US20240003614A1

US20240003614A1 - Monitoring device for a temperature-controlled storage device, monitoring device having a protective container and method for operation

Info

Publication number: US20240003614A1
Application number: US18/038,424
Authority: US
Inventors: Amin Zayani
Original assignee: Individual
Current assignee: Individual
Priority date: 2020-11-26
Filing date: 2021-11-11
Publication date: 2024-01-04
Also published as: DE102020214846A1; WO2022111773A1

Abstract

The invention relates to a monitoring device (1) for a temperature-controlled storage device (3), for example a refrigerator, having a first detection device with a first sensor (4) for detecting the activity curve of a cooling or heating unit (3a) and having a processing device (5) which determines at least one characteristic variable for the temperature curve in the storage device (3) from the activity curve. The processing device can be designed as a learning system with artificial intelligence, which is suitably trained.

Description

The text below is a translation of the German patent application with the reference No. 10 2020 214 846.5, which has been filed on Nov. 26, 2020 with the German patent office and which hereby is incorporated by reference in this application.
The invention is in the field of electrical engineering and refrigeration or heating engineering and is suitable for practical use, for example, in temperature-controlled storage devices, such as refrigerators or heating containers.
Many substances, such as special chemical or pharmaceutically active substances, but also foodstuffs and substances or objects required for industrial processes must be stored under certain temperature conditions. These include not only the storage temperature, but also, where applicable, temperature profiles, minimum and/or maximum temperatures or maximum times for which exceeding or falling below certain specified temperatures is permissible in each case.
Storage devices such as refrigerators usually have aggregates for heating or cooling, the activity of which, i.e. the operation or power for heating or cooling, is controlled with a control device, wherein the actual temperature is usually measured for the control. The threshold temperatures for turning a unit on or off, as well as the heating or cooling capacity and efficiency of the units, can change over time, and the temperature measurement can be subject to error and drift. The hysteresis between the genset turn-off threshold and the turn-on threshold can also fluctuate. In practice, for example, refrigerators used in households show surprisingly large temperature fluctuations that are often not noticed because the refrigerated items do not necessarily spoil if the temperature fluctuations do not last too long. However, pharmaceutically sensitive substances can easily spoil or become ineffective when temperatures exceed or fall below certain levels, which is often not apparent or detectable to the end user. This is true, for example, of insulin, which diabetics often have to store in a refrigerator for a certain period of time. Insulin, for example, can become ineffective in a short time if the temperature falls below 2 degrees Celsius or rises above 8 degrees Celsius.
In view of this situation, the present invention is based on the task of creating a monitoring device which determines certain characteristics of the temperature curve for the past and the present with as little effort as possible and predicts them for the future, so that certain temperature parameters can be easily determined.
The task is solved according to the invention by a monitoring device for a temperature-controlled storage device with a first detection device with a first sensor for detecting the course of activity of a cooling or heating unit and with a processing device which determines at least one characteristic variable for the temperature curve in the storage device from the course of activity.
The first detection device uses the first sensor to record a course of the activity of a unit over time. The sensor can detect the intensity of the activity, i.e. a heating or cooling capacity, or merely the information that the respective unit is in heating or cooling mode. Usually, in many storage devices, only the aggregate is switched on and off for control, so that the temperature reached results from the duration of operation or from the duty cycle, i.e. the ratio of the phase durations of operation and the phase durations of inactivity.
The first detection device can thus also determine characteristics of the temperature curve exclusively from the digital information as to whether and for how long the unit is switched on and off in each case. For example, these variables can be used to determine how large the temperature differences are between the individual operating phases of the unit. This depends, for example, on the absolute temperature in the storage facility, the outside temperature and the maximum period of inactivity of the aggregate, as well as on the contents of the refrigerator, e.g., if uncooled substances have just been stored. A cooling or heating unit may consist of either a compressor or a heating device, but it may also include means for distributing a heating or cooling medium as a sub-unit. For example, refrigerators often include additional fans that serve to distribute air within the refrigerator. Similar circulating devices may be provided in heating devices. In refrigerators, where a liquid cooling medium is usually provided for the cooling circuit, a pump may also be provided for transporting this cooling liquid. Accordingly, activity monitoring of a refrigeration unit may include, for example, monitoring and detection of the activity of the compressor when the fan is off and/or monitoring as well as detection of the activity of the fan when the compressor is off and/or joint detection of the activities of the compressor and the fan during operation, and additionally optionally monitoring and detection of the activity of a coolant pump.
For this purpose, a reference measurement with a direct temperature measurement can first take place in order to generate reference data for the activity combinations described above or even just the activities of one or more of the elements mentioned, with which the activity course of the unit can later be compared in order to derive the temperature data from this. For example, reference measurements or training of the equipment can basically take place with the fan off and later analyses can focus on monitoring one compressor alone.
It is also possible to alternatively or additionally collect fan activity data in isolation as reference data on its own and later compare it to operational data to determine an aging condition. Ultimately, the fan and compressor can also be collected for determining reference data or for training in joint operation.
For a reliable interpretation of measurement data, the monitoring device can receive and take into account information about certain exceptional states or also control states such as de-frost phases. De-frost phases are used for regular, temporary targeted heating to remove icing. During such phases, the measurements of the monitoring device can be suspended or interpreted in a different way than during normal operation.
These phases can also be taken into account when training the monitoring equipment and can be omitted, for example, for the acquisition of training data.
In addition to the activity data of the respective unit, the temperature curve can also depend on the state, for example the aging state of the unit or of the system, for example also on the state of a cooling or heating medium. When the system ages, the temperature control changes its behavior so that, for example, the respective aggregate is switched on more often or less often or the respective activity durations until a temperature threshold is reached become longer or shorter. These data can also be acquired by the acquisition device and taken into account by the processing device. Furthermore, in the monitoring device, for example in the processing device, predetermined static data can also be stored which have an influence on the interpretation of continuously recorded data, such as a device type of the storage device, construction type, for example with or without ventilation, year of manufacture and operating mode of the storage device, for example in the case of a refrigerator, whether this is for daily use and is opened approx. 20 times per day, or whether it is a storage refrigerator that is opened only a few times per week or a professional pharmacy refrigerator that is opened very frequently but only within certain pharmacy opening hours.
