Monitoring device for a solenoid valve of a compressed-air system of a vehicle, system composed of the solenoid valve, the monitoring device and a control device, and method for monitoring the solenoid valve

Abstract

The invention relates to a monitoring device (3) for a solenoid valve (2) of a compressed-air system of a vehicle, and to a method for monitoring the valve. The monitoring device (3) comprises: A current-measuring apparatus (5) which is designed to measure an electric supply current (I) of the solenoid valve (2), and an evaluation apparatus (6) which is designed to evaluate the change in the measured supply current (1) over time during a switching operation of the solenoid valve (2) and to provide state information relating to the state of the solenoid valve (2), which state information is dependent on this change over time.

Description

[0001] DESCRIPTION

[0002] Monitoring device for a solenoid valve of a compressed air system of a vehicle, system comprising the solenoid valve, the monitoring device and a control device, and method for monitoring the solenoid valve

[0003] The invention relates to a monitoring device for a solenoid valve of a compressed air system of a vehicle and a method for monitoring the solenoid valve, in particular for safety-relevant applications.

[0004] For system and safety reasons, it is necessary to perform functional monitoring of a solenoid valve. This functional monitoring monitors whether the solenoid valve is functioning properly, namely, whether the solenoid valve is switching properly.

[0005] With such monitoring, when the solenoid valve is switched, a pressure sensor is used to check whether a desired or expected pressure is reached upstream and / or downstream of the solenoid valve. For this purpose, the pressure is measured, and an evaluation device checks whether the desired pressure has been reached. However, this has previously required a pressure switch or pressure sensor for each individual solenoid valve to determine the pressure, as well as additional electronics for an evaluation device, which leads to high costs for such a monitoring device.

[0006] Another problem is that if there is no pressure on the solenoid valve, for example during a test before the vehicle starts driving, the effectiveness cannot be monitored by measuring the pressure.

[0007] To circumvent this problem, it is possible to use solenoid valves that have an additional sensor that monitors the position of the solenoid valve's armature and reports the position back as a status. This makes it possible to perform the monitoring function—that is, checking the switching of the solenoid valve—via the armature position. However, such solenoid valves are costly because they contain additional electronics. Furthermore, the signal output by the additional electronics must be read and evaluated on the control side.

[0008] This results in the object underlying the invention to eliminate the above disadvantages and to provide a monitoring device and a corresponding system and method for this purpose for a solenoid valve, so that the effective monitoring is possible without additional sensory expenditure.

[0009] This object is achieved by a monitoring device according to claim 1, a system comprising a solenoid valve, the monitoring device, and a control device according to claim 7, and a method according to claim 8. Advantageous developments of the invention are contained in the dependent claims.

[0010] According to one aspect of the invention, a monitoring device for a solenoid valve of a compressed air system of a vehicle has a

[0011] A current detection device which is designed to detect an electrical supply current of the solenoid valve, and an evaluation device which is designed to evaluate a temporal profile of the detected supply current during a switching operation of the solenoid valve and to provide status information about the status of the solenoid valve which is dependent on the temporal profile.

[0012] By evaluating the temporal progression of the supply current, the evaluation device can determine whether or not its temporal progression during the switching process of the solenoid valve is within the scope of correct operation, i.e. a typical temporal progression for the respective properties of the solenoid valve. If an atypical progression is detected, a faulty state of the solenoid valve is assumed. This makes it possible to determine whether the valve is functioning correctly or whether a potential fault exists, without the need for additional sensors. Since it is common practice in modern systems to monitor the supply current of solenoid valves for other functions as well, for example via a shunt resistor and corresponding detection electronics, a current detection device is already present and therefore does not require any additional costs.The solenoid valves are also usually controlled via an electronic control system, wherein the evaluation device according to the invention can be integrated into an already existing or already necessary electronic control system without great additional effort in series production, so that the evaluation according to the invention can also be carried out with no or very little additional effort.

[0013] The temporal profile of the supply current is defined as a temporal profile, i.e., a possibly multiple rise and fall within a predetermined period of time, during the operation of the solenoid valve and is transmitted to the evaluation device. However, a single sudden rise or fall in the supply current caused by switching on or off is not considered a temporal profile, since no status information related to the profile can be provided in this case.

