EP3512791B1 - Method for detecting a passenger entering a lift car of a lift assembly - Google Patents

Description

The invention relates to a method for recognizing that a passenger has entered an elevator car of an elevator installation according to the preamble of claim 1.

The WO 2013/130040 A1 describes a method for monitoring the use of an elevator system. In this process, the passengers in the elevator system are equipped with marking devices, so-called tags. Readers are attached to shaft doors or in the elevator cabins of the elevator system, which can recognize whether and, if so, which tag is in their vicinity. It can thus also be recognized when a passenger enters an elevator car. The reading devices forward the information to a traffic evaluation unit which, on the basis of this information, can monitor the use of the elevator system or record it for later analysis. The procedure according to the WO 2013/130040 A1 requires one day per passenger and at least one reader per landing door or per elevator car.

The US 2014/330535 A1 describes a method for detecting a movement of a passenger in an elevator car. In the method, a series of acceleration measurements is evaluated in order to identify the start and end of a journey in the elevator car. However, the method is not suitable for recognizing that a passenger has entered an elevator car of an elevator installation.

In contrast, it is the object of the invention in particular to propose a method by means of which entry into an elevator car by a passenger can be detected with as little additional hardware as possible and thus as cost-effectively as possible. According to the invention, this object is achieved with a method having the features of claim 1.

In the method according to the invention for recognizing that a passenger has entered an elevator car of an elevator installation, it is assumed that the passenger is carrying a mobile terminal device. The terminal has at least one, in particular however, it has several sensors with which the mobile device records and evaluates measured values. The recognition of entering the elevator car is then based on the stated measured values.

“Recognition of a passenger entering an elevator car of an elevator installation” is understood to mean that the point in time when the elevator car was entered is recognized. Entering the elevator car and thus the time of entering precedes a journey by the passenger in an elevator car or a movement and thus an acceleration of the passenger and the elevator car in the vertical direction. From the detection of a movement or acceleration of the passenger and the elevator car in the vertical direction, it is not possible to deduce the time of entering the elevator car. The period of time between entering the elevator car and the beginning of a journey by the passenger in the elevator car can be a few seconds or several minutes.

Nowadays, very many people and thus also many passengers in an elevator system carry a mobile device with sensors, for example in the form of a cell phone or smartphone. By using these terminals, which are carried along anyway, no additional hardware is necessary for carrying out the method, which would only be necessary for carrying out the method. Additional hardware may be necessary if the information generated by the method according to the invention about entering an elevator car is to be further evaluated. The method according to the invention can thus be carried out inexpensively.

The information that a passenger is entering an elevator car with a mobile terminal device can be evaluated in a wide variety of ways or used further, or can trigger a wide variety of actions. The terminal can for example pass on the information wirelessly to a traffic evaluation unit, which is then comparable with the traffic evaluation unit of the WO 2013/130040 A1 can analyze a traffic flow in the elevator system. The terminal can, for example, also be brought into a predetermined mode, that is to say, for example, a specific program, a so-called app, started or the app brought into a predetermined state. For example, an app can be started that displays certain content or a game can be started, which enables interaction with other passengers in the elevator car. In addition, it is possible for the terminal device to record measured variables with its sensors during the upcoming elevator journey, which are to be evaluated for monitoring the elevator system. As soon as entry into an elevator car is recognized, the terminal device can be put into a measuring mode and thus made ready for a measurement.

An exit from an elevator car can also be recognized in an analogous manner. Exiting is basically the opposite of entering an elevator car.

The evaluation of the recorded data and thus the recognition of entering the elevator car is carried out in particular by the mobile terminal. It is also possible, however, for the recorded data to be transmitted to an evaluation device and for the evaluation device to detect entry into the elevator car. In this case, the evaluation of the data by the terminal is limited to forwarding the data to the evaluation device. In addition, it is possible that at least part of the evaluation is carried out both by the mobile terminal and by the evaluation device. This enables mutual monitoring and / or supplementation, which enables a very high hit probability for the detection of entry into an elevator car.