This makes it possible to determine temperature trends during ongoing monitoring of the activities of a heating or cooling unit by the processing device, both for the past and for the future. This can be done by a direct comparison of the measured data with reference data, in particular based on a similarity metric, or also by a trained neural network or another automatic learning device or another device with artificial intelligence. For this purpose, the monitoring device receives data on the basis of which it can be seen whether there is significant misbehavior in operation. This can be done, for example, by signaling, for example inputting, a fault to the monitoring device when a fault is detected in operation. This may be the case, for example, during repairs.
The monitoring device then learns the activity patterns during training operation that lead to a repair. It can also be specified before training that a fault is assumed as soon as a certain threshold measured value of a detected variable is exceeded or undershot. This can be, for example, the duration of an activity phase of the compressor and/or fan or the duration between two activity phases of the compressor and/or fan. The threshold measurement value can also be a volume level of the activity of the compressor and/or fan or a volume level in a certain frequency range. Ultimately, an error can also be defined by the fact that a certain temperature value determined simultaneously at least during the training phase is exceeded or undershot at all or for a certain minimum duration.
Accordingly, in addition to the first detection device with the first sensor, a second detection device with a second sensor in the form of a temperature sensor can also be provided. In this case, the temperature is also detected directly, either only in a training phase or also continuously during operation. The temperature curve can also already be predicted for the future, since the state of the system can be determined precisely by comparison with reference data and thus both the aging state of the aggregate and, if necessary, the state of a heating or cooling medium as well as other boundary conditions can be determined and taken into account with or without their direct measurement.
To implement the first sensing device, it may be provided that the first sensor is an acoustic sensor, a vibration or acceleration sensor, a current or voltage sensor, or an electric or magnetic field sensor.
Each such sensor can basically detect the activity of a cooling or heating unit. For this purpose, either an operating noise or an electrical, magnetic or electromagnetic activity of a motor or a valve or switch is detected, for example.
From the running time of the aggregate per heating or cooling cycle, the efficiency can be determined and a signature, for example an acoustic signature, of the aggregate can also represent its aging condition. The acoustic signature may be in the form of the frequency spectrum of the operating noise of a compressor and/or fan or pump, reached for example at start-up or after a defined operating time, for example a few seconds, or before shutdown in a cooling phase. The signature may also consist of the evolution of the frequency spectrum during a period between start-up and shut-down of an element of the unit. For example, it can be monitored whether certain frequency components occur with high intensity in the spectrum during start-up and their intensity decreases towards the end or vice versa. However, it can also be decisive whether certain frequency components occur in the spectrum at all or not.
A sensor can also be used to detect the activity of a fan that is used, for example, to distribute or transport air in the storage device. For example, the noise of the air flow in the storage device and/or the running noise of the fan can be detected. This can be used to monitor the quality of the air distribution as well as the aging condition of the fan, for example, because the fan has an acoustic signature that reflects its aging condition or damage or defects.
The monitoring device can advantageously have a self-sufficient energy supply device that is independent of the energy supply of the storage device. This self-sufficient energy supply can be provided by a battery or a rechargeable battery, i.e. an electrochemical energy source, but also by an energy generation unit that generates electrical energy from temperature fluctuations, for example, from material movements caused by temperature fluctuations or from thermal voltages or from light radiation. Thus, the monitoring device can be operated without a connection to the storage device, for example electrically isolated from it and without an electrical coupling to the storage device. An installation effort for the monitoring device is thus eliminated, as is susceptibility to errors caused by faults in the storage device. The monitoring device can thus operate without a connection to the storage device and can simply be placed in the storage device for operation. It is also conceivable that a coupling to the energy supply of the storage device exists or is established exclusively for the regular charging of energy storage devices/batteries/accumulators. This can ensure that, during operation, the monitoring device is not connected to a 50 Hertz network and that the disturbances arising herefrom are avoided. Alternatively, a DC power supply can be used for power supply, which is very well electrically shielded and decoupled as well as acoustically isolated to suppress the interference frequency of 50 Hertz and harmonics thereof.
The monitoring device may additionally include a third detection device having a third sensor, wherein the third sensor detects whether the storage device is in an open or closed state.
This makes it possible to monitor whether and how long the refrigerator door is open and how often it is opened. In this way, the actual temperature can be determined or predicted. Reference data can also be used for this purpose, which can be obtained by repeatedly opening the storage device. For example, opening the refrigerator door can create a location-dependent temperature distribution in a refrigerator that influences the operating behavior of the refrigeration unit, in that the temperature sensor used to control the refrigerator is located either far away from the door or close to the door. A change in temperature due to the door being open is thus to be evaluated differently and leads to a different behavior of the control than a uniform increase in temperature when the refrigeration unit is inactive due to heat conduction. Monitoring the door or even other special occurrences, such as loading or unloading of the storage unit, can also be used when training the monitoring device to obtain the training data predominantly or exclusively during periods when the storage unit is not opened or loaded or unloaded, in order to focus the monitoring on the functioning of the compressor and/or the fan.
In principle, it can be provided that the third sensor is a light sensor or a position sensor for detecting the position of a door or a closing element of the storage device. Further sensors may also be provided for monitoring, for example an air humidity sensor in the storage device or in a protective container in the storage device. Humidity may be monitored along with other parameters to record or predict a comprehensive data profile for stored substances or items over time. One or more weight sensors may also be provided in shelves of the storage facility to record a loading or unloading event.
In addition, it may be provided that the processing device includes means for Fourier transforming the time-dependent data acquired by at least one of the sensors.
