WO2023169622A1

WO2023169622A1 - Method for determining a torque

Info

Publication number: WO2023169622A1
Application number: PCT/DE2023/100099
Authority: WO
Inventors: Kazuaki Kaneko; Jens Heim; Enrico CHIOVETTO
Original assignee: Schaeffler Technologies AG & Co. KG
Priority date: 2022-03-08
Filing date: 2023-02-08
Publication date: 2023-09-14
Also published as: DE102022105335A1

Abstract

The present invention relates to a method for determining a torque (T) of a drive train (1) by means of a mathematical model (9). The method may comprise capturing input data (E). The input data (E) may comprise a voltage (U) measured using a torque sensor (5). The input data (E) may also comprise at least one signal value (Ω, Ψ, I) of a motor controller (4). Furthermore, the method may comprise determining the torque (T) by means of a mathematical model (9) based on the voltage (U) and the at least one signal value (Ω, Ψ, I). The present invention also relates to a machine learning model for determining a torque (T) of a drive train (1).

Description

METHOD FOR DETERMINING TORQUE

TECHNICAL FIELD

The present invention generally relates to the technical field of methods for determining a torque of a device to be driven, in particular a drive train.

BACKGROUND

Methods for determining torque have a wide range of applications in drive technology. For example, to improve the driving behavior of motor vehicles used to transport people, the torque of the drive train is measured or determined mathematically. A powertrain typically includes a motor with a motor shaft, a motor controller, a torque sensor with a microcontroller, and a transmission with an output shaft. For the purposes of the invention, the person skilled in the art understands torque as the torque applied to the output shaft of the drive train. In addition to vehicle technology, torque is also an important parameter in robotics. Especially with industrial robots, high torque is often necessary to move the joints of a multi-link manipulator with sufficient force.

The previously known methods for determining the torque of a drive train typically consist of attaching a torque sensor to an output shaft of the drive train. Such a torque sensor is generally formed by attaching strain gauges to a component of the transmission, in particular an elastic transmission element with external teeth, and by determining a torque based on input data, the input data being measured 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.

Other known methods for determining torque are based on mathematical formulas that are derived from the physical behavior of the drive train. Such a method for determining the torque is disclosed, for example, in patent DE 199 63 279 B4 with the title “Method for determining an output torque of a drive device and drive device”. The output torque is derived using a mathematical formula, using a lubricant temperature of the transmission, a speed of the engine and an output torque of the engine as input data. It is not intended that a torque sensor is used.

Although the approaches mentioned above allow the torque of the drive train to be determined at least as a rough approximation, there are various disturbance variables acting on the drive train, such as inertia and temperature effects, which are not recorded by the known methods and which lead to an inaccurate determination of the torque.

It is therefore an object of the present invention to provide an improved method for determining a torque of a drive train and thus to at least partially overcome the above-mentioned disadvantages of the prior art. Another object of the present invention is to provide an improved mathematical model for determining torque.

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 determining a torque of a device to be driven, in particular a drive train. The method can include acquiring input data. The input data may include a voltage measured by a torque sensor. The input data can further include at least one signal value from an engine control system. Furthermore, the method may include determining the torque using a mathematical model based on the voltage and the at least one signal value. The drive train is preferably part of a vehicle or robot and may include a motor with a motor shaft, a motor controller, a torque sensor with a microcontroller and a transmission with an output shaft. The engine can be an internal combustion engine, an electric motor, a fuel cell, a hybrid drive or another energy converter.

The torque of the drive train is understood to mean the torque available to drive the vehicle, the robot or a manipulator member of the robot. The expert understands torque to be the torque applied to the output shaft of the drive train. The torque of the drive train is therefore the output torque of the transmission. The torque of the drive train can, for example, be transmitted directly to the wheels of the vehicle.

By determining the torque using the mathematical model, the person skilled in the art typically understands any type of determination and/or calculation in the mathematical sense. Any type of mathematical operation can be used, especially iterative procedures or solving differential equations. Reading from tables can also be used to determine torque.

