Method for carrying out a cleaning process and monitoring module

Abstract

The invention relates to a method for carrying out a cleaning process for cleaning the interior of a container (10), in which a cleaning fluid is sprayed in a periodically repeated jet pattern against the inner wall of the container (10) by means of a jet cleaner (20) introduced into the interior of the container (10) with a nozzle head (24) rotating about a main axis of rotation, wherein sensor values ​​are recorded by means of a sensor and output to a control unit (36). The invention is characterized in that the sensor is a microphone (34) which is configured and installed to record digital audio data of noises inside the container, wherein the audio data is fed to an analysis and assessment unit implemented in software in the control unit (36), analyzed by the latter, and the corresponding cleaning process is assessed as proper or improper on the basis of a result of this analysis. The invention further relates to a monitoring module (30) for use in such a cleaning method.

Description

Field of the invention

[0001] The invention relates to a method for carrying out a cleaning process for cleaning the interior of a container, in the course of which a cleaning fluid is sprayed against the inner wall of the container in a periodically repeated jet pattern by means of a jet cleaner introduced into the interior of the container and having a nozzle head rotating about a main axis of rotation, wherein sensor values ​​are recorded by means of a sensor and output to a control unit.

[0002] The invention further relates to a monitoring module for use in such a method. State of the art

[0003] A generic method is known from WO 2019 / 241261 A1.

[0004] The internal cleaning of containers, such as tanks, using so-called jet cleaners has long been known to those skilled in the art. A distinction is made between so-called surge cleaners and so-called targeted jet cleaners, both of which will be summarized here under the term jet cleaner. In both cases, a nozzle head connected to a cleaning fluid supply via fluid lines is introduced into the interior of the container to be cleaned. This can be done temporarily, specifically for the purpose of cleaning; alternatively, variants with a nozzle head permanently mounted inside the container are also known. The nozzle head is designed to spray cleaning fluid from the supply in a jet against the inner wall of the container, rotating during the process to reach as many wall areas as possible. The corresponding rotation drive can be motor-driven or caused by the recoil of the sprayed fluid.A surge cleaner is typically used to spray a jet, also known as a surge, with a comparatively large volume flow rate under comparatively low pressure. The rotation of the nozzle head is typically one-dimensional, i.e. around an axis of rotation that does not rotate relative to the container. The resulting jet pattern is comparatively simple; coverage of the entire inner wall surface is achieved by a correspondingly wide jet profile. A directional jet cleaner, on the other hand, typically produces a high-pressure jet with a lower volume flow rate. The nozzle head rotates around several axes of rotation simultaneously and reaches the inner wall by directing a jet with a comparatively narrow jet profile in a complex jet pattern. Here, too, there is usually an axis of rotation that does not rotate relative to the container and is referred to here as the main axis of rotation.In any case, the jet pattern is a periodic jet pattern, i.e., one that repeats at regular intervals. The period from the start of one jet pattern to the start of the next is referred to as the jet pattern period. The time required for one revolution of the nozzle head around its main axis of rotation is referred to as the period of rotation.

[0005] From the aforementioned generic document WO 2019 / 241261 A1, a surge cleaner is known whose rotatable nozzle head is additionally mounted for linear movement on a rail arrangement that can be inserted into the container interior. Due to the linear mobility of the nozzle head inside the container, the entire surface of the container's inner wall can be reliably reached, even in very long containers, such as railway tank cars in particular. In order to be able to specifically control the linear movement of the nozzle head, the previously known nozzle head is equipped with a laser-based distance meter, by means of which the current relative position of the nozzle head to other objects in the container, in particular to the container's inner wall, can be determined. Furthermore, the aforementioned document also describes the use of the distance meter to locate contamination detected by other means, for example, by an additional image sensor on the nozzle head.

[0006] WO 2018 / 218120 A1 discloses a jet cleaner whose nozzle head is also equipped with a dirt thickness sensor. This sensor can operate in various ways, particularly according to optical or acoustic principles, and serves to detect the thickness of a layer of dirt on the inner container wall that is to be removed during the cleaning process.