Since the undisturbed behavior of the temperature control basically has a periodic structure, it is possible to compress and structure the operating data by determining the operating frequency and other periods. In addition, this also anonymizes the measurement data, which can be useful, for example, when recording acoustic data, since this can also contain data subject to data protection, such as human communication. By means of a Fourier transformation, both the regularity of the switching on and off of the unit in the range of very low frequencies (cycle time e.g. 10 min or 20 min) and the acoustic signature or the frequency spectrum in the range of a few Hertz up to e.g. 16 kHz can be recorded.
It may also be useful to provide that the processing device has a storage device for the acquired data and/or the Fourier transformed data. This allows the data to be collected and evaluated over a longer period of time, which is particularly important when determining development trends over time.
In the case of the monitoring device, it can also be provided that the processing device has a device for comparing the acquired data or the Fourier-transformed data with reference data, which is set up, in particular trained, to determine from the comparison one or more characteristic variables of the temperature profile in the storage device and/or one or more characteristic variables for the state of the cooling or heating unit and/or the insulation state of the storage device or the state of a liquid coolant.
The recorded data can thereby, on the one hand, enable the assessment of the behavior of the storage device and the temperature for the past, in particular for the period of the recording of data. If necessary, time trends can also be used to determine or forecast characteristic values of the device, in particular the aging condition of the compressor, the fan, the cooling medium or drifts of the control system as well as the temperature or the temperature course, for the time before or after the acquisition of data—i.e. also for the future. From this, warning signals can be generated after automatic evaluation of the temperature course and determination of the exceeding or falling below of temperature and/or time thresholds. These can signal to an operator or user that conditions have occurred or are occurring in the past, present or future that endanger the unimpaired condition of substances or objects within the storage facility.
A simple implementation of such processing may provide that the processing device determines from the acquired data the temperature course in the storage device as well as, in particular, the course of the temperature of a substance or object stored therein. The nature of the stored substance, its heat capacity and, if applicable, a heat transfer coefficient and/or insulation may play a role. This can be particularly important if the substance or object stored in the storage device in turn still has a protective container or insulation container—possibly with its own considerable heat capacity—or is located in such an insulation container. For such cases, it is convenient to obtain corresponding reference data using a similar protective container, whereby the temperature of the stored substance can be measured inside the protective container for comparison. Such measurement data can also be used to appropriately train an AI module, i.e. a module with artificial intelligence, for example a neural network. Some of the aforementioned variables can also be determined from measurements of activity histories and/or temperatures. For example, the load condition can be determined by observing the activity period of the compressor and/or fan per cooling cycle set by the control system, or the rest period between two operating cycles of the compressor based on data obtained during training.
In the case of the monitoring device described, it can also be provided that it has a device for displaying or outputting the determined characteristic variables and/or the activity progress of a cooling or heating unit and/or an alarm signal. This can be, for example, a conventional digital display or a simple luminous display, which can also be designed as a color traffic light. A sound generator for the output of acoustic warning signals can also be provided.
The monitoring device may also have, for example, a transmitting and or receiving device for wireless communication, in particular for outputting alarm signals or display data on a terminal device. The wireless communication may use, for example, Ethernet, Wifi, long range radio, or a cellular communication standard or NB-IoT. Since storage devices often include metal enclosures, the wireless communications should either have an electrical feedthrough for an antenna line or an opening in the metal enclosure, such as a slot antenna or a glass window/non-metal window. In many cases, it will be sufficient to pass a thin antenna line out of the storage device enclosure, such as through a door seal.
In addition to a monitoring device of the type described above, the invention also relates to a protective device with a protective container for storing substances and/or objects and with a monitoring device of the type described above. Thereby, for example, the monitoring device can be integrated into the protective container and further, in addition to a temperature sensor in the storage device or in its place, a sensor for temperature monitoring can be provided inside the protective container and the measurement data of both temperature sensors can be processed by the processing device—both in the training phase and in normal operation. In particular, the protective container can have a container wall that is at least partially made of a material that exhibits a phase transition in a temperature range between 2 and 8 degrees Celsius. Moreover, the material may be integrated into the container wall, in particular enclosed therein, in such a way that the phase transition can be passed through several times without producing a change in shape of the container wall. It may be remembered that the container wall has one or more cavities in which the material undergoing a phase transition in the temperature range mentioned is located. The phase transition may take place between a solid phase and a liquid phase. In this case, the material is held in its liquid state in a cavity.
However, a material may also be provided that undergoes a phase transition between two different solid modifications in the critical temperature range so that no change in shape is to be feared. The protective container may be at least partially made of a typical insulating material, such as a foam or a natural insulating material such as cork or cork composite, or it may have evacuated cavities for thermal insulation.
In addition, a substance with a large specific heat capacity can be integrated into the protective container, for example inside the insulation, which has a storage capacity for a certain temperature and slows down temperature changes inside the protective container. The monitoring device, and in particular the protective container with such a monitoring device, may be connectable to a power supply of the storage device, but it/they may particularly advantageously have a self-sufficient power supply device that is independent of the power supply of the storage device. This self-sufficient energy supply can be provided by a battery or a rechargeable battery, i.e. an electrochemical energy source, but also by an energy generation unit which, for example, obtains electrical energy from temperature fluctuations, for example from material movements caused by temperature fluctuations or from thermal voltages or from light radiation. Bimetallic elements, Peltier elements or photosemiconductors/solar cells can be used for this purpose, for example. The Peltier element can, for example, be integrated into a wall of the protective container and convert the temperature differences between the interior and exterior of the protective container into electrical energy. In addition, it is useful to have a monitoring system for this self-sufficient energy supply, which issues alarm signals via a wireless connection if the energy supply to the monitoring device is not permanently present or is likely to be compromised.
The invention also relates to a method for operating a monitoring device of the type described above, in which it is provided that the measured temperature/time profiles and, in particular, also measured acoustic signals or measured signals from a vibration sensor are Fourier-transformed and compared in the Fourier-transformed form with reference data, and that at least one characteristic variable for the temperature profile in the temperature-controlled storage device is determined from the comparison.
