DE102022120560A1 - METHOD FOR CALIBRATION OF A TORQUE SENSOR SYSTEM FOR A TRANSMISSION - Google Patents

Abstract

The present invention relates to a method for calibrating a torque sensor system for a transmission (1). The transmission (1) can have an elastic transmission element (2) and a sensor (3) of the torque sensor system can also be arranged on the elastic transmission element (2). Furthermore, the method can include detecting at least one measurement signal (S1, S2) using the sensor (3) of the torque sensor system. Furthermore, in a further method step, a torque (B) can be determined from the at least one measurement signal (S1, S2) using a mathematical model. In further method steps, a total shaft rotation (T1) can be determined from the at least one measurement signal (S1, S2) and a total load torque (T2) from the torque (B). The mathematical model can be calibrated based on the total shaft rotation (T1) and the total load torque (T2). The present invention further relates to a transmission system, comprising a transmission (1), an elastic transmission element (2) and a torque sensor system with a sensor (3) and a mathematical model.

Description

TECHNICAL FIELD

The present invention relates generally to the technical field of methods for calibrating a torque sensor system for a transmission.

BACKGROUND

Torque sensor systems have a wide range of applications in drive technology. For example, to improve a transmission in robotics, the torque is measured or determined computationally using a torque sensor system. Torque sensor systems can help ensure that high torque can be provided using a transmission. Particularly in the case of industrial robots and multi-link robot joint systems, a high torque of a gearbox is often necessary in order to ensure optimal movement of the multi-link robot joint system for the application. In addition to robotics, torque sensor systems are also used in vehicle technology.

• Hashimoto, Minoru, Yoshihide Kiyosawa, and Richard P. Paul. „A torque sensing technique for robots with harmonic drives.“ IEEE Transactions on Robotics and Automation 9.1 (1993): 108-116 .
• Hashimoto, Minoru, and Yoshihide Kiyosawa. „Experimental study on torque control using Harmonic Drive built-in torque sensors.“ Journal of Robotic Systems 15.8 (1998): 435-445.

To determine the torque, a sensor of the torque sensor system is usually connected to a robot joint controller. Such a sensor is generally formed by attaching strain gauges to a component of the transmission of the robot joint system and determining a torque based on input data and parameters, the input data being detected by the strain gauges. Such procedures are described, for example, in the following scientific publications:

• Hashimoto, Minoru, Yoshihide Kiyosawa, and Richard P. Paul. “A torque sensing technique for robots with harmonic drives.” IEEE Transactions on Robotics and Automation 9.1 (1993): 108-116 .
• Hashimoto, Minoru, and Yoshihide Kiyosawa. “Experimental study on torque control using Harmonic Drive built-in torque sensors.” Journal of Robotic Systems 15.8 (1998): 435-445.

However, due to metal fatigue and degradation of transmission components, the physical properties of the transmission change. This effect is particularly pronounced in harmonic drive transmissions, where components such as the “Flex Spline” are repeatedly deformed during their service life.

This change results in a deterioration in the determination of torque by the torque sensor system. The service life of the transmission is also inaccurately estimated using the torque sensor system. After a certain period of time, the original parameters used to determine the torque are no longer valid.

It is therefore an object of the present invention to provide a method for calibrating a torque sensor system of the type mentioned at the outset. Another object of the present invention is to provide an improved torque sensor system.

SUMMARY

These tasks are solved by the subject matter of the independent claims. In its most general form, the invention relates to a method for calibrating a torque sensor system for a transmission. The transmission can have an elastic transmission element and a sensor of the torque sensor system can also be arranged on the elastic transmission element. Furthermore, the method can include detecting at least one measurement signal using the sensor of the torque sensor system. Furthermore, in a further method step, a torque can be determined from the at least one measurement signal using a mathematical model. In further method steps, a total shaft rotation can be determined from the at least one measurement signal and a total load torque from the torque. The mathematical model can be calibrated based on the total shaft rotation and total load torque.

