DE102023109691A1 - Inspection of elevator systems based on several measured variables - Google Patents

Abstract

A method for checking an elevator system comprises at least the following steps: operating the elevator system according to a predefined operating sequence; recording a plurality of measured values for several different measured variables, wherein the measured variables each represent an operating behavior of the elevator system during operation of the elevator system; and determining a test parameter on the basis of the recorded measured values.

Description

The present invention relates to a method for checking elevator systems, in particular rope elevator systems.

Operators of elevator systems are obliged to regularly check operational safety. In particular, it must be checked regularly whether the elevator system meets the specification for which the elevator system is approved. In the case of a rope elevator, for example, it must be ensured that the elevator system can reliably move and hold the maximum transport weight in the elevator car without the support element slipping on the traction sheave, which is critical to safety. For this purpose, the so-called traction capacity is determined, which must be checked regularly in accordance with the relevant regulations, in particular the technical rules for operational safety (TRBS 1201-4).

The traction capability generally represents the slip resistance of one or more support elements, in particular support ropes or belts, on the traction sheave. It can be determined using known methods by first loading the car with a nominal load plus an additional load for a safety margin. The loaded car is then moved by the traction sheave in various operating situations, preferably in an upward direction, and strongly decelerated in order to determine whether or not the support elements slip.

In practice, the described test procedure is very complex to carry out, as defined test loads have to be loaded into the elevator and then removed again after the test has been completed and, if necessary, taken to the next test location. The test also requires several people who need special training, which increases the effort required for the test accordingly.

Against the background described, the object of the present invention is to provide a method of the type mentioned at the outset and a testing device for an elevator system, which allows a simpler and more reliable testing of the driving ability of elevator systems.

The object is achieved according to a first aspect by a method having the features of claim 1. The method comprises at least the following steps: operating the elevator system according to a predefined operating sequence; recording a plurality of measured values for several different measured variables, wherein the measured variables each represent an operating behavior of the elevator system during operation of the elevator system; and determining a test parameter on the basis of the recorded measured values.

The method is based on various measured variables that together contain the necessary information for a simple yet reliable inspection of an elevator system. The measured variables can each have a traceable connection to a specific safety aspect, e.g. the traction capability. However, there can also be complex dependencies between the measured values that can be exploited to detect errors and safety risks in the elevator system, in particular using machine learning methods such as neural networks and the like.

The test parameter generally represents a specific safety aspect of the elevator system, in particular the traction capability, as described in more detail below.

Embodiments are disclosed in the dependent claims, the description and the figures.

According to one embodiment, the method includes determining whether the test parameter violates a predetermined test criterion. This step can be carried out by a person trained to check elevator systems, who evaluates the test parameter based on prior knowledge. However, the step, like determining the test parameter, is preferably carried out as a computer-implemented method step. For example, it can be determined with the aid of a computer whether the test parameter exceeds or falls below one or more safety-relevant threshold values.

According to one embodiment, the elevator system is formed by a cable elevator system. In this case, a first load element, such as the elevator car, is generally connected via a support means to a second load element, in particular a counterweight of the elevator system. During operation, the support means, which can generally be formed by one or more ropes or belts, is deflected and driven between the first load element and the second load element by means of a rotatably mounted drive pulley in order to move the first load element and the second load element simultaneously.

The test parameter determined in the case of rope lift systems can represent the traction capability in particular. As described above, this is of particular importance for the Operational safety and requires regular inspection.

According to a further embodiment, a first subset of the measured values is recorded by at least one sensor permanently installed on the elevator system. For example, measured values that are recorded anyway by an internal sensor system for the correct control of the elevator system can be read out via a communication interface and used to check the elevator system.

Extensive measurement data is particularly advantageous when using machine learning models, such as neural networks or regression models. Such models can benefit from large amounts of data even if individual data components have little or no influence on the test parameter to be determined.

According to a further embodiment, at least a second subset of the measured values is recorded by a mobile testing device arranged on the elevator system. This can be arranged flexibly on the elevator system for this purpose. It also has the necessary sensors, e.g. a distance sensor, a microphone and/or an acceleration sensor.