The characteristic variable for the temperature curve can be, for example, a minimum or maximum temperature of a cooling cycle or also a point in time in the past or future at which certain temperature threshold values are undercut or exceeded or at which the duration of undercutting or exceeding a temperature threshold value exceeds a certain predefined duration.
Appropriate monitoring of a protective container located in the storage device and its interior can also be provided. The protective container, when it is in the storage device, can influence the temperature behavior of the storage device through its thermal properties, for example by slowing down temperature changes in the storage device through the heat capacity of the protective container.
The same applies to the temperature prevailing in the protective container: the behavior of this temperature depends on the temperature in the storage facility and on the heat capacity and insulating capacity of the protective container. In order to be able to determine the temperature curve inside the protective container or outside the protective container in the storage facility, reference data for these variables should be available in conjunction with corresponding, simultaneously measured activity curves of heating and/or cooling units, so that the determination of the temperature curve for the past, present or future is possible on the basis of the comparison data depending on the actual activities of the unit(s) and the other sensor data currently measured.
Finally, the invention also relates to a method for training a monitoring device of the type described above for monitoring a storage device, in which at least the temperature in the storage device, in particular or additionally in a protective container within the storage device, and the activity curve of a heating or cooling unit are continuously measured and reference data are determined therefrom. In the context of the invention, the determination of reference data can in principle also be understood to mean the training of a learning device which, after the corresponding training, can determine temperature curves for the past or future from measured values of the activity curve of a heating and/or cooling unit and, if appropriate, additionally continuously measured temperature data and, if appropriate, further additionally measured data. Further additionally measured data may be, for example, opening hours of the storage facility. The learning device may, for example, have a neural network or a software module containing AI (artificial intelligence) and this may make use of machine learning (supervised or unsupervised).
The determination of reference data or the training of the learning device can take place in exactly the same storage facility whose temperature profile is to be determined later or in a similar storage facility. The determination of reference data or the training can also take place in a different type of storage facility and then the reference data or the learning state can be transferred by a transfer function if both boundary conditions of the facility at which training is carried out or at which reference data are determined and boundary conditions of the facility at which the monitoring device is used are sufficiently known. The data on which the processing is based can be divided into static and dynamic, continuously acquired data. The static data are those that can be specified and are largely unchangeable. These include the construction, year of manufacture and type of equipment of the storage facility, its size, type of ventilation and also, for example, its geographical location, which is decisive both for the energy supply (110 volt supply voltage or 220 volt) and for the usual temperature fluctuations outside the storage facility, such as daily, weekly or seasonal fluctuations.
For example, for training, the following 3 types of pattern recognition/classification/training are conceivable:
Fridge Profiling:
A classical unsupervised machine learning takes place, for which about one week should be sufficient (5 weekdays and nights+weekend) to establish a basis for a single refrigerator. After one year in operation, a “long term profile” is then obtained, which also covers the seasonal variability).
Fault and Defect Detection:
The goal here is to determine a fault diagnosis for refrigerators, based on sensor data. For this, a more extensive training is necessary, where certain faults must be visible, induced or simulated (e.g.: fan fault, compressor fault, lack of refrigerant, etc.).
Event Detection:
Extensive training is also necessary for this, whereby special conditions that are to be detected later, e.g.: introduction of larger uncooled masses, etc., can be specifically induced.
The training can start as Fridge Profiling and lead to the working ability of the monitoring device after a short time. Later, the training can be extended to points 2 and 3 to enable the monitoring device to detect more complex faults or conditions in general. For training and/or operation, the monitoring device can partially perform processing of data itself, but at least partially acquired data can also be transmitted to a server via data communication, if necessary after preprocessing, and processed there. In the process, data acquired in the server from different monitoring devices can be compared or linked with each other, so that, for example, a broad base of training data is obtained, which accelerates the training processes of each individual monitoring device and also the subsequent operation, and improves the quality of the processing of data, in particular the quality of predictions.
The temperature can be measured continuously in the training phase and/or in the operating phase, i.e., for example, also exclusively in the training phase, at least in the storage device, when a protective container is used outside this protective container and/or inside the protective container. In a particularly advantageous embodiment, the temperature is measured at at least 2 points:
Inside the protective container, because it is a matter of quality assurance of medicines stored inside the protective container.
Outside the protective container in the storage facility/in the refrigerator): so that the temperature trends in the refrigerator are detected without distortion.
For example, the parameters minimum and/or maximum temperature of a cycle inside and outside the protective container, cycle length and duration of operation of a unit as well as the time between two operating phases of the unit can be determined continuously. In addition, frequency spectra of a heating/cooling unit or also of a fan detected by an acoustic sensor or vibration sensor can be determined and characterized by parameters, such as peak heights at characteristic frequencies, total intensity, partial intensity in certain frequency ranges, and the ratios/quotients or differences between these partial intensities or Fourier transforms and their characterization parameters. In addition, humidity values inside and/or outside the storage facility, their range of variation, minima, maxima and cycle times can be determined. All these quantities and, for example, their rates of change, as well as possibly other quantities, can be used as the basis for training and later monitored and compared with the training data or further processed with the trained processing device for the potential generation of signals.
It is thus possible to calculate the time over which the temperature in the protective container will remain below 8 or above 2 degrees in the event of malfunctions of the storage device. A prognosis signal can then be output, e.g.: “Attention, refrigerator is no longer cooling properly and the temperature in the protective container will remain between 2 and 8 degrees for xx hours”. Such a prognosis is particularly possible if the temperature inside as well as outside the protective container has been recorded and analyzed in each case during the training phase. In addition, it can be advantageous if the heat capacity of the protective container and the heat transfer values or thermal conductivity values are known for this purpose.