[0014] According to an advantageous embodiment of the monitoring device, the evaluation device is designed to evaluate the supply current during an entire switching operation of the solenoid valve.

[0015] An entire switching process of the solenoid valve is preferably understood to mean that the supply current is evaluated from the time at which the solenoid valve is controlled by a supply voltage and until the switching process of the solenoid valve is completed, i.e. until an expected current flow for the solenoid valve is reached.

[0016] By evaluating the entire switching process, the behavior of the solenoid valve can be monitored throughout the entire switching process, allowing, for example, anomalies that occur only in certain phases of the switching process to be detected. Anomalies are preferably detected in the evaluation device by comparing the recorded temporal profile of the supply current with a typical, correct temporal profile and determining whether a deviation lies outside a predetermined tolerance range.

[0017] In a further advantageous embodiment of the monitoring device, the evaluation device is designed to determine a fault or no fault as status information within the scope of the evaluation of the entire switching process of the solenoid valve and to provide this status information as a result of an effective monitoring.

[0018] Effective monitoring refers to the monitoring that verifies the proper function of the solenoid valve. A positive result, i.e., "fault-free operation," is provided as status information related to the detected supply current profile, preferably when the supply current profile is considered typical and correct during the switching process. Alternatively or additionally, a "fault" of the solenoid valve is provided as status information, preferably when anomalies occur in the supply current profile during the switching process. Thus, effective monitoring can be carried out simply and without additional sensors.

[0019] According to a further advantageous embodiment of the monitoring device, the evaluation unit is designed to determine a "maintenance request" and to output or provide it as status information.

[0020] In this condition, the solenoid valve may continue to operate for a while, but failure is likely in the near future.

[0021] Based on the temporal profile of the detected supply current, either during the entire switching process or just during part of the switching process, wear or a malfunction can be inferred if predetermined characteristics of the profile occur, such as a higher current at a specific time due to increased friction. This allows the status information of the maintenance request to be provided in a simple manner, so that this status information can be used to consider the corresponding solenoid valve, for example, during subsequent maintenance, eliminating the risk that the corresponding solenoid valve will not be considered during subsequent maintenance and will fail shortly thereafter.

[0022] In a further advantageous embodiment of the monitoring device, the evaluation unit is designed to use the time profile of the supply current to determine an input-side pressure at the solenoid valve and to provide this pressure as the status information.

[0023] Since an actuation effort for the solenoid valve by the magnet depends on the input-side pressure at the solenoid valve, the input-side pressure can be estimated by detecting the supply current, which depends on the actuation effort, without the need for additional sensors for the input-side pressure.

[0024] According to a further advantageous embodiment of the monitoring device, the evaluation unit has a neural network which is designed to determine, by independent learning, typical current profiles for a correct state of the solenoid valve as well as atypical current profiles for malfunctions of the solenoid valve and to provide them as the state information.

[0025] When evaluating the measured supply current via the neural network, current profiles during fault-free operation of the solenoid valve can be easily learned, compared to other methods that, for example, estimate the states of the solenoid valve based on a model. This can also be done outside of the neural network, so that corresponding data sets are made available for effective monitoring within the neural network. For example, effective monitoring can be performed in such a way that a malfunction is determined by detecting anomalies in the current profile, and other properties or, if applicable, potential problems can be easily identified.According to a further aspect of the invention, a system comprising a solenoid valve, a monitoring device according to the invention as described above, and a control device configured to control the solenoid valve is provided, wherein the system is configured to control a supply voltage for the solenoid valve in dependence on the status information provided by the monitoring device.

[0026] Such a system makes it possible, for example, to control the solenoid valve with the minimum required energy input, thus ensuring reliable switching of the solenoid valve with minimal energy consumption. This makes it possible, on the one hand, to save energy and, on the other hand, to operate the solenoid valve as gently as possible to increase its service life. This is explained in more detail below.

[0027] According to a further aspect of the invention, a method for monitoring a solenoid valve of a compressed air system of a vehicle includes the steps of: detecting a supply current of the solenoid valve by means of a current detection device; evaluating a temporal profile of the detected supply current during the duration of a switching operation of the solenoid valve by means of an evaluation device in the form of status information of the solenoid valve related to the temporal profile; and providing the status information.