Preferably, at least a subset of the measured variables relates to a movable element of the elevator system that is moved during operation of the elevator system. For example, one or more measured variables relate to technical properties of the elevator car and/or a counterweight of the elevator system. The measured variables can also relate to other movable elements, e.g. a support means, a speed limiter or parts thereof or a shaft copy of the elevator system.

The measured variables can represent at least one of the following measured variables: distance or position of the movable element relative to a fixed point; speed or acceleration of the movable element; mass of the movable element; acoustic sound emitted by the elevator system due to the movement of the element; and/or an electrical operating variable of an electric motor of the elevator system. The electrical operating variable can in particular be formed by an electrical current or an electrical voltage that is applied to a drive motor of the elevator system.

According to a further embodiment, the measured values for at least one of the measured variables are recorded using different sensor types. Advantageously, a single measured variable can be measured redundantly using different sensor types in order to compensate for measurement inaccuracies of the individual sensor types. It has proven to be particularly advantageous to determine the distance of the movable element using both a sensor of a mobile testing device and an internal distance sensor of the support means. In addition, it is possible to record the distance and/or acceleration of the movable element using several sensors of the mobile testing device. For this purpose, the testing device can have, for example, a LIDAR module and an ultrasonic sensor.

In a further embodiment, the measured values for determining the test parameter are evaluated depending on reference measured values. These can be based, for example, on a learning operation of the elevator system. In addition or alternatively, reference measured values can be included on the basis of predetermined specification data of the elevator system. The specification data includes, for example, data provided by a manufacturer of the elevator system or a component of the elevator system (e.g. by a sensor manufacturer). The data contains information about an expected and reliable behavior of the underlying component. A comparison of the current measured data with the expected reference data can therefore provide information about the quality of the measured data and safety risks. The data comparison can therefore be used advantageously for a reliable inspection of the elevator system.

The measured values can be compared with the reference values in various ways. Preferably, a predetermined mathematical relationship is formed between the current measured values and the reference values, for example a correlation function (e.g. in the form of a cross-correlation or convolution function). Additionally or alternatively, a statistical machine learning model can be used, which also forms mathematical relationships between the measured values, e.g. by applying kernel functions. However, simpler function types can also be used, e.g. a large number of threshold comparisons in a neural network or a regression model.

As mentioned at the beginning, the elevator system is operated according to a predefined operating sequence while the measured values are being recorded. On the one hand, this can be specified by a regular operating sequence programmed into the elevator system, for example an elevator ride between two levels. In addition or alternatively, a safety operating sequence can be provided. in particular an emergency stop in which the car is braked with maximum braking acceleration in the downward direction (travel towards the ground).

According to a further embodiment, the measured values are processed depending on predetermined tolerance ranges. For example, a subset of the measured values can be excluded from the determination of the test parameter or modified according to a predetermined filter function if the measured values of the subset violate an assigned tolerance range, i.e. exceed or fall below corresponding threshold values. The reliability of the method can be significantly increased in this way; on the one hand because outliers in the measured values cannot influence the test result; on the other hand because the test model does not have to be designed to compensate for outliers. Measured values contaminated with noise or other measurement uncertainties can therefore also be used to check the elevator system.

According to a further embodiment, the method can be carried out at least partially on a central server which receives the measured values via a data network and determines the test parameter.

In a further embodiment, the method is at least partially embodied by a computer program in the form of instructions which, when the program is executed by a computer, cause the computer to carry out the method according to one of the described embodiments. In particular, the test parameter can be determined by a computer program.

The computer program is preferably stored as application software, in particular as an app on a mobile testing device or sensor unit that can be used to implement the method. For this purpose, the testing device can be equipped with a non-volatile memory and a processor unit for interpreting and executing the commands stored in the memory.

A further aspect of the invention relates to a mobile testing device for checking an elevator system. The testing device is designed in accordance with the method aspects described above to record a large number of measured values for several different measured variables. The measured values can be recorded via sensors permanently mounted on the elevator system and/or via sensors built into the testing device. The testing device preferably also comprises an evaluation unit for determining a test parameter on the basis of the recorded measured values. One or more of the method aspects described above can be implemented here.