In the following, the invention is shown by means of figures of a drawing and described in the following. Thereby shows

FIG. 1 : a storage device in the form of a refrigerator,

FIG. 2 : a protective container in perspective view,

FIG. 3 : a protective container in a top view,

FIG. 4 : schematic of a monitoring device,

FIG. 5 : first measurement data serving as training data,

FIG. 5 a : a temperature curve with characteristic disturbances,

FIG. 6 : second measured data being evaluated,

FIG. 7 : third measured data, which are evaluated, as well as

FIG. 8 schematically shows a flow diagram of the operation of the monitoring device.

In FIG. 1 , as a typical application, a refrigerator 3 is shown as a storage facility in which a protective container 2, for example containing a drug such as insulin, is stored. Other possible storage facilities can be cold rooms, pharmacy refrigerators, but also warming containers, for example incubators or fermentation containers, for example for alcohol production. Such facilities typically have a heating or cooling unit that operates in intermittent temperature-controlled mode. In FIG. 1 , the cooling unit is designated 3 a.
Finally, a refrigerator door is labeled 3 b. Opening the refrigerator door interferes with the temperature control in such a way that additional or more frequent operation of the refrigeration unit becomes necessary.
The invention is based at least in part on the fact that the operation of the cooling or heating unit and the opening or closing of a door/opening can be monitored non-invasively, that is, without an electrical connection to the device.
For this purpose, the operation of the cooling unit 3 a can be monitored acoustically by a microphone, for example. In extreme cases, only the operation itself can be recorded, i.e. the start, end and duration of the individual operating phases. More informative is the monitoring of noise and frequency spectra by one or more microphones, which can complement each other by the suitable frequency characteristics. Condenser microphones, for example, which can be particularly well encapsulated, can be used. However, it is also possible to use other types of microphones such as MEMS microphones, which are used in cell phones, for example. It has been shown that the analysis of acoustic signatures picked up by microphones due to sound transmitted through the air is more successful than the analysis of structure-borne sound picked up by a microphone. For the analysis of structure-borne sound, a microphone can be brought into direct contact with a fixed component of the storage device. The analysis of structure-borne sound picked up in this way can, for example, be performed in addition to the analysis of sound picked up through the air, and structure-borne sound detection can identify noise that is primarily unrelated to a cooling or heating unit and can be eliminated for the analysis of sound picked up through the air.
Alternatively, a sensor may be provided to detect the power consumption of the unit, voltage, or an electric, magnetic, or electromagnetic field. A cooling unit may be designed as an electric compressor with a motor that, in addition to operating noises, i.e., acoustic signals, also consumes electricity and produces electric, magnetic, and electromagnetic fields. The cooling unit may additionally have a fan, the operating noise of which can be monitored, as can the airflow noise generated by the ventilation.
FIG. 2 shows a protective container 2 in the form of a box with a lid in which, for example, medicines can be stored. The protective container can have thermally insulating walls, for example with a foam or foam material, an insulating natural material such as cork or an aerogel. In addition, the protective container may also have a high thermal capacity element, such as a gel or water container, that dampens the temperature fluctuations in the protective container, i.e., equalizes the temperature, in the event of external temperature fluctuations in the storage facility. Both the additional insulation and the element with high heat capacity can be arranged in the area of the container walls, on the inner side of the walls of the container or in the walls of the protective container.
FIG. 2 shows a monitoring device 1 installed in the protective container 2, which will be discussed in more detail below. Side walls of the protective container are marked 2 a, 2 b. Such a monitoring device 1 can also be used by itself without a protective container for monitoring a storage facility.
FIG. 3 shows a protective container 2 in an opened view from above with side walls 2 a, 2 b as well as a phase transition material 2 c integrated in a side wall 2 a as a heat accumulator with an object or substance 11 stored therein and with a monitoring device 1. The material 2 c can have a phase transition in the range between 2 and 8 degrees Celsius, preferably between 2 and 4 degrees Celsius. The monitoring device 1 has a first sensor 4 that detects the activity of the cooling unit, for example a microphone, and a second sensor 6 for detecting the temperature in the environment of the protective container. This second sensor is located on the outside of the protective container 2. In the design shown in FIG. 2 , the monitoring device is integrated into the wall of the protective container in such a way that one part of the monitoring device faces the outside of the protective container, in particular forms part of the outer wall, and another part is in communication with the inside of the protective container. The monitoring device also has a third sensor 7 which monitors the opening state of a door of the refrigerator, i.e. is designed, for example, as a light sensor in order to detect either ambient light entering the storage device from the outside or the light of a lamp of the storage device itself, which is automatically switched on when the door is opened. However, the sensor for the opening state of the door could also be formed, for example, by a position sensor on the door or by a sensor for a draught in the storage device.