Calibration typically means adapting the torque sensor system to changes in the transmission. The torque sensor system can include a mathematical model or at least components of the mathematical model. The mathematical model can also be the changes to the transmission can be adapted. For this purpose, the mathematical model can include calibration parameters. The mathematical model can be calibrated or adjusted by changing the calibration parameters of the mathematical model. The calibration parameters of the mathematical model can be changed in such a way that the calibration parameters match the physical behavior of the transmission. This means that the torque can also be precisely determined using the mathematical model.

Changes to the transmission can occur, for example, due to material fatigue of the components of the transmission. Material fatigue can be caused by changing mechanical stress, corrosion, changing temperature, UV radiation and/or ionizing radiation.

The transmission is preferably part of a vehicle or robot. The transmission can be connected to a motor. The transmission typically has a drive shaft connected to the motor. The engine can be an internal combustion engine, an electric motor, a fuel cell, a hybrid drive or another type of energy converter.

Helical gears, bevel gears, helical gears, planetary gears or harmonic drive gears can be used as gears.

The elastic transmission element can, for example, comprise a “flex spline” of a harmonic drive transmission. The elastic transmission element can deform with every rotation of the drive shaft. The elastic transmission element therefore typically experiences changing mechanical loads, which can lead to material fatigue.

The torque sensor system may include a variety of sensors. The torque sensor system can, for example, include a torque sensor in the form of a strain gauge. A sensor can also include a variety of strain gauges.

The measurement signal detected by the sensor of the torque sensor system can be, for example, an electrical voltage. A torque can be calculated using the electrical voltage. The measurement signal typically changes with rotation of the drive shaft, with, for example, a high voltage being detected at a drive shaft position of 0° and a low voltage at a drive shaft position of 90°.

The total shaft rotation can be the sum of the shaft rotations of the drive shaft. A shaft rotation can be a rotation of the drive shaft through 360°. The total shaft rotation can be used to determine the total number of shaft rotations since the start of detection by the torque sensor system. In other words, the total number of shaft rotations since the transmission was put into operation can be recorded using the total shaft rotation.

The total load torque can be a sum of torques. Torques can vary during shaft rotations. Torque can mean the torque applied to the elastic transmission element. In other words, the total load torque can be used to record the torques since the transmission was put into operation.

The dynamic fatigue of the elastic transmission element can be detected using the total shaft rotation and the total load moment.

The measurement signals detected by the sensor of the torque sensor system can be transmitted from the sensor to the torque sensor system using a communication line. The communication line can be, for example, an EtherCAT, RS485 or SPI system.

The torque can be understood as the torque available to drive the vehicle, the robot or a manipulator member of the robot. The torque of the transmission can therefore be, for example, the output torque of the transmission. The torque from the gearbox can be transmitted directly to the manupulator links of the robot or to the wheels of the vehicle. Torque can also be understood as the torque applied to the drive shaft of the transmission.

The person skilled in the art typically understands determining the torque from the measurement signal to mean any type of determination and/or calculation in the mathematical sense. Any type of mathematical operation can be used for this purpose, especially iterative procedures. Reading from tables can also be used to determine torque.

The torque sensor system may include a microcontroller. Some or all of the method steps can be implemented on the microcontroller of the torque sensor system. The microcontroller can be integrated in the torque sensor system. An advantage of this implementation is that the method steps can be carried out in their entirety using the torque sensor system and no further components are necessary.

Some or all of the method steps may be performed by (or using) a hardware device, such as a dedicated microcontroller. It is also possible for the hardware device to include a processor, a microprocessor, a programmable computer, or an electronic circuit. In some embodiments, one or more of the key method steps may be performed by such a device.

Of course, the process can be fully automated. It is also possible for the process to be carried out continuously in the real time that is usual for robotics or motor technology.

The method according to the invention advantageously leads to an improvement in the calculation of the torque. Furthermore, the process determines the service life of the gearbox more precisely. The method can thus contribute to the transmission having an improved, simpler structure. At the same time, the process can achieve lower production costs.