[0007] A disadvantage of these methods is the lack of control over whether a cleaning process is proceeding properly or whether errors occur, for example, due to mechanical blockage of the nozzle head and / or interruption of the cleaning fluid flow. Only after the cleaning process has been completed can the result be determined using additional sensors, and the cleaning process subsequently classified as proper or improper. However, detecting errors afterward is very time-consuming.

[0008] DE 10 2020 131 124 A1 discloses an acoustic method for externally detecting deposits on the inner walls of pipelines. For this purpose, a surface vibration is applied to the pipeline from the outside, which is measured at a distance from the source. The changes the wave experiences during its propagation path allow conclusions to be drawn about the pipeline section being passed through, in particular the presence of deposits. The conclusions drawn are used to assess the success of a current pipeline flushing process.

[0009] DE 10 2016 104 980 A1 discloses an acoustic monitoring system for the operation of a household appliance, particularly a washing machine. A microphone located in the appliance housing records the general operating noises and compares them with stored noise patterns that are typical for proper operation. In the event of deviations, warning signals can be generated and / or the appliance's operation can be interrupted. Task

[0010] It is the object of the present invention to further develop a generic method in such a way that the assessment of the cleaning process as proper or improper can be carried out while it is being carried out. Description of the invention

[0011] This object is achieved in conjunction with the features of the preamble of claim 1 in that the sensor is a microphone which is set up and installed to record digital audio data of noises inside the container, wherein the audio data is fed to an analysis and assessment unit implemented in software in the control unit, is analyzed by the latter and the corresponding cleaning process is assessed as proper or improper on the basis of a result of this analysis.

[0012] For this purpose, in particular a monitoring module with the features of claim 11 can be used, ie a monitoring module for use in a method for carrying out a cleaning process for cleaning the inside of a container, in the course of which a cleaning fluid is sprayed in a periodically repeated jet pattern against the inner wall of the container by means of a jet cleaner introduced into the interior of the container with a nozzle head rotating about a main axis of rotation, comprising - a housing that can be flanged to the outer wall of the container, - a microphone arranged in the housing, which is designed to record digital audio data of noises inside the container when the housing is flanged to the outer wall of the container, and - a control unit arranged in the housing and operatively connected to the microphone, with a software-implemented analysis and evaluation unit for analyzing and evaluating audio data recorded by the microphone.

[0013] Preferred embodiments are subject of the dependent claims.

[0014] The basic idea of ​​the present invention is to acoustically monitor the cleaning process inside the container by recording and automatically analyzing the noises generated during cleaning. Depending on the result of the analysis, the corresponding cleaning process is then assessed as proper or improper, and suitable compensatory or corrective measures can be initiated, particularly in the latter case. In particular, in the simplest case, a warning signal can be generated that indicates the need for intervention to operating personnel. More complex measures are also conceivable, in particular the initiation of more detailed testing or automated correction routines.

[0015] In principle, it is conceivable to conduct the automated analysis according to criteria explicitly specified by the user. For example, target values ​​can be specified for physically descriptive parameters such as volume, pitch, pattern repetition frequency, etc., and compared with the actually measured values ​​for these parameters. However, this is disadvantageous in that the overall acoustic result of a cleaning process depends heavily on the circumstances of the individual case. For example, the same jet cleaner, used consecutively to clean different containers, may produce a completely different acoustic scenario each time despite the same control. Likewise, the acoustic scenarios that arise in the same container when different jet cleaners are used can be very different.Measuring a target value table “by hand” in advance for each constellation and storing it as a reference involves considerable effort.