The mobile terminal can be designed, for example, as a mobile phone, a smartphone, a tablet computer, a smart watch, a so-called wearable, for example in the form of an electronic, smart textile, or as another portable terminal. The sensor of the mobile terminal can be designed, for example, as a microphone, an acceleration sensor, a rotation rate sensor, a magnetic field sensor, a camera, a barometer, a brightness sensor, a humidity sensor or a carbon dioxide sensor. The acceleration, yaw rate and magnetic field sensors are designed in particular as so-called three-dimensional or 3D sensors. Such sensors deliver three measured values in the x, y and z directions, the x, y and z directions being arranged perpendicular to one another. The terminal has in particular several and in particular different types of sensors, that is to say for example a microphone, a three-dimensional acceleration sensor, a three-dimensional rotation rate sensor and a three-dimensional magnetic field sensor. In the following, acceleration, yaw rate and magnetic field sensors are understood to mean three-dimensional acceleration, yaw rate and magnetic field sensors.

The passenger can carry the terminal with him in completely different orientations, so that in the first approach it is not clear how the acceleration, yaw rate or magnetic field sensors are oriented in space. However, since the acceleration due to gravity is always measured, the vertical direction, that is to say the absolute z-direction, can be clearly determined from this, at least when the passenger is not moving. With knowledge of the absolute z-direction, the measured values of the acceleration, yaw rate and magnetic field sensors can be converted into values that are aligned along the absolute z-direction and absolute x- and y-directions. The absolute x, y and z directions are each arranged perpendicular to one another. All of the following statements about accelerations, rotation rates or magnetic field strengths relate to measured values converted in this way and statements about x, y and z directions refer to absolute x, y and z directions. Instead of determining the values in the absolute x, y and z directions, the three measured values can be viewed as vectors and a resulting vector can be formed from the individual vectors. Instead of using the three individual measured values, the resulting vector can also be used.

In an embodiment of the invention, the mobile terminal with the sensor or sensors detects measured values that characterize movements of the passenger and evaluates them. The stated measured values are, in particular, accelerations, that is to say transverse accelerations and / or rotation rates, three accelerations and / or rotation rates being measured in the x, y and z directions in particular. From the measured values characterizing the movements of the passenger, conclusions can be drawn about the movements of the passenger, and it can be recognized from the movements of the passenger that the passenger is entering an elevator car. It is generally assumed that the passenger carries the terminal with him in such a way that the measured values measured by the terminal characterize not only the movements of the terminal but also of the passenger.

In an embodiment of the invention, a movement pattern of the passenger is derived from the measured values and compared with at least one stored signal pattern.

The recognition of entering the elevator car then takes place on the basis of the comparison mentioned. Entering an elevator car can thus be recognized particularly reliably.

In this case, the mentioned stored signal patterns are movement patterns. In this context, a movement pattern should be understood to mean, for example, a time sequence in particular of accelerations or rotation rates. A movement pattern can also be described with a so-called feature or, in particular, with a plurality of features. Such features can be, for example, statistical parameters such as mean values, standard deviations, minimum / maximum values or the results of a Fast Fourier analysis of the accelerations or rotation rates mentioned. In this case, a movement pattern can also be referred to as a so-called feature vector. The features mentioned can be determined in particular for individual time segments, with individual measured values being formed in particular based on values or progressions. For example, such a time segment can be characterized in that the passenger is not moving, that is to say, for example, waiting in front of the shaft door. In particular, not just a single acceleration or rate of rotation is considered, but the combination of several accelerations and / or rates of rotation, in particular of three accelerations and rates of rotation.

A stored signal pattern can, for example, be characteristic of accelerations, rotation rates and / or magnetic fields or features when a person walks to a landing door, waiting in front of the landing door until the elevator car is available and access is possible, entering the elevator car and turning around in the direction of the car door contain. The signal patterns can be created by specialists on the basis of their experience or, in particular, can be determined by one or more experiments. In particular, methods of so-called machine learning are used to recognize or classify movement patterns. For example, a so-called support vector machine, a random forest algorithm or a deep learning algorithm can be used. These classification methods must first be trained. For this purpose, in experiments typical movement patterns for entering an elevator car, in particular based on the features mentioned, are generated and the algorithms mentioned are used for training Provided. After the algorithms have been trained with a sufficient number of training patterns, they can decide whether or not an unknown movement pattern indicates entering an elevator car. In this case the signal pattern is stored in the parameters of the algorithm.

The generation of the typical movement patterns for the training can be carried out by a passenger who uses the mobile terminal in daily use. He only has to mark the beginning and the end of entering an elevator car. It is also possible that, after the actual training has been completed, the passenger provides feedback as to whether entering an elevator car was not recognized or whether entering an elevator car was incorrectly recognized. This feedback can be used for further training of the algorithm.
Since not all people move in the same way, e.g. turn around at different speeds, and waiting times vary in length, the measured movement pattern is compared not just with one signal pattern, but with a whole series of slightly different signal patterns.