In addition, another temperature sensor 60 may be provided in the monitoring device to monitor the temperature inside the protective container and point toward the interior of the protective container.
In FIG. 4 a monitoring device 1 is schematically shown with some further elements. In addition to the first, second and third sensors 4, 6, 7 and the sensor 60 for the temperature in the protective container, the monitoring device has a processing device 5 for the sensor data, which is connected to the sensors via signal lines or wireless connections. The processing device comprises means 5 a for Fourier transforming the substantially periodic data. The Fourier transformation causes the data to be compressed, which also ensures, for example, that any acoustic data recorded via a microphone is sufficiently alienated so that, for example, randomly recorded human speech is only processed further in a highly alienated form. The processing device also has a memory device 5 b in which training/reference data is stored with which currently measured data is compared. The memory device may also be directly formed in part by a neural network in which data is stored by structuring the nodes and their connections through training. Results of comparison and estimation of the state of the storage device or its elements may also be stored in the storage device 5 b. The processing device further comprises a module 5 c, which may be a learning system if this is not already realized by a neural network and which includes elements of artificial intelligence. This module can assign to the acquired or Fourier-transformed data various states of the storage device or its elements, such as the heating/cooling unit or the insulation of the storage device or a heat transport fluid/cooling fluid states or parameters that can result in the output of an alarm signal. For this purpose, the processing device further comprises a display device 8 and/or a transmitting/receiving device 9 for wireless communication according to a common standard. Alarm signals can then be received by a terminal device 10, for example in the form of a smartphone or tablet computer or a wearable worn on the wrist, and signaled to a person visually, haptically or acoustically.
The following measured values or parameters can generally be recorded by the individual sensors:
The temperature and/or humidity sensor in the protective container allows a log to be created of the progress of these variables and is thus used, for example, for quality control of medications.
The temperature and/or humidity sensor in the storage facility outside the protective container allows monitoring of the facility's performance and, together with the recording of other variables, future forecasts.
Through the latter temperature sensor, the time cycles of cooling, cooling rate, maximum and minimum temperature of temperature control cycles can be detected.
Through an acoustic sensor, the compression cycles, the acoustic signature of heating and cooling units and of fans, volume and operating time within a compression cycle can be recorded. In addition, the switching of relays or on/off switches and the speed of fans can be detected acoustically.
A vibration and/or acceleration sensor can be used to monitor vibrations and movements in the environment of the storage device, vibrations caused by the operation of units, movements of the protective container and also the orientation of the protective container and the storage device. This means, for example, that incorrect positioning of the protective container or inadequate horizontal alignment of the storage device can be detected.
Overall, some variable quantities can be determined by using the sensors and learning the underlying structure in the training phase, such as the length and profile of temperature cycles, the cooling/heating efficiency of a unit or the entire storage facility, the thermostat drift as a function of time, and the usage profile when loading and unloading goods to be tempered.
For example, as a result of processing continuously acquired data based on static, predetermined parameters, training operations, and continuously acquired measurement data, the following 3 types of outputs/signals can be produced:
Reports on conditions of elements of the storage facility or substances/items stored in it, such as medicines, alarm signals to inform about malfunctions and initiate countermeasures, as well as indications and recommendations for optimization, for example warnings of impending malfunctions, forecasts and indications for optimized maintenance.
In FIG. 5 , time is plotted on the horizontal axis and various measurable signals are plotted on the vertical axis in a coordinate system. The step function 4 a represents the operating state of a cooling unit with the function values zero (switched off) and 1 (switched on). It can be seen that the unit according to FIG. 5 is switched on for a constant, short period of time.
The signal from the door opening sensor is indicated by 7 a. Only one period of time during which the door of the storage device/refrigerator is open can be seen in the diagram.
Curve 6 a shows the temperature curve in the storage device superimposed on the other signals. On the one hand, this is characterized by a slow temperature rise inside the storage device as long as the cooling unit is switched off. This is due to the heat transport through the insulation of the storage device from the warmer exterior space to the cooler interior space. During operation of the refrigeration unit, the temperature drops again in each case, and then slowly rises again after the unit is switched off. Opening the refrigerator door usually results in an upward temperature jump, denoted by 16 in the example shown, as warm air enters the refrigerator. If the refrigerator is only open for a short time, the temperature drops again partially after the door is closed, since the opening time is usually not sufficient to heat the solid components of the refrigerator, so that these partially cool the air in the refrigerator again after closing. Where 17 denotes a temperature rise that is due to the introduction of drugs that have a higher temperature than the refrigerator temperature. The temperature is lowered in the next operating phase of the refrigeration unit, but the heating effect of the drugs can continue for several cycles.