[0016] It is therefore considered more advantageous to implement the software and assessment unit as an AI (artificial intelligence) algorithm. In particular, implementation as an AI algorithm operating according to the support vector machine (SVM) principle has proven successful. This type of AI application is familiar to those skilled in the art. Corresponding algorithms can be found as open-source programs, although they still require adaptation to the specific problem. In this case, such an AI algorithm can be trained, for example, by using a monitoring module installed on the container to record the acoustic appearance of one or more cleaning processes and feed it into the AI ​​algorithm. A separate examination is performed after each recorded cleaning process to determine whether it was carried out correctly or not. The results of this examination are also fed into the AI ​​algorithm.It is particularly advantageous if the supplied data includes both properly and improperly performed cleaning processes. The KL algorithm then independently determines criteria by which a distinction can be made and stores them as references. The user does not need to have explicit knowledge of these criteria and / or reference values. With a KL algorithm as a software-implemented analysis and assessment unit, future cleaning processes can then be reliably categorized and assessed as proper or improper, adapted to the circumstances of the individual case.

[0017] To avoid changing these individual circumstances between different cleaning processes, it is advantageous to permanently install a separate monitoring module on each container to be cleaned and to train its analysis and assessment unit accordingly. This is readily possible since the components required to construct a monitoring module according to the invention are readily and inexpensively available.

[0018] The audio data from the cleaning process is preferably stored and processed as WAVE files, particularly .wav files. These are standards known to those skilled in the art for data recording, particularly acoustic vibrations, which can be provided by most commercially available digital microphones and read and processed by appropriately configured audio algorithms.

[0019] The microphone preferably operates at a sampling rate between 10 kHz, particularly 20 kHz and 80 kHz, and especially between 40 kHz and 60 kHz. These frequency ranges have proven to be sufficiently high for a differentiated analysis and reliable assessment of typical cleaning processes, while also being low enough to be handled with cost-effective electronic hardware. It has proven sufficient for the microphone to record the audio data as a sequence of sound level measurements. The term "sound level" is to be understood broadly here and depends on the respective microphone type. In particular, sound pressure values, sound intensity values, or sound power values ​​can be measured or recorded.

[0020] Ideally, the audio data is fed to the software-implemented analysis and evaluation unit as a plurality of successively recorded files, each representing a sound sequence with a length corresponding to a multiple of the period of rotation of the nozzle head around the main axis of rotation. This corresponds to a kind of averaging of the acoustic scenario and suppresses the significant acoustic differences that can occur at the location of the microphone during the rotation of the nozzle head around the main axis of rotation. However, these averaging windows must not be too long; this would be detrimental to the response speed of the monitoring system.

[0021] It has proven particularly advantageous if the audio data is fed to the software-implemented analysis and assessment unit as a plurality of successively recorded files, each representing a noise sequence with a length between 1 and 10 seconds, particularly between 2 and 5 seconds, and especially 3 seconds. For typical jet cleaners with their typical rotation frequencies, such a selection represents a good compromise between sufficient acoustic averaging on the one hand and sufficient response speed of the monitoring system on the other.

[0022] In principle, the positioning of the microphone relative to the container is largely arbitrary, as long as acoustic coupling is present. However, it has proven particularly advantageous if the microphone is installed on the outer wall of the container, in particular if it is flanged there and is acoustically coupled to the interior of the container via a fluid-tight and acoustically permeable window. Even better acoustic coupling would be achieved if installed inside the container. However, this is disadvantageous for obvious practical reasons. Installation away from the container, on the other hand, would compromise the acoustic coupling. Fixed installation on the outer wall is therefore considered optimal overall. The additional provision of an acoustic window, which of course must not lead to a critical local weakening of the container wall, represents a further improvement in the acoustic coupling.

[0023] Preferably, the control unit is contained together with the microphone in a monitoring module, which, as mentioned, is preferably flanged or can be flanged to the outer wall of the container. This monitoring module preferably also has a signaling unit operatively connected to the control unit, by means of which the current status of the assessment of the cleaning process is or can be signaled visually and / or acoustically. In particular, the signaling unit can have an LED display integrated into the wall of the housing.

[0024] Despite the microphone and control unit preferably being housed together in the housing of the monitoring module, it is preferably provided that the interior of the housing is divided into at least two acoustically isolated chambers, one of which houses the microphone and the other the control unit. This acoustic decoupling is particularly important in the usual cases where the control unit, e.g., a microcontroller, is equipped with its own cooling device, particularly in the form of a fan. The noise generated by the fan in the immediate vicinity of the microphone and picked up by the latter can be disruptive. Especially if the fan is switched on and off based on cooling requirements, this can adversely affect reliable training and / or reliable operation of the analysis and assessment unit. The spatial and acoustic separation within the housing counteracts this.