In a further aspect of the invention, the sensor of the torque sensor system comprises a first strain gauge and a second strain gauge. A first measurement signal can be detected using the first strain gauge and a second measurement signal can be detected using the second strain gauge.

Strain gauges can be made of a semiconductor and can detect a voltage across a resistor based on the piezoresistive effect. A change in resistance and tension can occur due to a deformation of the strain gauge. The advantage of the strain gauge used compared to conventional metal strain gauges is its high sensitivity.

In a further aspect of the invention, the first strain gauge and the second strain gauge are offset by 90° in a rotation direction.

By offsetting the strain gauges by 90°, the measurement signals recorded by the strain gauges can also have a phase shift of 90°.

In a further aspect of the invention, determining the total shaft rotation and determining the total load torque comprises, as a method step, determining a difference signal from the first measurement signal and the second measurement signal. Furthermore, a change in the difference signal to a previous state can be determined using predefined sections. If a change in the difference signal has been determined, the total shaft rotation and the total load torque can be determined.

In a further aspect of the invention, the mathematical model is calibrated using a predefined calibration function and based on the total shaft rotation and the total load torque.

The mathematical model can include calibration parameters that can be calculated using the calibration function. The calibration function may have been determined and predefined during the development phase using a lifespan test.

In a further aspect of the invention, a sensor system is provided that includes means for carrying out one of the methods described herein.

The sensor system can include a strain gauge. Further, the sensor system may include a microcontroller and/or a storage medium (or a data carrier or a computer-readable medium) that includes a computer program stored thereon for executing one of the methods described herein when executed by a processor. The data carrier, digital storage medium or recorded medium is usually tangible and/or non-seamless. Another embodiment of the present invention is a sensor system as described herein that includes a processor and the storage medium.

In a further aspect of the invention, a computer program is provided that includes instructions that, when executed by a computer, cause the program to carry out one of the methods described herein.

According to the invention, a transmission system is also provided. The transmission system may include a gearbox, an elastic transmission element and a torque sensor system with a sensor and a mathematical model. The sensor of the torque sensor system can be arranged on the elastic transmission element. Furthermore, the torque sensor system can be designed such that a torque can be determined using the mathematical model and the mathematical model can be calibrated using at least one measurement signal detected by the sensor.

The transmission according to the invention can advantageously have a particularly space-saving and compact design. The transmission can also advantageously have a small number of components. The calibration also makes the gearbox particularly flexible to external influences.

In a further aspect of the invention, the sensor of the torque sensor system comprises a first strain gauge and a second strain gauge.

In a further aspect of the invention, the mathematical model is designed to be calibratable using a calibration function.

In principle, it is possible that the calibration function includes a regression model and/or a multi-layer perceptron model.

Artificial neural networks can be used as mathematical models for the methods described herein. Furthermore, the calibration function can also include an artificial neural network.

Artificial neural networks (ANN) are systems inspired by biological neural networks, such as those found in a retina or brain. ANNs include a plurality of interconnected nodes and a plurality of connections, called edges, between the nodes. There are usually three types of nodes, input nodes that receive input values, hidden nodes that are (only) connected to other nodes, and output nodes that provide output values. Each node can represent an artificial neuron. Each edge can send information from one node to another. The output of a node can be defined as a (nonlinear) function of the inputs (e.g. the sum of its inputs). A node's inputs can be used in the function based on a "weight" of the edge or node providing the input. The weight of nodes and/or edges can be adjusted in a learning process. In other words, training an artificial neural network may include adjusting the weights of the nodes and/or edges of the artificial neural network, i.e. to achieve a desired output for a particular input.

Alternatively, the mathematical models and the calibration function can be a support vector machine, a random forest model or a gradient boosting model. Support vector machines are supervised learning models with associated learning algorithms that can be used to analyze data (e.g. in a classification or regression analysis). Support vector machines can be trained by providing input with a plurality of training input values belonging to one of two categories. The support vector machine can be trained to assign a new input value to either category. Alternatively, the mathematical models may be a Bayesian network, which is a probabilistic directed acyclic graphical model. A Bayesian network can contain a set of random variables and their conditional dependencies represent using a directed acyclic graph. Alternatively, the mathematical models may be based on a genetic algorithm, which is a search algorithm and heuristic technique that mimics the process of natural selection.