• 1 eine schematische Ansicht von Teilen einer Aufzugsanlage;
• 2 ein schematisches Blockschaltbild von Aspekten eines Verfahrens zur Überprüfung einer Aufzugsanlage;
• 3 ein Betriebsablaufdiagramm; und
• 4 ein schematisches Blockschaltbild von weiteren Aspekten eines Verfahrens zur Überprüfung einer Aufzugsanlage.

The described aspects of the invention are described below, purely by way of example, with reference to the drawings, in which:

• 1 a schematic view of parts of an elevator system;
• 2 a schematic block diagram of aspects of a method for checking an elevator installation;
• 3 an operational flow diagram; and
• 4 a schematic block diagram of further aspects of a method for checking an elevator system.

1 shows schematically sketched components of a cable lift system 10 comprising a car 12 designed as a passenger cabin, which is connected to a counterweight 14 via one or more support means 16. The support means 16 can be designed as a rope or belt, for example, and is guided over a rotatably mounted traction sheave 18 and a deflection roller 20. By driving the traction sheave 18 by means of an electric motor (not shown), the car 12 can be moved either vertically up or down in order to transport people or other payloads within an elevator shaft between predefined stopping positions, as is generally known in elevator systems.

Aspects of a procedure for safety checking of the elevator system 10 are described below.

First, the traction sheave 18 is driven to move the elevator car 12 up or down in a vertical direction within an elevator shaft 31. Preferably, the empty elevator car 12 is moved upwards and decelerated to a standstill by activating the emergency stop with maximum braking acceleration.

During the movement of the elevator car 12, a large number of measured values are measured for various measured variables that are related to the operation of the elevator system 10. One of the measured variables represents a distance 30 between the elevator car 12 and a ceiling section 33 of the elevator shaft 31 (see dashed line in 1 ). The distance 30 is measured at regular time intervals to determine the speed and/or acceleration of the car 12. In addition, the distance 30 serves as a measure for the Length of the support section 22 which extends between the traction sheave 18 and the car 12.

A mobile test device 24 is used to measure some of the measured values. This is arranged on the roof of the elevator car 12 and has various sensors (not shown). On the one hand, a distance sensor is provided which emits a laser beam to the ceiling section 33 and calculates the distance 30 by measuring the travel time until the reflected laser beam is received. On the other hand, the test device 24 is equipped with a microphone which measures the acoustic emissions of the elevator system 10. In particular, acoustic vibrations can be measured which arise on the support element section 22 and provide conclusions about the driving ability of the traction sheave 18.

In addition to the above-mentioned measured values, which are recorded directly by the test device 24, further measured values are measured. These are measured by means of sensors permanently installed on the elevator system 10 and transmitted by a control device of the elevator system 10 to the test device 24 via a wireless communication interface (in 1 not shown). This includes the current mass of the car 12 as well as other constant technical properties of the elevator system 10, in particular the mass of the counterweight 14, the total length and the mass of the support means 16.

The measured values are processed in the test device 24 to obtain a reliable estimate of the driving ability of the traction sheave 18. This will be explained below with reference to 2 and 4 described in more detail.

Firstly, referring to 2 a method for determining a test parameter is described which represents the traction capability. On the one hand, measured values 32 are used for this purpose, which are recorded directly by the test device 24. On the other hand, measured values 34 are used which are read out by the elevator system 10 as described above.

The measurement value groups 32 and 34 are first subjected to preprocessing 36 in which the measurement values are compared with predefined threshold values and a portion of the measurement values that violate one or more threshold values are excluded from further processing. Furthermore, the measurement values 32, 34 are standardized.