In an embodiment of the invention, the mobile terminal uses the sensor or sensors to record measured values that characterize activities of the elevator installation and to evaluate them. Activities of the elevator system are to be understood here, for example, as movements of individual components of the elevator system, such as movements of the elevator car, a shaft door, a car door or a control of a door drive. The terminal device detects in particular noises and / or magnetic fields, three magnetic fields in particular being measured in the x, y and z directions. The changes in the measured magnetic fields can be caused, for example, by the activity of the door drive, which has an electric motor, and / or by the car and / or shaft door, which has ferromagnetic material. From the stated measured values it can be concluded, for example, that the car door of an elevator car has opened in front of a passenger and closed behind him.

In an embodiment of the invention, an activity pattern of the elevator installation is derived from the measured values and compared with at least one stored signal pattern. The recognition of entering the elevator car is then based on the mentioned comparison. Entering an elevator car can thus be recognized particularly reliably.

In this case, the mentioned stored signal patterns are activity patterns. In this context, an activity pattern should be understood to mean, for example, a time sequence in particular of measured noises and / or magnetic fields. An activity pattern can also be described with a feature described in connection with movement patterns or, in particular, with a plurality of features. In particular, not only a single measurement of a magnetic field in one direction is considered, but the combination of several measurements of magnetic fields in several, in particular three, directions.

A signal pattern can describe, for example, a noise of a car door when opening or a noise when the elevator car drives into a floor or features derived therefrom. The signal patterns can be created by specialists on the basis of their experience or, in particular, can be determined by one or more experiments. In order to determine the signal patterns, in particular methods of so-called machine learning can be used analogously to the above description in connection with movement patterns. The signal patterns can also be divided into time segments and features can be determined individually for each segment.

Since similar activities of elevators, such as opening the car door, can vary, for example take different lengths of time, the measured activity pattern is compared in particular not only with one signal pattern, but with a whole series of slightly different signal patterns.

In an embodiment of the invention, the mobile terminal with the sensor detects measured values characteristic of the surroundings of the mobile terminal and evaluates them. For example, magnetic fields, air pressure, brightness, humidity or the carbon dioxide content of the air can be measured.

In an embodiment of the invention, a property pattern of the elevator installation is derived from the measured values and compared with at least one stored signal pattern. The recognition of entering the elevator car is then based on the mentioned comparison. Entering an elevator car can thus be recognized particularly reliably.

In this case, the mentioned stored signal patterns are property patterns. In this context, a property pattern should be understood to mean, for example, a chronological sequence of measured values which describe the surroundings of the terminal, that is to say in this case properties of the elevator system. A property pattern can also be described with a feature described in connection with movement patterns or, in particular, with a plurality of features. In particular, not only the course of a single measurement of one of the properties mentioned is considered, but the combination of several measurements.

A signal pattern can describe, for example, the change in the magnetic field from outside to inside the elevator car or features derived therefrom. Changes in the magnetic field can for example be caused by different uses of ferromagnetic materials or different electrical components, such as coils outside and inside the elevator car. The ferromagnetic materials can themselves generate a magnetic field and / or influence the earth's magnetic field.

A signal pattern can, for example, describe the change in the CO2 content of the air from outside to inside the elevator car or features derived therefrom. The CO2 content of the air increases due to the air exhaled by the passengers in the locked elevator car. This means that the CO2 content of the air in the cabin is generally higher than outside. In addition, the CO2 content increases slowly during the journey, which means that a journey in an elevator car can be detected. This increase is a rather slow process, but it can be recognized during longer journeys.

A signal pattern can, for example, describe the change in humidity from outside to inside the elevator car or features derived therefrom. This increases slowly, analogous to the CO2 content inside the cabin, due to the exhaled air, so that the evaluation can proceed analogously to the CO2 content.

A signal pattern can, for example, describe the change in temperature from outside to inside the elevator car or features derived therefrom. Due to the heat given off by the passengers, the temperature rises slowly, so that the evaluation can run analogously to the CO2 content.

A signal pattern can, for example, describe the change in brightness from outside to inside the elevator car or features derived therefrom. It is usually less bright inside an elevator car than outside.