[0028] By analyzing the supply current's temporal profile, a typical temporal profile for fault-free operation can be determined without the need for additional sensors. This allows the solenoid valve to be monitored with little or no additional effort and / or provides status information about potential problems or malfunctions of the solenoid valve.

[0029] In an advantageous embodiment of the method, the supply current profile is evaluated using a neural network based on results trained using machine learning. An evaluation algorithm can be generated based on input data from the measured supply current and results determined during a learning process for a particular piece of state information, without the need for complex definition of relationships between the input data.

[0030] According to an advantageous embodiment of the method, the supply current is evaluated during an entire switching operation of the solenoid valve.

[0031] By evaluating the entire switching process, the behavior of the solenoid valve can be monitored throughout the entire switching process so that, for example, anomalies that only occur in certain phases of the switching process can be detected.

[0032] The detection of an anomaly preferably occurs in the evaluation device by preferably comparing the detected temporal profile of the supply current with a typical, correct temporal profile and a deviation lies outside a predetermined tolerance range.

[0033] According to a further advantageous embodiment of the method, an expected service life of the solenoid valve is provided as the status information.

[0034] From the time course of the measured supply current, the expected service life of the solenoid valve can be easily provided as the status information.

[0035] In a further advantageous embodiment of the method, a failure probability of the solenoid valve is provided as the state information.

[0036] The probability of failure of the solenoid valve can be easily provided as status information from the temporal profile of the detected supply current. In an advantageous embodiment of the method, a result of an effective monitoring of a switching function of the solenoid valve is provided as the status information.

[0037] It is easy to determine the functionality of the solenoid valve without additional sensors.

[0038] In a further advantageous embodiment of the method, a maintenance request for required maintenance is provided as the status information.

[0039] The status of the solenoid valve can be easily determined from the temporal course of the detected supply current and thus the maintenance request can be provided without additional sensors.

[0040] According to another advantageous embodiment of the method, a pressure applied to the input side of the solenoid valve is provided as the status information.

[0041] Since a power for actuating the solenoid valve depends on the input-side pressure at the solenoid valve, the input-side pressure can be estimated by detecting the supply current without the need for additional sensors for detecting the input-side pressure.

[0042] The invention is explained below using exemplary embodiments with reference to the attached figures.

[0043] In particular,

[0044] Fig. 1 is a circuit diagram of a system according to the invention;

[0045] Fig. 2 shows a diagram with current curves during a switching operation; and Fig. 3 shows a flowchart of a method for monitoring a solenoid valve of a compressed air system of a vehicle.

[0046] Fig. 1 shows a circuit diagram of a system 1 according to the invention. The system 1 has a solenoid valve 2 of a compressed air system of a vehicle, a monitoring device 3 for the solenoid valve 2 and a control device 4 which is designed to control the solenoid valve 2.

[0047] The monitoring device 3 comprises a current detection device 5 and an evaluation device 6. In alternative embodiments, the monitoring device 3 is integrated into at least one of the control device 4, the current detection device 5, and the evaluation device 6.

[0048] The solenoid valve 2 comprises a switching valve that allows the flow of a fluid through the switching valve to be blocked or closed via a valve body that rests on a valve seat. The valve body is connected to an armature that is moved by a magnet to rest on the valve seat and block or open the flow.

[0049] The current detection device 5 detects an electrical supply current I of the solenoid valve 2. For this purpose, the current detection device 5 comprises a shunt resistor and a cooperating detection electronics. In alternative embodiments, the supply current I is detected via a current transformer.

[0050] The evaluation device 6 evaluates a temporal profile of the detected supply current I and provides status information related to this temporal profile. The temporal profile of the supply current I is defined as a temporal profile, i.e., a possibly multiple rise and fall. However, a mere sudden rise or fall of the supply current I due to switching on or off is not considered a temporal profile, since no status information related to the profile can be provided in this case. In particular, the evaluation device 6 evaluates the temporal profile of the supply current I during an entire switching process. This means that the supply current I is evaluated from the time at which the solenoid valve 2 is controlled with a supply voltage until the switching process of the solenoid valve 2 is completed, i.e., until an expected current for the solenoid valve is reached.In alternative embodiments, the temporal profile of the supply current is not evaluated during the entire switching process, but only in a section of the switching process in which experience shows that relevant status information can be expected.