In the following step 38, the measured values 32, 34 are processed with a machine-learned model that outputs the test parameter 40. The model was previously determined on the basis of a large number of training data that represent a desired operating behavior based on reference measured values. Alternatively or additionally, the model can have a fixed calculation rule for the driving ability, which can be specified as follows: $$\frac{F1}{F2}=\frac{\text{G}\left(\text{G}+\ddot{\text{x}}\right)}{\text{F}\left(\text{G}-\ddot{\text{x}}\right)}$$

Here, x denotes the time-dependent distance covered by the car 12 or the suspension element 16 and the counterweight 14 during the measurement process. Accordingly, ẍ denotes twice the time derivative of the distance, which corresponds to the acceleration of the car 12. The variable G denotes the mass of the counterweight 14, the variable F the mass of the car 12 and g the acceleration due to gravity. The test parameter is determined as a function of these variables as a ratio between F1 and F2, where F1 denotes the suspension element force on the side of the counterweight 14 and F2 denotes the suspension element force on the side of the car 12. Optionally, the proportional suspension element masses can be added to the car weight or the counterweight depending on the distance position. The proportional suspension element masses can be determined, for example, by multiplying the proportional suspension element lengths by the predetermined suspension element mass per unit length.

In general, it is assumed that the accuracy of the test parameter increases with the number of measured variables that are included in the calculation. In order to achieve a certain level of significance for the test parameter, conditions can be set for the number of measured variables and the quality of the measured values recorded for each of them. For example, it can be required that a deviation of < x ₁ % is maintained for two measured variables, a deviation of < x ₂ % for three measured variables, and a deviation of < x _N % for N measured variables, where x _N represents the deviation based on the measurement uncertainty of the individual measured variables.

It is also preferable to measure a respective measurement variable using several sensors. In this way, deviations from individual sensors can be detected and compensated for when evaluating the corresponding measured values. This means that the elevator system can be checked with greater reliability and accuracy.

With reference to 4 Further aspects of the procedure of 2 described, which can be optionally implemented. In step 56, current measured values are recorded, in particular the measured values recorded by a sensor unit or test device 54 arranged on the elevator car 12 during operation of the elevator system 10. The current measured values do not necessarily have to be recorded by known sensors, but can also be recorded by unknown sensors of the elevator system. The measured values recorded in step 56 can therefore in particular be the values recorded in connection with 2 described measured values 32 and/or 34.

In step 58, reference measurement values are recorded that were determined as part of a learning operation of the elevator system 10. In step 60, manufacturer measurement values are used that were determined in advance independently of the specific elevator system 10. The measurement values recorded in steps 58 and 60 are then combined in step 62 to form a reference signal. The reference signal serves as a signal-based reference for an intrusive check of the elevator system 10.

In step 64, a relation signal is formed on the basis of the current measured values recorded in step 56 and the reference signal formed in step 62, which relates the current measured values and the reference signal to one another and represents agreements or deviations between the data groups involved. For example, the relation signal can be formed as a correlation signal.

In step 66, the relation signal is evaluated and examined for correspondences between the reference signal and the current measured values. A machine learning model is preferably used for this purpose, preferably a neural network or comparable models. In step 68, the evaluation result obtained in step 66 is evaluated. In particular, the 2 The test parameters 40 described in this document are assessed in order to draw conclusions about the operational safety of the lift system 10. All of the 2 and 4 The data processing steps described can be carried out by an evaluation unit of the test device 24 or alternatively on a central server. The test parameter 40 can thus be determined particularly quickly and at the same time very reliably.

With reference to 3 An exemplary operating flow diagram 42 for the elevator system 10 is described below. First, it is assumed that the elevator system 10 is at a standstill and accordingly none of the movable elements are moving.

In diagram 42, the speed v of the elevator car 12 is plotted against time t. At the beginning, the speed v of the elevator car 12 is zero and the time when it is ready to depart is marked by reference number 44. An elevator ride is then initiated by entering an operating command. To do this, the doors of the elevator car 12 are first closed over time 46. The elevator car 12 is then accelerated to a constant travel speed greater than zero and braked to a standstill before reaching the desired level (see curve 48). The doors are then opened again over time 50. This results in a total travel time which is designated by reference number 52.

The exemplary operating procedure of 3 is programmed into the elevator system 10 with respect to the underlying control commands and can be represented by measured values and other data which are obtained in the process steps 58 and 60 of 4 taken into account. If the programmed operating procedures are changed, the data used to check the elevator system 10 can be adjusted accordingly. This means that the inspection procedure can be adjusted without any problems if the operating procedures are modified. It is therefore not necessary to develop the inspection procedure from scratch.