The mathematical models and/or the calibration function can be trained using a training data set or can be trained in advance. The training data set can, for example, be created on a test bench.

For example, “supervised learning” can be used as a training method for the mathematical models and/or the calibration function. In supervised learning, the mathematical models and/or the calibration function are trained using a plurality of training samples, where each sample can include a plurality of input data values and a plurality of desired output values, i.e. a desired output value is assigned to each training sample. By specifying both training samples and desired output values, the machine learning model “learns” what output value to provide based on an input sample that is similar to the samples provided during training. In addition to supervised learning, semi-supervised learning can also be used. In semi-supervised learning, some of the training samples lack a desired output value. Supervised learning can be based on a supervised learning algorithm (e.g. a classification algorithm, a regression algorithm or a similarity learning algorithm). Classification algorithms can be used when the outputs are restricted to a limited set of values (categorical variables), i.e. the input is classified as one of the limited set of values. Regression algorithms can be used when the outputs show any numerical value (within a range). Similarity learning algorithms can be similar to both classification and regression algorithms, but are based on learning from examples using a similarity function that measures how similar or related two objects are. In addition to supervised learning or semi-supervised learning, unsupervised learning can be used to train the machine learning model. In unsupervised learning, input data may be provided (only) and an unsupervised learning algorithm may be used to find structure in the input data (e.g. by grouping or clustering the input data, finding commonalities in the data). Clustering is the assignment of input data comprising a plurality of input values into subsets (clusters) such that input values within the same cluster are similar according to one or more (predefined) similarity criteria, while they are dissimilar to input values comprised in other clusters.

BRIEF DESCRIPTION OF THE DRAWINGS

• 1A: Getriebe 1 gemäß Ausführungsformen der Erfindung.
• 1B: Messsignale S1 und S2 gemäß Ausführungsformen der Erfindung.
• 2A: Differenzsignal gemäß Ausführungsformen der Erfindung.
• 2B: Bestimmen von Abschnitten M des Differenzsignals D gemäß einer Ausführungsform der Erfindung.
• 3A: Bestimmen von diskreten Werten des Differenzsignals D in den vordefinierten Abschnitten gemäß einer Ausführungsform der Erfindung.
• 3B: Feinjustierung der Abschnitte M des Differenzsignals D gemäß einer Ausführungsform der Erfindung.
• 3C: Feinjustierung in Unterabschnitte des Differenzsignals D gemäß der in 3B dargestellten Ausführungsform.
• 4: Verfahren gemäß Ausführungsformen der Erfindung im Überblick.
• 5: Verfahren gemäß einer weiteren Ausführungsform der Erfindung im Überblick.

Preferred embodiments of the present disclosure are described below with reference to the following figures:

• 1A : Transmission 1 according to embodiments of the invention.
• 1B : Measurement signals S1 and S2 according to embodiments of the invention.
• 2A : Differential signal according to embodiments of the invention.
• 2 B : Determining sections M of the difference signal D according to an embodiment of the invention.
• 3A : Determining discrete values of the difference signal D in the predefined sections according to an embodiment of the invention.
• 3B : Fine adjustment of the sections M of the difference signal D according to an embodiment of the invention.
• 3C : Fine adjustment into subsections of the difference signal D according to in 3B illustrated embodiment.
• 4 : Overview of methods according to embodiments of the invention.
• 5 : Method according to a further embodiment of the invention at a glance.

DESCRIPTION OF PREFERRED EMBODIMENTS

Embodiments of the invention are described below, in which a method for calibrating a torque sensor system for a transmission is used in order to make the torque sensor system and the transmission particularly efficient.

1A shows a transmission 1 according to an embodiment of the invention. In the example shown, the transmission 1 is used in an industrial robot and includes an elastic transmission element 2 and a sensor 3 of a torque sensor system integrated in the transmission 1. The sensor includes a first strain gauge DS1 and a second strain gauge DS2.