[0025] Further details and advantages of the invention will become apparent from the following specific description and drawings. Brief description of the drawings

[0026] They show: Fig. 1: a highly schematic sectional view of a container during its cleaning according to the invention, Fig. 2: a highly schematic sectional view of a monitoring module according to the invention. Detailed description of preferred embodiments

[0027] The same reference symbols in the figures indicate the same or analogous elements.

[0028] Fig. Figure 1 shows a highly schematic sectional view through a container 10, in particular a tank, such as is used, for example, in the pharmaceutical or food industry, during a cleaning process according to the invention. The container 10 has a main nozzle 12 through which the rod 22 of a surge cleaner 20 is introduced into the interior of the container 10. The rod 22 is connected, in a manner not shown, to a supply of cleaning fluid, from which the cleaning fluid can be pumped to a nozzle head 24 arranged at the end of the rod 22. The rod 22 thus serves both to position the nozzle head 24 and to supply cleaning fluid. At the outlet of the nozzle head 24, a jet 26 of cleaning fluid is produced, which, in the illustrated surge cleaner 20, has a high volume flow and a very wide jet profile.As indicated by the rotation arrow 28, the nozzle head 24 rotates during the cleaning process around the main rotation axis defined by the rod assembly 22, which in this case represents the only rotation axis. In this way, the inner wall of the container 10 is sprayed all around with the cleaning fluid.

[0029] An acoustic window 14 is arranged in the wall of the container 10, through which a monitoring module 30, described below, "listens" into the interior of the container 10. The window 14 can, for example, consist of a thin, metallic membrane that is integrated into the wall of the container 10 in a fluid-tight manner. The membrane of the acoustic window 14 is intended to enable acoustic coupling between the interior and exterior of the container 10 without allowing fluid to enter or exit. The specific design and positioning of the acoustic window 10 must be selected by the person skilled in the art in light of the circumstances of the individual case, in particular the fluids used, the pressures used, and other parameters.

[0030] Fig. 2 shows a highly schematic representation of a monitoring module 30, which is used in the Fig. 1, the window 14 is fixed to the outer wall of the container 10 in an arched manner or - as a separate structural unit - can be fixed there. The monitoring module 30 has a housing 31, which can be flanged to the outer wall of the container 10 via a flange 32. The interior of the housing 31 is divided transversely to its axial direction by a partition wall 33. This preferably provides acoustic decoupling between the two sections of the housing 31 separated by it. Fig. 2 lower, ie flange-near, part of the housing 31 is a microphone 34. This is preferably on a Fig. 2, which closes the flange-side opening of the housing 31, is fixed to the coupling plate 35 shown below. When the monitoring module 30 is flange-mounted to the outer wall of the container, this coupling plate 35 rests against the window 14, thus creating a continuous acoustic coupling between the interior of the container and the microphone 34.

[0031] The microphone 34 generates digital audio data which is forwarded to a control unit 36 ​​which is located in the Fig. 2 upper part, i.e., above the partition wall 33. The electronic control unit 36 ​​is thus largely acoustically decoupled from the microphone 34. This means that any interference noise, such as that generated by a fan of the control unit 36, cannot reach the microphone 34 or can only do so to a significantly reduced extent.

[0032] The control unit 36 ​​is in turn connected to a signal unit 37, which in the embodiment shown is realized as an LED strip on the upper edge of the housing 31.

[0033] As already discussed in detail in the general part of the description, the noises generated during the cleaning process are recorded by microphone 34 and forwarded to control unit 36, where these noises are automatically categorized and the cleaning process is assessed as proper or improper. The result of this assessment can then be displayed by appropriately controlling display unit 37.