Some or all of the method steps may be performed by (or using) a hardware device, such as the torque sensor 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 engine technology.

Typically, the person skilled in the art understands input data to be the data that is made available to the mathematical model for determining the torque.

The engine control is usually used to control and/or regulate the engine. The engine control can read out, measure and/or save all signal values relating to the engine. The engine control can have means for reading, measuring and/or Save the signal values. Signal values can be electrical signals such as voltages and/or currents. The signal values can also be pressures, temperatures, rotational speeds, accelerations and/or speeds. The engine control can have access to sensors, such as temperature and/or pressure sensors, depending on the application, or signal values can be transmitted from the sensors to the engine control. It is also possible for the engine control system to measure a large number of signal values at the same time and transmit them to the other components of the drive train.

The engine control can also be designed to control the engine and/or other components of the device to be aborted. The engine control can, for example, regulate the speed of a vehicle. The control can be implemented by a control loop implemented as an algorithm.

The torque sensor is preferably equipped with a microcontroller. The mathematical model or at least parts of the mathematical model can be implemented on the microcontroller.

Helical gears, bevel gears, helical gears, planetary gears or Harmonie Drive gears can be used as gears.

The method according to the invention advantageously leads to an improvement of the measuring method using a torque sensor. In addition to the voltage measured by the torque sensor, the method can provide additional data from the engine control system in order to determine the torque of the drive train even more precisely. Furthermore, the method according to the invention has the advantage that input data from various components of the drive train, in particular the engine control and the torque sensor, can be made available to the mathematical model at the same time. In other words, measured signal values from different sensors can be merged (multi-sensor fusion) and further processed by the mathematical model.

In a further aspect of the invention, the at least one signal value of the motor control comprises a motor position and/or a direction of rotation. The motor position refers to the angle of rotation of the motor shaft leading out of the motor. The motor shaft can for example, initially have a rotation angle of approximately 0°. After half a revolution of the motor, the rotation angle is essentially 180°. The direction of rotation means the direction of rotation of the motor shaft. The direction of rotation can be clockwise or counterclockwise. The motor position and the direction of rotation have the advantage that they are preferably available as stored data in the motor control or are measured by the motor control.

In another aspect, the input data includes a temperature. Temperatures can be lubricant temperatures in the transmission or a temperature measured in the engine. Advantageously, the method can thus detect possible temperature effects acting on the torque of the drive train.

In another aspect, the mathematical model includes a finite element model. A finite element model has the advantage that any complex geometry of the drive train and its physical behavior can be mapped based on differential equations and continuum theory. Non-linearities regarding the material and/or geometry can also be easily mapped.

In another aspect, the mathematical model includes a machine learning model. The machine learning model can in particular be an artificial neural network.

Artificial neural networks (ANN) are systems inspired by biological neural networks, such as those found in a retina or brain. AN Ns 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 can involve adjusting the weights of the Include nodes and/or edges of the artificial neural network, ie to achieve a desired output for a particular input.

Alternatively, the machine learning model 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 machine learning model may be a Bayesian network, which is a probabilistic directed acyclic graphical model. A Bayesian network can represent a set of random variables and their conditional dependencies using a directed acyclic graph. Alternatively, the machine learning model may be based on a genetic algorithm, which is a search algorithm and heuristic technique that mimics the process of natural selection.

In a further aspect, the machine learning model is trained using at least one training data set. The at least one training data set can include a voltage measured by a torque sensor, at least one signal value from a motor control and a reference torque.

The training data set can, for example, be created on a test bench. The voltage measured by the torque sensor, the at least one signal value from the engine control and the reference torque are measured on the test bench in different operating states of the drive train.

For example, “supervised learning” can be used as a training method for the machine learning model. In supervised learning, the machine learning model is 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, that is, a desired output value is assigned to each training sample. By specifying both training samples and desired output values (here the reference torque mentioned above), the machine learning model “learns” what output value should be based on an input sample that is similar to the Sample values provided during training must be provided. 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 from 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.