A signal pattern can describe, for example, the change in acoustics from outside to inside the elevator car or features derived therefrom. Since an elevator car is a comparatively narrow, closed space, the echo or the sound attenuation changes, for example. In particular, special test signals can be used to determine this change.

The signal patterns can be created by specialists on the basis of their experience or, in particular, can be determined by one or more experiments. In order to determine the signal patterns, analogous to the above description in connection with movement patterns, in particular methods of so-called machine learning can be used. The signal patterns can also be divided into temporal segments and features can be determined individually for each section.

Since not all elevator systems have identical property patterns, but can vary these, the measured property pattern is compared in particular not only with one signal pattern but with a whole series of slightly different signal patterns.

For the detection of entering an elevator car, not only measured values characterizing individual movements of the passenger, measured values characterizing activities of the elevator system or measured values characterizing properties of the elevator system are recorded and evaluated, but a combination of these different types of measured values. Entering an elevator car can thus be recognized particularly reliably.

In an embodiment of the invention, at least one of the mentioned stored signal patterns is changed, in particular all stored signal patterns are changed. A learning process takes place through which the stored signal patterns are better and better adapted to the actual circumstances. This enables particularly precise detection of a passenger entering an elevator car.

In particular, a journey in an elevator car is recognized from the measured values measured by at least one of the sensors of the mobile terminal. As soon as a journey in an elevator car has been recognized, the movement, activity and / or property patterns recorded prior to the journey are compared with stored signal patterns and the stored signal patterns are adapted on the basis of the comparison. In particular, the stored signal patterns are changed in the direction of the movement, activity and / or property patterns recorded before the journey. In particular, the so-called machine learning methods described above can be used. This enables particularly effective learning and thus also particularly precise detection of a passenger entering an elevator car.

If a journey in an elevator car has been detected, an exit from the elevator car can also be detected with a very high probability of being hit. As soon as the passenger is moving significantly transversely to the vertical direction, that is to say either in the x and / or y direction, it can be assumed that he is leaving the elevator car. This movement can be detected, for example, by means of the acceleration sensor. As an alternative to detecting a movement in the x / y direction, the resultant vector of the accelerations in the x, y and z directions described above can also be used.

A journey in an elevator car has a characteristic course of the acceleration in the vertical direction. The elevator car is first accelerated up or down, then mostly travels for a while at a quasi constant speed and is then braked to a standstill. This acceleration profile can be recognized with high accuracy in the measured values of one or more acceleration sensors of the mobile terminal. In this way, reliable detection of a journey by the passenger and thus by the mobile terminal in an elevator car is possible. Based on this reliable detection, a reliable adaptation of the stored signal pattern is possible, which ultimately leads to a particularly reliable detection of the entry of a passenger into an elevator car.

As an alternative or in addition, the air pressure measured by a barometer can also be evaluated to detect a journey in an elevator car. Driving in the vertical direction results in a change in the air pressure, the gradient of the change being significantly greater in terms of amount than when climbing stairs or with weather-related changes in air pressure.

Further advantages, features and details of the invention emerge from the following description of exemplary embodiments and from the drawings, in which identical or functionally identical elements are provided with identical reference symbols.

Fig. 1

eine sehr schematische Darstellung einer Aufzuganlage mit einem Passagier,

Fig. 2a, b, c

zeitliche Verläufe von Drehraten beim Einsteigen eines Passagiers in eine Aufzugkabine,

Fig. 3a, b, c

zeitliche Verläufe von magnetischen Feldstärken beim Einsteigen eines Passagiers in eine Aufzugkabine, und

Fig. 4

einen zeitlichen Verlauf einer Beschleunigung in vertikaler Richtung bei einer Fahrt einer Aufzugkabine.

Show:

Fig. 1

a very schematic representation of an elevator system with a passenger,

Figures 2a, b, c

Temporal progressions of rotation rates when a passenger boarding an elevator car,

Figures 3a, b, c

Time curves of magnetic field strengths when a passenger boarding an elevator car, and

Fig. 4

a time profile of an acceleration in the vertical direction when an elevator car is traveling.