FIG. 5 a shows measured data for the temperature over a few hours and a larger number of operating cycles of a refrigeration unit, where 17 again indicates a malfunction due to storage of substances at a higher temperature and 18 indicates an operating malfunction in the control system of the refrigerator.
The described processes and their correspondence in the temperature course can be recorded and analyzed over a longer period of time, such as days and weeks. After a short time, the processing device itself can generate predictions about a future temperature course and compare these with the real temperature course. This trains the processing device to such an extent that the predictions about the temperature become reliable. It is also conceivable thereafter that the temperature is not measured further and the behavior of the storage device is determined from the behavior of the aggregate alone. Past states can also be determined in this way by analyzing stored activity data, so that it is also possible to determine whether the temperature has fallen below or exceeded certain thresholds in the past.
Certain additional information can also be stored in the processing device before, during or after the training, which is helpful for the analysis and prognosis, for example, how long a cooling unit of a certain model of a refrigerator should be in operation at maximum until the lower temperature threshold is reached, or how far apart two activity/operation phases of a unit should be at least in time. These variables depend, among other things, on the quality of the insulation, the performance and efficiency of the unit, the age of the cooling medium, and the load condition of the refrigerator, with the structure of the dependencies being similar for a number of refrigerators or models. Thus, the training time can be significantly reduced. The special cases mentioned, that the interval between two operating phases of the aggregate varies and that one operating phase of an aggregate is longer than usual, are exemplarily shown in the diagram of FIG. 6 .
FIG. 7 shows an example of a case in which the refrigerator door is left open (7 a) for a longer period of time after an activity phase 4 a of the refrigeration unit. This can, for example, lead to a temperature increase which triggers an alarm signal.
FIG. 8 schematically shows steps of the operating procedure of the monitoring device. In a first phase 12, the monitoring device is trained by continuous data acquisition of operating data and storage of known additional data. In the following steps 13, 14 and 15, operating data is acquired in each case and compared with the stored data structures. If a temperature threshold is exceeded in each case for currently measured operating data, an alarm signal is output. If no temperature threshold is exceeded and this is also not foreseeable for immediately upcoming periods, no alarm signal is output and operating data continues to be measured in steps 14, 15. If a temperature is continuously measured in the storage device and/or in the protective container, the processing device can continue to learn and continuously improve its predictions.
Longer-term trends can also be identified, such as an increase in the operating times of the unit as this decreases in performance or the coolant transports heat less well. Leaks in a compressor can also be detected or predicted at an early stage in this way.
Basically, in a first phase, the monitoring device can record the measured data with the available sensors and allow remote monitoring via a radio link. At the same time, the measured values can be stored in order to create suitable protocols that allow quality management for goods stored in the storage facility. At this stage, temperatures in the storage facility inside and outside the protective container can also be directly compared to confirm the effectiveness of the protection provided by the protective container. Already at this stage, certain critical situations can be indicated by alarm signals, such as a power failure, a door opening for too long, an incorrect temperature setting. In addition, it can be shown how long an acceptable temperature can be guaranteed in the protective container in the event of a temperature control failure.
In a second phase during and after a training phase, further information or signals may be generated by a processing device or a server supplied with the data, including the following: Suggestions for seasonal settings of the thermostat, information on predictive maintenance with this supporting information, for example, on cooling efficiency, indication of failure probabilities of units based on the acoustic signature, which depends on a degree of wear or damage that has occurred.
As training methods, for example, “supervised machine learning” can be used in the laboratory, in which the system is specifically brought into special states, for example error states, and correct or desired reactions are specified. Alternatively or additionally, especially after “supervised machine learning”, training based on “deep learning” can be carried out with a large number of situations that have actually occurred and the associated data. For this purpose, data can also be used which have not been processed by the individual processing device itself, but which originate from other systems/processing devices, are stored on a server to which the processing device is connected, and are applicable to the individual system at least to a certain degree.
In operation, the monitoring device or a protective container with a monitoring device can be supplied with energy by means of a battery or accumulator. This achieves a complete decoupling from the power supply of the aggregates or further elements of the storage device. Only for recharging the batteries can the operation be interrupted and a connection to an AC network be established, for example also by means of the internal supply lines of the aggregates of the storage facility.