[0051] In this exemplary embodiment, the evaluation device 6 comprises a neural network. In particular, an autoencoder is used as the neural network in this exemplary embodiment. The autoencoder can detect deviations from a "normal" state. The results of machine learning can thus be used to provide the state information. In alternative embodiments, a model-based observer is used instead of the neural network to estimate the state of the solenoid valve 2 and to output the state information. Deviations from the simulated physical behavior are monitored. Deviations in the current / time behavior can be inferred from a solenoid valve malfunction.

[0052] The system 1 uses the status information to control the supply voltage for the solenoid valve 2. In alternative embodiments, the supply voltage is not controlled depending on the status information, but the status information is only used to detect states of the solenoid valve 2 that require other actions, such as immediate repair, consideration during the next maintenance, or similar.

[0053] Fig. 2 shows a diagram with various temporal current profiles of the supply current I during a switching operation. The abscissa represents the time after the activation of the solenoid valve 2 in milliseconds [ms], and the ordinate represents the supply current I of the solenoid valve 2 in amperes [A]. These temporal profiles of the supply current I are recorded by the current detection device 5, made available to the evaluation device 6, and evaluated by the evaluation device 6.

[0054] The diagram shows possible ranges in which the supply currents I vary depending on the time at which solenoid valve 2 is actuated. In particular, the area hatched with vertical lines shows supply currents I that result when solenoid valve 2 is actuated at an inlet pressure of 0 bar to 2 bar. The area hatched with horizontal lines shows supply currents I that result when solenoid valve 2 is actuated at an inlet pressure of 2 bar to 4 bar.

[0055] The time course of the respective supply current I can essentially be divided into three sections i, ii, iii.

[0056] In a first section i, the supply current I increases. A gradient G of the increase in the supply current I corresponds to an electrical constant, such as an impedance of the solenoid coil of the solenoid valve 2. During this increase, the armature of the solenoid valve 2 does not yet move.

[0057] In a second section ii, the armature of the solenoid valve 2 moves from a time B, which depends on the input-side pressure of the solenoid valve 2 and is shown as an oblique line. After the armature movement begins, the supply current I initially drops and then rises again. Finally, in a third section iii, the supply current I reaches an expected current value that is present when the solenoid valve 2 is open and which can lie within a predetermined range. Both the time B and the subsequent level of the supply current I, particularly in the second section ii, depend on the input-side pressure of the solenoid valve 2. However, the basic temporal progression of the supply current I is characteristic of the solenoid valve 2.

[0058] If the detected temporal profile of the supply current I corresponds to a typical, namely the characteristic temporal profile of the supply current I of a normally operating solenoid valve 2, it is assumed that the solenoid valve 2 is switching normally, i.e., that there is no malfunction in the solenoid valve 2, and corresponding status information is output. This enables the evaluation device 6 to monitor the effectiveness of the solenoid valve 2.

[0059] A fault occurs, for example, if the current strength is detected as zero, in which case it can be assumed that there is, for example, a line break in a supply line or within the solenoid valve 2 and corresponding status information is provided.

[0060] If an increase in the supply current I above the stored value, which is usually present at the beginning of the armature movement, is detected, it can be concluded that the armature is not moving and the solenoid valve 2 is blocked and therefore defective, so that corresponding status information of such a fault is provided.

[0061] If the supply current I curve shows an unusual increase in the supply current I at a specific time or period, it can be concluded that increased friction is present within the solenoid valve 2. The evaluation device 6 then provides the status information for a maintenance request.

[0062] In a further development of this embodiment, a previously stored empirical value with regard to future wear is included and status information with regard to an expected service life of the solenoid valve 2 is provided.

[0063] From the temporal progression of the supply current I at the time B, when the armature of the solenoid valve 2 begins to move, the evaluation device 6 also outputs the input-side pressure applied to the solenoid valve 2 and provides corresponding status information. The input-side pressure is determined from empirical values. In alternative embodiments, either all of the above

[0064] Status information is provided, or only a selection of it.