According to Fig. 1 an elevator system 10 has an elevator car 11 which can be moved up and down in an elevator shaft 12 in the vertical direction 13. For this purpose, the elevator car 11 is connected to a counterweight 16 via a flexible suspension element 14 and a drive roller 15 of a drive (not shown). The drive can move the elevator car 11 and the counterweight 16 up and down in opposite directions via the drive roller 15 and the suspension element 14. The elevator shaft 12 has three shaft openings 17a, 17b, 17c and thus three floors, which are closed with shaft doors 18a, 18b, 18c. In the Fig. 1 the elevator car 11 is located at the shaft opening 17a, that is to say on the lowest floor. When the elevator car 11 is on one floor, that is to say at one of the shaft openings 17a, 17b, 17c, the corresponding shaft door 18a, 18b, 18c can be opened together with a car door 19, thus making it possible to enter the elevator car 11. In order to open the car door 19 and the corresponding shaft door 18a, 18b, 18c, door segments (not shown) are pushed on laterally, so that the door segments are shifted to the side. The car door 19 and the corresponding shaft door 18a, 18b, 18c are operated by a door drive 20 which is controlled by a door control unit 21. The door control unit 21 is in signal connection with an elevator control unit 22, which controls the entire elevator installation 10. The elevator control unit 22 controls, for example, the drive and can thus move the elevator car 11 to a desired floor. It can also, for example, send the door control unit 21 a request to open the car door 19 and the corresponding shaft door 18a, 18b, 18c, which the door control unit 21 then executes by means of a corresponding activation of the door drive 20.

On the lowest floor, that is to say in front of the shaft door 18a, there is a passenger 23 who is carrying a mobile terminal in the form of a cell phone 24 with him. The mobile phone 24 has several sensors, of which only one microphone 25 is shown. The mobile phone 24 also has three-dimensional acceleration, yaw rate and magnetic field sensors, which can detect measured values in the x, y and z directions. As stated above, the measured values recorded by the acceleration, yaw rate and magnetic field sensors can be converted in a simple manner into values relating to absolute x, y and z directions. All of the following statements about accelerations, rotation rates or magnetic field strengths thus relate to measured values converted in this way and statements about x, y and z directions on absolute x, y and z directions.

It should be recognized on the basis of the measured values recorded by the sensors of the mobile phone 24 when the passenger 23 enters the elevator car 11. To this end, the mobile phone 24 continuously records measured values and evaluates them. The mobile phone 24 detects, for example, the rotation rates about the x, y and z axes. These measured rates of rotation characterize not only movements of the mobile phone 24, but also movements of the passenger 23. Measured values are continuously recorded and, by combining the individual measured values of the various acceleration sensors, continuous ones Movement patterns of the passenger 23 generated. The measured values are filtered in particular by means of a low-pass filter. In this case, the named movement pattern thus contains the courses of the rotation rates around the x, y and z axes. The mobile phone 24 compares the continuous movement pattern thus generated with stored signal patterns which are typical of a movement pattern when entering an elevator car 11. In order to be able to carry out the comparison, features in the form of mean values, standard deviations and minimum / maximum values of the individual rotation rates or time segments of the rotation rates are determined and compared with stored values. If the differences between the features of the measured profiles and the stored features are smaller than definable threshold values, then sufficient correspondence between a movement pattern and a stored signal pattern is recognized. The mobile phone 24 concludes from this that the passenger 23 has entered the elevator car 11. The mobile phone 24 can use this information in very different ways. In this example, it is supposed to switch to a measurement mode in which it is ready for measurements during the upcoming journey in the elevator car 11 for monitoring the elevator installation 10. The measurements are only started at a later point in time.

The comparison between a measured movement pattern and a stored signal pattern and thus the detection or classification of movement patterns can also be carried out using methods of so-called machine learning. For example, a so-called support vector machine, a random forest algorithm or a deep learning algorithm can be used.

The transverse accelerations in the x, y and z directions can also be taken into account, so that the movement pattern also contains the progressions of the accelerations in the x, y and z directions.

It is also possible that the mobile phone does not carry out the detection of entry into an elevator car completely on its own, but rather transmits the recorded data to an evaluation device. The detection of entry into the elevator car is then carried out by the evaluation device. As soon as entry is detected, the evaluation device sends a corresponding signal to the mobile phone.

In the Figures 2a, 2b and 2c a measured movement pattern and a stored signal pattern are shown over time, with in Fig. 2a the rotation rates α about the x-axis, in Figure 2b around the y-axis and in Figure 2c is shown around the z-axis. The measured rate of rotation is shown with a solid line and the saved rate of rotation of the signal pattern is shown with a dashed line. The solid lines 26a, 26b, 26c therefore represent the measured rates of rotation and the dashed lines 27a, 27b, 27c the stored rates of rotation about the x, y and z axes. The measured values are shown smoothed.