Gearbox 1 is a harmonic drive gearbox that is characterized by a high gear ratio and high rigidity. The elastic transmission element 2 is also referred to as a “flex spline”. The transmission 1 also has a drive shaft and an output shaft. The torque of the output shaft is passed on to a manipulator on the industrial robot.

The strain gauges each consist of a semiconductor and detect a voltage across a resistor based on the piezoresistive effect. A change in resistance and tension occurs due to a deformation of the strain gauge. The advantage of the strain gauge used compared to conventional metal strain gauges is its high sensitivity. The strain gauge of the torque sensor system is arranged on an elastic transmission element 2 of the transmission 1. The sensor, comprising the first strain gauge DS1 and the second strain gauge DS2, is thus embedded in the transmission 1.

The first strain gauge DS1 and the second strain gauge DS2 are offset by 90° in the rotation direction of the drive shaft.

The measurement signals detected by sensor 3 are transmitted to the torque sensor system and further processed by the torque sensor system. The sensor 3 detects a first measurement signal S1 using the first strain gauge DS1 and a second measurement signal S2 using the second strain gauge DS2.

The torque sensor system includes a mathematical model and further means to calculate the torque B from the measurement signals S1 and S2.

All calculation times are within the usual range for robotics applications. All calculations and transmission of signals meet the real-time requirement of industrial robots.

1 B shows the measurement signals S1 and S2 detected by the sensor 3 of the torque sensor system according to embodiments of the invention.

The first measurement signal S1 and the second measurement signal S2 are at least approximately sine oscillations. The first measurement signal S1 and the second measurement signal S2 are shifted by 90°. The first measurement signal S1 comprises strain gauge signal detected by means of the first strain gauge DS1 in the form of a voltage that was detected during the rotation of the drive shaft. Analogously to this, the second measurement signal S2 includes the strain gauge signal detected by means of the second strain gauge DS2. The measurement signals S1 and S2 were recorded from 0° drive shaft angle to 360° drive shaft angle, i.e. during rotation of the drive shaft.

2A shows a difference signal D according to embodiments of the invention. The difference signal D is formed by subtracting the first measurement signal S1 from the second measurement signal S2: $$\text{D}=\text{<figure-callout id="2" label="S" filenames="DE102022120560A1\_0020.png" state="{{state}}">S</figure-callout>}2-\text{<figure-callout id="1" label="S" filenames="DE102022120560A1\_0020.png" state="{{state}}">S</figure-callout>}1$$

The difference signal D has approximately the shape of a sine wave.

2 B shows determining portions of the difference signal D according to an embodiment of the invention. Ten sections M were determined along the Y axis.

Furthermore, the difference signal D was divided into four areas along the X-axis. One range includes 90° rotation of the drive shaft.

The sampling frequency of sensor S is 1 kHz. This results in a sampling interval of the sensor S of 1 ms. The maximum speed of the Harmonic Drive transmission is 1500 rpm.

$$\text{Drehungsschritt}\left[\text{Grad}\right]=90\left[\text{Grad}\right]/\text{Max}\text{. Anzahl der Abschnitte}$$

In this case the following applies: $$\begin{array}{l}\text{Max}\text{. Number of sections}=\left(60\ast 1000/\text{maximum speed of the gearbox}\right)/(\text{Abtanstinter}-\\ \text{vall of sensor S*4})\end{array}$$

$$\text{rotation step}\left[\text{Degree}\right]=90\left[\text{Degree}\right]/\text{Max}\text{. Number of sections}$$

The maximum number of sections can thus be calculated: $$\text{Max}\text{. Number of cuts}=60*1000/1500/1*4=10$$