Claims

1. Monitoring device (1) for a temperature-controlled storage device (3) with

a first detection device with a first sensor (4) for detecting the course of activity of a cooling or heating unit (3 a) and with

a processing device (5) which determines at least one characteristic variable for the temperature curve in the storage device (3) from the course of activity.

2. Monitoring device according to claim 1,

characterized in that

the first sensor (4) is an acoustic sensor, a vibration sensor, a current or voltage sensor or a sensor for an electric or magnetic field, wherein in particular a second detection device with a second sensor (6, 60) in the form of a temperature sensor is additionally provided.

3. Monitoring device according to claim 1,

characterized by a third detection device having a third sensor (7) for detecting whether the storage device (3) is in an open or closed state.

4. Monitoring device according to claim 3,

characterized in that

the third sensor (7) is a light sensor or a position sensor for detecting the position of a door (3 b) or a closing element of the storage device (3).

5. Monitoring device according to claim 1,

characterized in that

the processing device (5) comprises a device (5 a) for Fourier transformation and/or other signal processing of the time-dependent data acquired by at least one of the sensors (4, 6, 7, 60).

6. Monitoring device according to claim 1,

characterized in that

the processing device (5) comprises a storage device (5 b) for the acquired data (4 a, 6 a, 7 a) and/or the Fourier transformed data.

7. Monitoring device according to claim 1,

characterized in that

the processing device (5) has a device (5 c) for comparing the acquired data (4 a, 6 a, 7 a) or the Fourier-transformed data with reference data, which is set up, in particular trained, to determine from the comparison one or more characteristic variables of the temperature profile in the storage device (1) and/or one or more characteristic variables for the state of the cooling or heating unit (3 a) and/or the insulation state of the storage device (3) or the state of a liquid coolant.

8. Monitoring device according to claim 1,

characterized in that

the processing device (5) determines from the acquired data (4 a, 6 a, 7 a) the course of temperature in the storage device (3) and in particular the course of temperature of a substance or object (11) stored therein.

9. Monitoring device according to claim 1,

characterized in that

it has a device (8) for displaying or outputting the determined characteristic variables and/or the activity curve of a cooling or heating unit (3 a) and/or an alarm signal.

10. Monitoring device according to claim 1,

characterized in that

it has a transmitting and/or receiving device (9) for wireless communication, in particular for outputting alarm signals or display data on a terminal (10).

11. Protective device with a protective container (2) for storing substances and/or objects as well as a monitoring device (1) according to claim 1,

wherein in particular the monitoring device (1) is integrated into the protective container (2) and further in particular a sensor (60) for temperature monitoring is provided in the interior of the protective container (2), wherein in particular further the protective container has a container wall (2 a, 2 b) which consists at least partially of a material (2 c) which exhibits a phase transition in a temperature range between 2 and 8 degrees Celsius, and in that in particular the material is integrated, in particular enclosed, in the container wall (2 a, 2 b) in such a way that the phase transition can be passed through several times without producing a change in shape of the container wall.

12. Method of operating a monitoring device according to claim 1,

characterized in that

the measured data (4 a, 6 a, 7 a), in particular in the form of activity and temperature/time profiles and in particular also measured acoustic signals or measured signals of a vibration sensor, are Fourier transformed and compared with reference data, and in that at least one characteristic variable for the temperature profile in the temperature-controlled storage device (3) is determined from the comparison.

13. Method of training a monitoring device (1) according to claim 1 in a storage device (3),

characterized in

continuously measuring at least the temperature in the storage device, in particular in a protective container (2) within the storage device, as well as the course of activity of a heating or cooling unit (3 a), and determining reference data therefrom.