[0065] Fig. 3 shows a flowchart of a method for monitoring the solenoid valve 2 of the vehicle's compressed air system.

[0066] During operation, the monitoring device 3 detects the supply current I of the solenoid valve 2 in a step S1 by means of the current detection device 5.

[0067] This detected supply current I is then made available to the evaluation device 6 in a step S2 and evaluated by the evaluation device 6 in the form of status information of the solenoid valve related to the temporal progression during the duration of a switching operation of the solenoid valve 2, with data about the switching operation optionally being provided by the control device 4. The supply current I of the solenoid valve 2 flowing during the entire switching operation is evaluated, i.e., from the time at which the solenoid valve 2 is actuated with a supply voltage and until the switching operation of the solenoid valve 2 is completed, i.e., until an expected current for the solenoid valve is reached. In alternative embodiments, only the supply current I flowing in a partial range of the switching operation is evaluated. The partial range of the switching operation is determined on the basis of empirical values.For example, an area is evaluated in which the input-side pressure can be determined from the current curve.

[0068] The evaluation device 6 then provides the status information related to the temporal progression in a step S3.

[0069] For the functional monitoring of the solenoid valve, i.e., the effective monitoring, a neural network is used in one embodiment. This neural network uses a feed-forward neural network, in which information flows only in one direction without loops, or an autoencoder. When neural networks are used, training takes place beforehand. Data corresponding to correct ("normal") operation is collected with a statistical dispersion. This data is then used to train an anomaly detection system. In a further step, different fault patterns are simulated during training, which can be used to classify the fault patterns. The training or learning takes place as so-called "supervised learning," in which the data is assigned a so-called label based on previously acquired knowledge.

[0070] The autoencoder features a neural network that adapts so-called hyperparameters during learning so that, despite compression, the network's reconstructed output data remains very close to the input data. A deviation of the reconstructed output data from the input data is a measure of the quality of the function monitoring, namely, how close the input data is to a "normal / correct" function of solenoid valve 2. By deliberately generating faulty data sets, the anomaly detection function can be tested. At the same time, the faulty data sets are used to train a classifying neural network.

[0071] A so-called convolutional neural network is used to determine the input pressure and remaining lifetime. In alternative embodiments, other types of neural networks are used.

[0072] Machine learning, in particular so-called supervised learning, is carried out by means of the neural network of the evaluation device 6. In this case, by entering data at an input level, a result is calculated at an intermediate level, which is then output at an output level. During learning or training, i.e. when data is entered and the expected result for the entered data is checked, the parameters for the calculation are gradually changed by the neural network so that the desired result is achieved. By repeating the training several times, the parameters are optimized so that the result is achieved more reliably. Furthermore, other influencing factors can also be subsequently added as input, which are then taken into account when determining the result.With existing training data, for example, a prediction of the function can be derived, a prediction of the remaining service life can be determined and an inlet pressure of the valve can be estimated.

[0073] The graph shown in Fig. 2 is used as input, and the image of the graph is used as input both for training and in the application, using a convolutional neural network, for example, which is primarily used in the machine processing of image or audio data. For graphs with different curves, the result is learned, which corresponds to the properties of the different curves, and the result is then provided as state information. The recorded current curves are thus evaluated by the neural network.

[0074] The status information is either displayed on a central display for the respective solenoid valve 2 or, alternatively, the status information is transmitted to a central computer.

[0075] Furthermore, an expected service life of the solenoid valve 2 is provided as the state information from the time course of the supply current I.

[0076] In addition, a failure probability of the solenoid valve 2 and / or the maintenance requirements for the required maintenance of the solenoid valve 2 or a replacement of the solenoid valve 2 are provided as the state information from the time course of the supply current I.

[0077] Furthermore, the pressure applied to the input side of the solenoid valve is provided as the status information. Finally, a result of the effective monitoring of the switching function of solenoid valve 2 is provided as the status information.

[0078] In alternative embodiments, not all state information is output, but only one or a few of the state information.

[0079] If the monitoring results in a positive result, namely the proper functioning of solenoid valve 2, the supply voltage for solenoid valve 2 is optionally reduced, thus also reducing the supply current. This makes it possible to operate solenoid valve 2 at a reduced power level to save energy and increase the service life of solenoid valve 2. The supply voltage is iteratively reduced as long as reliable function of solenoid valve 2 is guaranteed.

