unit within the module is illustrated schematically according to a flowchart.
[0039] In step S1, it is first checked whether the module has been activated for a cold start. If the module has not been activated during a predetermined period of time, and it can be assumed that the components in the module 1 and their housing 4 are at ambient temperature, a cold start is present.
[0040] If it is determined in step S1 that the module has been activated (alternative: yes), the recording of the temperature and humidity values of the gas within the housing 4 is initiated in step S2, and the corresponding time series is buffered. In addition, the ambient temperature can be detected separately, or the first detected temperature value is assumed to be the ambient temperature. The temperature and humidity values are detected by means of temperature and humidity sensors 5 on the circuit board 2 in the module 1. Otherwise (alternative: no), jump back to step S1.
[0041] In step S3, it is checked whether a predetermined time corresponding to the initial phase has elapsed since the start of the recording. If this is the case (alternative: yes), the time series is evaluated in step S4 to determine the maximum dew point temperature. Otherwise, jump back to step S2.
[0042] The evaluation of step S4 is carried out by means of a data-based dew point model, which enables the maximum dew point temperature to be determined directly by regression, or by means of a prediction of the time curve of the dew point temperature from the end of the initial phase.
[0043] The data-based dew point model can be constructed as a deep neural network or a recurrent neural network (e.g. LSTM, etc.). In the case of a deep neural network, one or more curve features are extracted from an input vector composed of input quantities from the time series of temperature and humidity during the initial phase, such as the maximum, minimum and / or average gradient of temperature and humidity during the initial phase, the maximum change in temperature and humidity during the initial phase, the dew point temperature determined from temperature and humidity at the beginning of the cold start phase according to a preset mathematical model or a preset characteristic map, the values of temperature and humidity at preset times during the initial phase, the maximum and minimum curvatures of the temperature and humidity curve during the initial phase.
[0044] In the case of an LSTM, the time series of temperature and humidity during the initial phase are preset and the time curve of the temperature and humidity of the gas inside the housing after the initial phase is predicted. The dew point temperature is derived from the respective temperature and humidity values, so that a time series of the predicted dew point temperature until the end of the cold start phase is obtained. The maximum dew point temperature can be determined from the predicted dew point temperature curve.
[0045] Now, in step S5, it is checked in dependence on the ambient temperature whether the ambient temperature is below a threshold amount depending on the maximum dew point temperature, which can be derived, for example, from the maximum dew point temperature plus a tolerance amount of, for example, between 1 °C and 3 °C. It is assumed here that the component that warms up slowest during the entire cold start phase, which can correspond to a duration of between 5 minutes and 30 minutes, is at the assumed temperature corresponding to the ambient temperature. If this is the case (alternative: yes), this is correspondingly signaled in step S6, for example as a warning output to the outside of the module in the form of an error message. Otherwise, the method is ended until a reactivation in the cold start phase is determined.
[0046] In Figure 3 A flowchart for illustrating a training method for a data-based dew point model is described in
[0047] In step Sll, measurements are made on the active module at different ambient temperatures and different humidity contents inside the housing 4 during a defined duration of the cold start phase of, for example, 30 minutes. The measurement is made by recording the time series of temperature and humidity during the cold start phase, each time starting from a module that is completely cooled to the ambient temperature. The time series is based on a sampling rate of, for example, between 1 second and 15 seconds.
[0048] In step S12, a training data set is determined from the time curves of temperature and humidity during the cold start phase of the active module 1 to be modeled. For this purpose, input quantities that are characteristic of the time curve of temperature and humidity during the initial phase as input quantities (as illustrated in the example above) can be determined from the time curve as input vectors of the training data set. The characteristic of the time curve can thus be produced by extracting features from the time curve of temperature and humidity of the initial phase of the cold start phase, for example, the phase within the first three minutes.
[0049] The combination of temperature and humidity at each detection instant of the time series can be associated with a dew point temperature, respectively. The maximum dew point temperature can be determined therefrom as a label of the training data set concerned.