A rotation step can therefore be calculated for a range of rotation of the drive shaft: $$\text{rotation step}\left[\text{Degree}\right]=90\left[\text{Degree}\right]/10=9\left[\text{Degree}\right]$$

$$1/25=40\left[\text{ms}\right]\text{ Periode f}\ddot{\text{u}}\text{r eine Wellendrehung}$$

$$40\left[\text{ms}\right]/1\left[\text{ms}\right]=40\text{ Abtastpunkte f}\ddot{\text{u}}\text{r eine Wellendrehung}$$

$$40\text{ Abtastungen/4 Bereiche}=10\text{ Abtastungen pro 90-Grad-Bereich}$$

Further parameters can be calculated: $$1500\left[\text{rpm}\right]/60\left[\text{sec}\right]=25\text{Shaft revolutions per sec}$$

$$1/25=40\left[\text{ms}\right]\text{period f}\ddot{\text{u}}\text{r a shaft rotation}$$

$$40\left[\text{ms}\right]/1\left[\text{ms}\right]=40\text{sampling points f}\ddot{\text{u}}\text{r a shaft rotation}$$

$$40\text{Scans/4 areas}=10\text{Samples per 90 degree area}$$

The number of sections can therefore be defined as 10. This guarantees that even at maximum shaft speed, the step size of the section is only plus/minus 1.

One range corresponds to a shaft rotation of 90 degrees. Thus, one rotation step is equal to 9 degrees (90 degrees / 10 sections).

3A shows the determination of discrete values of the difference signal D in the predefined sections M according to an embodiment of the invention. After ten sections M have been determined along the Y axis, a value for the respective section is then determined for each of the sections M using the difference signal D. This results in a step-like difference signal S.

3B shows fine adjustment of the sections M of the difference signal D according to an embodiment of the invention. The first and tenth sections of the step-like difference signal S are divided into further subsections.

3C shows the fine adjustment in subsections of the step-like difference signal S according to in 3B illustrated embodiment. The first section was divided into three subsections, labeled 1, 1.25 and 1.5. The tenth section has also been divided into three subsections, labeled 9.5, 9.75 and 10.

4 shows a method according to an embodiment of the invention in a schematic representation.

• Abschnitt 10, wenn Abschnittswert 10 min <= Differenzsignal D < Abschnittswert 10 max
• Abschnitt 9, wenn Abschnittswert 9 min <= Differenzsignal D < Abschnittswert 9 max
• Abschnitt 8, wenn Abschnittswert 8 min <= Differenzsignal D < Abschnittswert 8 max

First, the difference signal D is calculated by forming the difference between the second measurement signal S2 and the first measurement signal S1. The current section membership of the difference signal D is then determined based on the predefined section values of the individual sections:

• Section 10, if section value 10 min <= difference signal D < section value 10 max
• Section 9, if section value 9 min <= difference signal D < section value 9 max
• Section 8, if section value 8 min <= difference signal D < section value 8 max

For example, if a value of the difference signal D is 10 V and section 10 is defined by a minimum section value (“section value 10 min”) of 9.5 V and a maximum section value (“section value 10 max”), the value of the difference signal is assigned to section 10 assigned. The value of the difference signal D therefore has a current section affiliation to section 10.

It is then checked whether there is a change in the section membership of the value of the difference signal D. If the value of the difference signal D, for example, had a previous section affiliation with section 9 and a current section affiliation with section 10, the result is a change of 1. The change in the section affiliation results from the rotation of the drive shaft.

In a further step, the shaft rotation A is calculated. The absolute value of the difference between the current section membership and the previous section membership is calculated. The absolute value of the difference is multiplied by a rotation step: $$\begin{array}{l}\text{A}=\text{Section}\left(\text{current section access}\ddot{\text{O}}\text{rity-previous section affiliation}\ddot{\text{O}}\text{accuracy}\right)*\\ \text{rotation step}\left[\text{Degree}\right]\end{array}$$

The rotation step is, for example, 9° if the difference signal was divided into ten sections over 90°.

$$\text{T2}\left[\text{Nm}\right]:\text{ Gesamtlastmoment}+=\text{B}$$

In a next step, the shaft rotation A in the ROM of the microcontroller or in the external EEPROM is added to the total shaft rotation T1. Furthermore, the torque B is added to the total load torque T2. $$\text{T1}\left[\text{Degree}\right]:\text{Total shaft rotation}+=\text{A}$$

$$\text{T2}\left[\text{Nm}\right]:\text{Total load moment}+=\text{b}$$

The shaft rotation A and the torque B are added if a change in the section affiliation is detected, i.e. H. the static load moment is ignored due to the lower fatigue compared to dynamic fatigue and also for simplicity.