The data used in steps 58 and 60 are preferably stored on a central server (“cloud”) so that they are stored in a secure location and can be quickly available and centrally maintained, for example in the event that the test procedure needs to be adapted.

LIST OF REFERENCE SYMBOLS

10

Elevator system

12

Car

14

Counterweight

16

Carrying means

18

Traction sheave

20

Deflection pulley

22

Suspension section

24

Test device

30

Distance

31

Elevator shaft

32

Measurement values of the sensor unit

34

Measurement values of the elevator system

36

Filtering the measured values

38

Determining the test parameter

40

Test parameters

42

Operational flow chart

44

Ready to depart

46

close the door

48

Elevator ride

50

open doors

52

Total travel time

54

Test device

56

Recording of current measured values

58

Recording of reference measurements

60

Entering manufacturer measurements

62

Creating a reference signal

64

Forming a relation signal

66

Evaluating the relation signal

68

Evaluating the evaluation results

v

speed

t

Time

Claims (13)

Method for checking an elevator system, in particular a cable elevator system, comprising: - operating the elevator system (10) according to a predefined operating sequence; - recording a plurality of measured values (32, 34, 56, 58, 60) for several different measured variables, wherein the measured variables each represent an operating behavior of the elevator system (10) during operation of the elevator system (10); and - determining a test parameter on the basis of the recorded measured values (32, 34, 56, 58, 60). Procedure according to Claim 1 , wherein a first load element (12) is connected to a second load element (14) of the elevator installation (10) via at least one support means (16), and wherein the support means (16) is deflected and driven between the first load element (12) and the second load element (14) by means of a rotatably mounted drive pulley (18) in order to move the first load element (12) and the second load element (14), and wherein the test parameter represents the driving ability of the elevator installation (10). Procedure according to Claim 1 or 2 , wherein the measured values or a first subset (34) of the measured values are recorded by at least one sensor permanently installed on the elevator installation (10). Method according to at least one of the preceding claims, wherein the measured values or a second subset (32) of the measured values are recorded by at least one mobile testing device (24) arranged on the elevator installation (10). Method according to at least one of the preceding claims, wherein at least a subset of the measured variables relates to a movable element (12) of the elevator system (10) which is moved during operation of the elevator system (10), and wherein the measured variables of the subset represent at least one of the following measured variables: - distance (30), in particular position; - speed (v); - acceleration; - mass; - acoustic sound. Method according to at least one of the preceding claims, wherein the measured values (32, 34) for at least one of the measured variables are recorded by means of different sensor types. Method according to at least one of the preceding claims, wherein the measured values for determining the test parameter (40) are evaluated as a function of reference measured values (58, 60), and wherein the reference measured values (58, 60) are based on a learning operation of the elevator installation (10) and/or have predetermined specification data. Method according to at least one of the preceding claims, wherein the measured values for determining the test parameter (40) are evaluated by means of a function based on a machine learning model (66). Method according to at least one of the preceding claims, wherein the predefined operating sequence comprises a regular operating sequence determined by input on the elevator system (10) and/or a safety operating sequence, in particular an emergency stop. Method according to at least one of the preceding claims, wherein the measured values (32, 34, 56) are processed depending on predetermined tolerance ranges, in particular wherein a subset of the measured values (32, 34, 56) are excluded from the determination of the test parameter (40) or are changed according to a predetermined filter function (38) if the measured values (32, 34, 56) of the subset violate an associated tolerance range. Method according to at least one of the preceding claims, wherein the method is carried out at least partially on a central server which receives the measured values via a data network and determines the test parameter (40). Computer program comprising instructions which, when the program is executed by a computer, cause the computer to carry out at least one step of the method according to one of the preceding claims. Mobile testing device for checking an elevator system, wherein the testing device (24) is designed to record a plurality of measured values for several different measured variables, wherein the measured variables each represent an operating behavior of the elevator system (10) while the elevator system (10) is operated according to a predefined operating sequence, and wherein the testing device (24) has an evaluation unit for determining a test parameter (40) on the basis of the recorded measured values.

Images