[0034] Of course, the embodiments discussed in the specific description and shown in the figures are only illustrative embodiments of the present

[0035] In the light of the disclosure herein, a broad spectrum of possible variations is available to the person skilled in the art. List of reference symbols 10 containers 12 nozzles 14 windows 20 surge cleaners 22 rods 24 nozzle head 26 beam 28 Rotation arrow 30 Monitoring module 31 housings 32 flange 33 Partition wall 34 Microphone 35 coupling plate 36 Control unit 37 Signal unit

Claims

[1] Method for carrying out a cleaning process for cleaning the interior of a container (10), in the course of which a cleaning fluid is sprayed in a periodically repeated jet pattern against the inner wall of the container (10) by means of a jet cleaner (20) introduced into the interior of the container (10) with a nozzle head (24) rotating about a main axis of rotation, wherein sensor values ​​are recorded by means of a sensor and output to a control unit (36), characterized by , that the sensor is a microphone (34) which is arranged and installed to record digital audio data of sounds inside the container, wherein the audio data is fed to an analysis and assessment unit implemented in software in the control unit (36), analyzed by the latter, and the corresponding cleaning process is assessed as proper or improper on the basis of a result of this analysis. [2] Method according to claim 1, characterized by that the software-implemented analysis and evaluation unit is a Kl algorithm operating according to the support vector machine principle. [3] Method according to one of the preceding claims, characterized by that the audio data is saved and processed as WAVE files, in particular as .wav files. [4] Method according to one of the preceding claims, characterized by that the microphone (34) records the audio data with a sampling rate between 10 kHz and 100 kHz, in particular between 20 kHz and 80 kHz, in particular between 40 kHz and 60 kHz. [5] Method according to claim 4, characterized by that the microphone (34) records the audio data as a sequence of sound level measurements. [6] Method according to one of the preceding claims, characterized bythat the audio data are supplied to the software-implemented analysis and assessment unit as a plurality of successively recorded files, each representing a sound sequence of a length corresponding to a multiple of a period of rotation of the nozzle head (24) about the main axis of rotation. [7] Method according to one of the preceding claims, characterized by that the audio data are fed to the software-implemented analysis and assessment unit as a plurality of successively recorded files, each representing a sound sequence of a length between 1 and 10 seconds, in particular between 2 and 5 seconds, in particular 3 seconds. [8] Method according to one of the preceding claims, characterized by that the microphone (34) is installed on the outer wall of the container (10) and is acoustically coupled to the interior of the container via a fluid-tight and acoustically permeable window (14). [9] Method according to one of the preceding claims, characterized by that the control unit (36) together with the microphone (34) is contained in a monitoring module (30) flanged to the outer wall of the container (10). [10] Method according to claim 9, characterized by that the monitoring module (30) has a signal unit (37) which is operatively connected to the control unit (36), by means of which the current status of the assessment of the cleaning process is signaled optically and / or acoustically. [11] Monitoring module (30) for use in a method for carrying out a cleaning process for cleaning the interior of a container (10), in the course of which a cleaning fluid is sprayed in a periodically repeated jet pattern against the inner wall of the container (10) by means of a jet cleaner (20) introduced into the interior of the container (10) with a nozzle head (24) rotating about a main axis of rotation, comprising - a housing (31) which can be flanged to the outer wall of the container (10), - a microphone (34) arranged in the housing (31) which is designed to record digital audio data of noises inside the container when the housing (31) is flanged to the outer wall of the container (10), and - a control unit (36) arranged in the housing (31) and operatively connected to the microphone (34) with a software-implemented analysis and assessment unit for analyzing and assessing audio data recorded by means of the microphone (34). [12] Monitoring module (30) according to claim 11, characterized by that a signal unit (37) is further comprised, which is arranged in the housing (31) and is operatively connected to the control unit (36), by means of which the respective current status of the assessment can be signalled optically and / or acoustically. [13] Monitoring module (30) according to claim 12, characterized bythat the signal unit (37) has an LED display integrated into the wall of the housing (31). [14] Monitoring module (30) according to one of claims 11 to 13, characterized by that the interior of the housing (31) is divided into at least two acoustically insulated chambers, in one of which the microphone (34) is arranged and in the other of which the control unit (36) is arranged.

Images