The total shaft rotation T1 is therefore used to determine the total rotation of the drive shaft since the start of detection.

In a next step, the calibration parameter α is updated based on the last value of T1 and T2.

If a previous section membership does not exist, it is initialized with the current section membership. This usually happens in the first calculation cycle after switching on the torque sensor system or with the start of recording.

5 shows an overview of the method according to a further embodiment of the invention.

A microcontroller first receives the first measurement signal S1 and the second measurement signal S2, which were converted into digital signals using an analog-digital converter ADC. In a first step, the torque B is determined from the measurement signals S1 and S2 using a mathematical model, using a calibration parameter α: $$\text{b}=\mathrm{\alpha }\left(\text{<figure-callout id="1" label="S" filenames="DE102022120560A1\_0020.png" state="{{state}}">S</figure-callout>}1+\text{S}2\right)$$

In a second step, the shaft rotation A is calculated according to the steps already mentioned.

Thus, for the calculation of the shaft rotation A, only rotation steps are taken into account in which dynamic fatigue of the elastic transmission element 2 was detected by means of the section membership.

After calculating the shaft rotation A, the value of the shaft rotation is stored in a non-volatile electronic memory component of the microcontroller, for example ROM or EEPROM, if a change in the difference was detected.

The total shaft rotation T1 and the total load torque T2 are then calculated.

In a third step, the mathematical model is calibrated based on the total shaft rotation T1 and the total load torque T2.

The calibration parameter α is updated based on the last value of the total shaft rotation T1 and the total load torque T2 using a predefined calibration function k: $$\mathrm{\alpha }=\text{k}\left(\text{<figure-callout id="1" label="T" filenames="DE102022120560A1\_0020.png" state="{{state}}">T</figure-callout>}\mathrm{1,}\text{T}2\right)$$

The updated calibration parameter α is used in the torque calculation in the next calculation cycle.

The calibration function k was determined and predefined using a service life test in the development phase. In the present case, a linear regression model is used as the calibration function k.

berechnet werden, wobei β1 und β2 Regressionskoeffizienten des linearen Regressionsmodells sind. Die Regressionskoeffizienten wurden dabei vorab mittels Tests berechnet.Using a linear regression model as the calibration function k, the calibration coefficient α can be obtained using the formula $$\mathrm{\alpha }=\mathrm{<figure-callout\; id="1"\; label="\beta "\; filenames="DE102022120560A1\_0020.png"\; state="\{\{state\}\}">\beta </figure-callout>}1*\text{<figure-callout id="1" label="T" filenames="DE102022120560A1\_0020.png" state="{{state}}">T</figure-callout>}1+\mathrm{<figure-callout\; id="2"\; label="\beta "\; filenames="DE102022120560A1\_0020.png"\; state="\{\{state\}\}">\beta </figure-callout>}2*\text{<figure-callout id="2" label="T" filenames="DE102022120560A1\_0020.png" state="{{state}}">T</figure-callout>}2$$

can be calculated, where β1 and β2 are regression coefficients of the linear regression model. The regression coefficients were calculated in advance using tests.

It is understood that the order of the steps mentioned may vary depending on the use case. For example, the total load torque T2 and then the total shaft rotation T1 can be determined first. It is also possible to simultaneously determine the total load torque T2 and the total shaft rotation T1.

The torque B can be further processed in further steps in the torque sensor system. The torque B can be transmitted to a control loop, for example. The control circuit can be intended to regulate a motor characteristic value and/or a speed. The control circuit can also be intended to regulate a torque. Advantageously, control based on the control loop can be improved by calibrating the torque sensor system. Furthermore, the torque B can be transmitted from the torque system to an engine control.