[0080] On the other hand, the supply voltage for solenoid valve 2 can optionally be increased if the result of the effective monitoring shows that solenoid valve 2 is not switching properly in order to ensure its function again.

[0081] Although the present invention has been described with reference to specific features and embodiments, it is obvious that various modifications and combinations may be made thereto without departing from the spirit and scope of the invention. Accordingly, the description and drawings are to be considered merely as an illustration of the invention as defined by the appended claims and are intended to cover all modifications, variations, combinations, or equivalents that fall within the scope of the present invention. LIST OF REFERENCE SYMBOLS

[0082] 1 system

[0083] 2 Solenoid valve 3 Monitoring device

[0084] 4 Control device

[0085] 5 Current measuring device

[0086] 6 Evaluation device B Time when the armature of the solenoid valve starts to move

[0087] G Gradient

[0088] I supply current i first section of the time profile of the supply current ii second section of the time profile of the supply current iii third section of the time profile of the supply current

Claims

PATENT CLAIMS 1. Monitoring device (3) for a solenoid valve (2) of a compressed air system of a vehicle, comprising: a current detection device (5) configured to detect an electrical supply current (I) of the solenoid valve (2), and an evaluation device (6) configured to evaluate a temporal profile of the detected supply current (I) during a switching operation of the solenoid valve (2) and to provide status information about the state of the solenoid valve (2) that depends on the temporal profile.

2. Monitoring device (3) according to claim 1, wherein the evaluation device (6) is configured to evaluate the temporal profile of the supply current (I) during an entire switching operation of the solenoid valve.

3. Monitoring device (3) according to claim 2, wherein the evaluation device (6) is configured to determine whether or not a fault exists as status information during the evaluation of the entire switching process of the solenoid valve (2) and to provide this status information as a result of functional monitoring.

4. Monitoring device (3) according to one of the preceding claims, wherein the evaluation unit (6) is configured to detect a "maintenance request" and to provide the status information for required maintenance.

5. Monitoring device (3) according to one of the preceding claims, wherein the evaluation unit (6) is configured to monitor the temporal progression of the to close the supply current to an input-side pressure at the solenoid valve (2) and to provide this pressure as the status information.

6. Monitoring device (3) according to one of the preceding claims, wherein the evaluation unit (6) comprises a neural network which is configured to independently learn typical current patterns for a correct state of the solenoid valve (2) as well as atypical current patterns for malfunctions of the solenoid valve (2) and to provide them as the state information.

7. System (1 ) comprising a solenoid valve (2), a monitoring device (3) according to one of the preceding claims and a control device (4) configured to control the solenoid valve (2), wherein the system (1 ) is configured to control a supply voltage for the solenoid valve (2) depending on the status information provided by the monitoring device (3).

8. Method for monitoring a solenoid valve (2) of a compressed air system of a vehicle comprising the steps: Detection of a supply current (I) of the solenoid valve (2) by means of a current detection device (5) (S1 ); Evaluating the temporal profile of the detected supply current (I) during the duration of a switching operation of the solenoid valve (2) by means of an evaluation device (6) in the form of a state information of the solenoid valve (S2) related to the temporal profile; and Providing status information related to the time course (S3).

9. Method according to claim 8, wherein the evaluation of the course of the supply current (I) is carried out using a neural network based on results trained by machine learning.

10. Method according to claim 8 or 9, wherein the supply current (I) is evaluated during an entire switching operation of the solenoid valve (2).

11. Method according to any one of claims 8 to 10, wherein an expected lifetime of the solenoid valve (2) is provided as the status information.

12. Method according to any one of claims 8 to 11, wherein a failure probability of the solenoid valve (2) is provided as the status information.

13. Method according to one of claims 8 to 12, wherein a result of an operational monitoring of a switching function of the solenoid valve (2) is provided as the status information.

14. Method according to any one of claims 8 to 13, wherein a maintenance request for required maintenance is provided as the condition information.

15. Method according to one of claims 8 to 14, wherein a pressure applied to the input side of the solenoid valve (2) is provided as the status information.

Images