The stored signal pattern (dashed lines 27a, 27b, 27c) contains typical curves of rotation rates as they occur when entering an elevator car. From the point in time t0 to the point in time t1, the passenger runs towards the shaft door in order to stop at point in time t1 and wait until point in time t2 for the shaft and car door to be opened. There are virtually no rotation rates. From the point in time t2 the passenger enters the elevator car and then turns around in the direction of the car door. This reversal primarily leads to a clear deflection of the rotation rates around the z-axis (line 27c), with a brief undershoot occurring in the opposite direction at the beginning and at the end of the deflection. As in the Figures 2a, 2b and 2c As can be seen, the measured movement pattern (solid lines 26a, 26b, 26c) follows the stored signal pattern very precisely. The comparison of the movement patterns with stored signal patterns proceeds as described above. On the basis of this correspondence, the mobile phone concludes that the passenger has entered the elevator car.

Since not all people move in the same way, e.g. turn around at different speeds, and waiting times vary in length, the measured movement pattern is compared in particular not only with one signal pattern, but with a whole series of slightly different signal patterns.

In addition to the rotation rates, the accelerations in the x, y and z directions can also be taken into account in a comparable manner. This makes it easier to identify walking in the direction of the shaft door and into the elevator car, as well as waiting in front of and in the elevator car.

In order to make the detection of entry into an elevator car more reliable, further measured values recorded by sensors of the cell phone are evaluated in particular. The mobile phone 24 detects the magnetic field strength in the x, y and z directions, in particular with the three-dimensional magnetic field sensor. The measured values thus identify a property of the elevator system. It is very difficult to infer from measured values at a single point in time that the mobile phone and thus the passenger is in an elevator car. For this reason, a property pattern is created from the temporal progressions of the three field strengths, with the measured values being filtered in particular by means of a low-pass filter. The mobile phone 24 compares the continuous property pattern thus generated with stored signal patterns which are typical of a property pattern when entering an elevator car 11. If sufficient correspondence between a movement pattern and a stored signal pattern is detected, the mobile phone 24 concludes from this that the passenger 23 has entered the elevator car 11. The comparison of the movement patterns with stored signal patterns proceeds as described above.

In the Figures 3a, 3b and 3c a measured property pattern and a stored signal pattern are shown over time, where in Fig. 3a the magnetic field strength H in the x-direction, in Figure 3b in y-direction and in Figure 3c are shown in the z-direction. The measured field strengths are each shown with a solid line and the stored field strengths of the signal pattern are each shown with a dashed line. The solid lines 28a, 28b, 28c thus represent the measured field strengths and the dashed lines 29a, 29b, 29c the stored field strengths in the x, y and z directions. The measured values are shown smoothed.

The stored signal pattern (dashed lines 29a, 29b, 29c) contains typical curves of field strengths as they occur when entering an elevator car. Shortly before to shortly after time t2, at which the passenger enters the elevator car, a significant increase can be seen in the field strengths in the y and z directions, while the field strength in the x direction remains virtually unchanged for the entire time. The change in field strengths is due in particular to the use of ferromagnetic materials in the elevator car. As in the Figures 3a, 3b and 3c As can be seen, the measured property pattern (solid lines 28a, 28b, 28c) follows this very precisely stored signal pattern. This correspondence is a further indication for the mobile phone that the passenger has entered the elevator car. The comparison of the property pattern with stored signal patterns runs analogously to the comparison of the movement patterns with stored signal patterns described above.

Since not all elevator systems have identical property patterns, but can vary these, the measured property pattern is compared in particular not only with one signal pattern but with a whole series of slightly different signal patterns.

In addition, additional measured values such as the air pressure, the brightness, the humidity or the carbon dioxide content of the air can be taken into account.

A further increase in the reliability of the recognition of entry into an elevator car can be achieved by additionally taking into account measured values which characterize an activity of the elevator installation. For example, an activity pattern can be derived from the magnetic field strengths described above, which is compared with a signal pattern that is typical for the opening of the car and shaft door. Another possibility is to derive an activity pattern from the noise measured with the microphone and to compare this with a signal pattern that is typical for opening the car and landing door. As with the movement and property patterns, it can be useful to compare the activity patterns with several, slightly different signal patterns. A sufficient correspondence between the measured activity patterns and a stored signal pattern can in turn be assessed as an indication that the passenger has entered an elevator car.