In addition to use in passenger cars, the principles disclosed here can also be implemented in other areas of application, e.g. in lorries, trucks, motorcycles, axle drives in robots and drives in flight simulators.

The algorithms or computer programs required for the above functions can expediently be implemented in whole or in part in a microcontroller, in the torque sensor system and/or in another computer system connected to these devices. The computing system may be a local computing device (e.g., personal computer, laptop, tablet computer, or cell phone) with one or more processors and one or more storage devices, or may be a distributed computing system (e.g., a cloud computing system with one or more processors or one or more Storage devices distributed at various locations, for example at a local client and/or one or more remote server farms and/or data centers).

Depending on particular implementation requirements, embodiments of the invention may be implemented in hardware or software. Some or all of the method steps may be performed by (or using) a hardware device, such as a processor, a microprocessor, a programmable computer, or an electronic circuit.

Although some aspects have been described in the context of a device, it is clear that these aspects also represent a description of the corresponding method, where a block or a device corresponds to a method step or a function of a method step. Analogously, aspects that are described as part of a method step also represent a description of a corresponding block or element or a property of a corresponding device.

QUOTES INCLUDED IN THE DESCRIPTION

This list of documents listed by the applicant was generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Non-patent literature cited

• Hashimoto, Minoru, Yoshihide Kiyosawa, and Richard P. Paul. “A torque sensing technique for robots with harmonic drives.” IEEE Transactions on Robotics and Automation 9.1 (1993): 108-116 [0003]

Claims (10)

A method for calibrating a torque sensor system for a transmission (1), wherein the transmission (1) has an elastic transmission element (2) and a sensor (3) of the torque sensor system is arranged on the elastic transmission element, the method having the following method steps: Detecting at least one measurement signal (S1, S2) using the sensor (3) of the torque sensor system; Determining a torque (B) from the at least one measurement signal (S1, S2) using a mathematical model; Determining a total shaft rotation (T1) from the at least one measurement signal (S1, S2); Determining a total load torque (T2) from the torque (B); Calibrate the mathematical model based on total shaft rotation (T1) and total load torque (T2). The procedure according to Claim 1 , wherein the sensor (3) of the torque sensor system comprises a first strain gauge (DS1) and a second strain gauge (DS2), wherein a first measurement signal (S1) is detected by means of the first strain gauge (DS1) and a second by means of the second strain gauge (DS2). Measurement signal (S2) is detected. The procedure according to Claim 2 , wherein the first strain gauge (DS1) and the second strain gauge (DS2) are offset by 90 ° in a rotation direction. The procedure according to one of the Claims 2 or 3 , wherein determining the total shaft rotation (T1) and determining the total load torque (T2) has the following method steps: determining a difference signal from the first measurement signal (S1) and the second measurement signal (S2); Determining a change in the difference signal (D) to a previous state using predefined sections (M); Determine the total shaft rotation (T1) and the total load torque (T2) if a change in the difference signal (D) was determined. The procedure according to one of the Claims 1 until 4 , where the mathematical model is calibrated using a predefined calibration function (k) and based on the total shaft rotation (T1) and the total load torque (T2). A sensor system comprising means for carrying out the method according to one of Claims 1 until 5 . A computer program comprising instructions which, when the program is executed by a computer, cause it to carry out the method according to one of the Claims 1 until 5 to carry out. A transmission system, comprising a transmission (1), an elastic transmission element (2) and a torque sensor system with a sensor (3) and a mathematical model, wherein the sensor (3) of the torque sensor system is arranged on the elastic transmission element (2), wherein the Torque sensor system is designed such that a torque (B) can be determined using the mathematical model and the mathematical model can be calibrated using at least one measurement signal (S1, S2) detected by the sensor (3). The transmission system after Claim 8 , wherein the sensor (3) of the torque sensor system comprises a first strain gauge (DS1) and a second strain gauge (DS2). The transmission system according to one of the Claims 8 or 9 , wherein the mathematical model is designed to be calibratable by means of a calibration function (k), wherein the calibration function (k) comprises a regression model, a multi-layer perceptron model and / or an artificial neural network.

Images