[0050] Alternatively, the data-based dew point model can also be constructed as a recurrent neural network, for example an LSTM, which can be used to predict time series during the cold start phase, each time starting from the initial time series or starting value of the open state of the active model. By correspondingly evaluating the predicted time series, the maximum dew point temperature can be found.
[0051] The training of the recurrent neural network can be carried out from the time series of various measured temperatures and humidities as training data sets.
[0052] After the training, the data-based dew point model can be implemented in step S13 by transferring the parameters into the evaluation unit 6 of the module 1.
Claims
1. A method for monitoring humidity inside a closed housing (4) of an active module (1), comprising the following steps: - After the active module (1) is cold-started, the temperature and humidity change curves of the gas inside the housing (4) are detected (S2) during the preset cold-start phase; - Provide the ambient temperature at the cold start moment of the active module (1) described in (S2); - Based on the detected temperature and humidity change curves and the ambient temperature, using the provided data-based dew point model, a signal is used to indicate (S6) that the humidity inside the housing (4) is too high, wherein the data-based dew point model is trained to output the maximum dew point temperature based on the temperature and humidity change curves, or in particular to output an indication of a possible lower than the maximum dew point temperature based on the ambient temperature at the start of the cold start phase.
2. The method according to claim 1, wherein indicating excessive humidity inside the housing (4) by signaling comprises the following steps: - Using the trained data-based dew point model, the maximum dew point temperature modeled during the cold start phase (S4) is determined based on the temperature and humidity variation curves. - When the ambient temperature is below a threshold determined based on the maximum dew point temperature, a signal (S6) is used to indicate (S6) that the humidity inside the housing (4) is too high.
3. The method according to claim 1 or 2, wherein the temperature and humidity of the gas are determined by means of a combination of humidity and temperature sensors.
4. The method according to any one of claims 1 to 3, wherein the ambient temperature, as the temperature at the cold start moment of the active module (1), is provided by a temperature sensor inside or outside the active module (1), or as the temperature value inside the housing (4) first detected.
5. The method according to any one of claims 1 to 4, wherein the data-based dew point model is configured as a regression model, and the regression model is configured in the form of a deep neural network or a recurrent neural network.
6. The method according to any one of claims 1 to 4, wherein the data-based dew point model is trained with a training dataset to correlate one or more curve features of the time-varying temperature and humidity during the initial phase with the maximum dew point temperature during the entire cold start phase.
7. The method according to any one of claims 1 to 4, wherein the data-based dew point model is trained as a recursive model with a training dataset to determine predicted temperature and humidity change curves based on the time-varying temperature and humidity curves during the initial phase, and to determine the maximum dew point temperature from the predicted temperature and humidity change curves.
8. The method according to any one of claims 1 to 6, wherein the threshold is higher than the maximum dew point temperature by a preset temperature offset.
9. A method for training a data-based dew point model based on a training dataset that correlates temperature and humidity variation curves inside the housing (4) of an active module (1) with an indication of maximum dew point temperature, or in particular, correlates ambient temperature at the start of a cold start phase with an indication of a temperature that may be below the maximum dew point temperature.
10. A device for monitoring humidity inside a closed housing (4) of an active module, particularly an evaluation unit (6), wherein the device is configured as follows: - After the active module (1) is cold-started, the temperature and humidity change curves of the gas inside the housing (4) are detected during the preset cold-start phase. - Provide the ambient temperature at the cold start moment of the active module (1); - Based on the detected temperature and humidity change curves and the ambient temperature, a signal is used to indicate that the humidity inside the housing (4) is too high by means of a provided data-based dew point model, wherein the data-based dew point model is trained to output the maximum dew point temperature based on the temperature and humidity change curves, or in particular to output an indication of a possible lower than the maximum dew point temperature based on the ambient temperature at the start of the cold start phase.
11. A computer program comprising instructions that, when executed in a data processing apparatus or a computer, cause the data processing apparatus or the computer to perform the method according to any one of claims 1 to 8.
12. A machine-readable storage medium having a computer program as claimed in claim 11 stored thereon.