The mobile phone can be designed in such a way that it already recognizes entry into an elevator car if there is a single sufficient match of a movement pattern, a property pattern or an activity pattern with a stored signal pattern. But it is also possible that entry is only recognized when there are at least two, three or more matches.

In order to make the detection of entry into an elevator car more reliable, the stored signal patterns can be adapted. With an adaptation, the method can be adapted in particular to the behavior of the owner of the cell phone. For this purpose, the mobile phone recognizes, in particular, a journey in an elevator car. This can be recognized very reliably by monitoring the acceleration in the z direction and thus in the vertical direction 13. In Fig. 4 a curve of the acceleration a in the z-direction upwards is shown by way of example with the line 30, the acceleration due to gravity not being taken into account. The elevator car 11 and thus also the passenger 23 with his mobile phone 24 are accelerated with an almost constant acceleration from time t4. Shortly before the desired speed of the elevator car 11 is reached, the acceleration drops in order to reach the zero line at time t5. The elevator car 11 then travels at a constant speed up to the point in time t6, in order then to be braked with a quasi constant negative acceleration up to the point in time t7. This typical course with acceleration in the vertical direction, constant travel and braking to a standstill can be seen very well in the measured values.

As soon as a journey in an elevator car has been detected, the movement, activity and / or property patterns recorded prior to the journey are compared with stored signal patterns and, on the basis of the comparison, the stored signal patterns are adapted using machine learning methods. The stored signal patterns are changed in the direction of the movement, activity and / or property patterns recorded before the journey.

Finally, it should be pointed out that terms such as “having”, “comprising”, etc. do not exclude other elements or steps and that terms such as “a” or “an” do not exclude a plurality. Furthermore, it should be pointed out that features or steps that have been described with reference to one of the above exemplary embodiments can also be used in combination with other features or steps of other exemplary embodiments described above. Reference signs in the claims are not to be viewed as a restriction.

Claims (11)

1. Method for detecting a passenger (23) entering an elevator car (11) of an elevator system (10), in which

   - the passenger (23) carries a mobile terminal (24) having at least one sensor (25),

   - the mobile terminal (24) records and evaluates measured values by means of the sensor (25), and

   - a point in time at which the elevator car (11) is entered is detected on the basis of the mentioned measured values
2. Method according to claim 1, characterized in that the mobile terminal (24) records and evaluates measured values which indicate movements of the passenger (23) by means of the sensor (25).
3. Method according to claim 2, characterized in that the mobile terminal (24) records and evaluates accelerations, rotation rates and/or magnetic fields.
4. Method according to either claim 2 or claim 3, characterized in that a movement pattern (26a, 26b, 26c) of the passenger (23) is derived from the measured values and is compared with at least one stored signal pattern (27a, 27b, 27c), and the entry into the elevator car (11) is detected on the basis of said comparison.
5. Method according to any of claims 1 to 4, characterized in that the mobile terminal (24) records and evaluates measured values which indicate activities of the elevator system (10) by means of the sensor (25).
6. Method according to claim 5, characterized in that an activity pattern is derived from the measured values and is compared with at least one stored signal pattern, and the entry into the elevator car (11) is detected on the basis of said comparison.
7. Method according to any of claims 1 to 6, characterized in that the mobile terminal (24) records and evaluates measured values which indicate properties of the surroundings of the mobile terminal (24) by means of the sensor (25).
8. Method according to claim 7, characterized in that a property pattern (28a, 28b, 28c) is derived from the measured values and is compared with at least one stored signal pattern (29a, 29b, 29c), and the entry into the elevator car (11) is detected on the basis of said comparison.
9. Method according to any of claims 5 to 8, characterized in that the mobile terminal (24) records and evaluates noises, magnetic fields, CO2 content of the air, air humidity, temperature, air pressure, brightness and/or noises.
10. Method according to any of claims 4 to 9, characterized in that at least one of said stored signal patterns (27a, 27b, 27c; 29a, 29b, 29c) is changed.
11. Method according to claim 10, characterized in that a journey in an elevator car (11) is detected from the measured values, and measured values recorded before the journey are compared with stored signal patterns (27a, 27b, 27c; 29a, 29b, 29c), and the stored signal patterns (27a, 27b, 27c; 29a, 29b, 29c) can be adjusted on the basis of the comparison.

Images