In a further aspect, the mathematical model is implemented on a microcontroller of the torque sensor. An advantage of this implementation is that determining the torque using a mathematical model can be carried out in its entirety on the microcontroller and no further components are necessary to execute the mathematical model.

In a further aspect, the torque determined using the mathematical model is transmitted to a control loop. The control circuit can be intended to regulate a current strength of the motor and/or a speed. Advantageously, a control system based on the control loop can thus be improved.

In a further aspect, the plausibility check procedure has further procedural steps. In a further step, an estimated torque can be obtained based on a current strength of the motor. Furthermore, an upper torque limit and/or a lower torque limit can be determined based on the estimated torque. The torque determined using the mathematical model can then be evaluated based on the upper torque limit and/or the lower torque limit. The current strength of the motor can be read out by the motor control as a signal value.

The estimated torque can be determined, for example, using the following formula, whereby a torque constant and a gear ratio of the gearbox have previously been determined by the motor control: estimated torque = current of the motor * torque constant * gear ratio of the gearbox

The evaluation of the torque determined using the mathematical model can be checked, for example, by the following condition: estimated torque - A < torque determined using the mathematical model < estimated torque + A

Where A is a calibration parameter taking into account the current level of accuracy of the estimated torque. Parameter A can be defined and saved in advance in the microcontroller of the torque sensor. As long as the torque determined by the mathematical model is within the range “estimated torque (current-based) ± A”, it can be assessed whether the torque determined using the mathematical model is plausible. If the torque determined by the mathematical model is out of range, the error condition is output as a diagnostic condition or status.

In a further aspect, a device for data processing is provided, in particular a microcontroller of a torque sensor, comprising means for carrying out one of the methods described herein.

The data processing device may comprise a storage medium (or a data carrier or a computer-readable medium) which comprises, stored thereon, a computer program for carrying out one of the methods described herein when it is used by one processor is running. The data carrier, digital storage medium or recorded medium is usually tangible and/or non-seamless. Another embodiment of the present invention is an apparatus as described herein that includes a processor and the storage medium.

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

According to the invention, a machine learning model for determining a torque of a drive train is also provided. The machine learning model can have input means for receiving a voltage measured by a torque sensor and at least one signal value from a motor control. Furthermore, the machine learning model may have means for determining the torque based on the voltage and the at least one signal value.

The machine learning model can refer to algorithms and statistical models that computer systems can use to perform a specific task without using explicit instructions, instead of relying on models and inference. For example, in machine learning, instead of a rule-based transformation of data, a transformation of data can be used that can be derived from an analysis of historical and/or training data. For example, the physical behavior of the powertrain can be analyzed using a machine learning model or using a machine learning algorithm. In order for the machine learning model to analyze the physical behavior of the powertrain, the machine learning model can be trained using training data sets as input and training content information as output. By training the machine learning model with a large number of training data sets and/or training sequences (e.g. signal values) and associated training content information (e.g. labels or annotations), the machine learning model “learns” to recognize the physical behavior of the powertrain, such that the physical behavior contained in not included in the training data can be predicted using the machine learning model. The same principle can be used for other types of sensor data as well: By training a machine learning model using From training sensor data and a desired output, the machine learning model “learns” a transformation between the sensor data and the output, which can be used to provide an output based on non-training sensor data provided to the machine learning model. The provided data (e.g. sensor data and/or metadata) can be preprocessed to obtain a feature vector, which is used as input for the machine learning model.

In a further embodiment, the invention provides a training data set for training the machine learning model. The at least one training data set can include a voltage measured by a torque sensor, at least one signal value from a motor control and a reference torque.

BRIEF DESCRIPTION OF THE DRAWINGS

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

Fig. 1: Drive train 1 according to embodiments of the invention.

Fig. 2: Method according to embodiments of the invention at a glance.

Fig. 3A: Mathematical model 9 according to embodiments of the invention.

Fig. 3B: Motor control 4 and torque sensor 5 according to embodiments of the invention.

Fig. 4: Method for plausibility checking according to embodiments of the invention.

DESCRIPTION OF PREFERRED EMBODIMENTS

Embodiments of the invention are described below, in which a method for determining a torque of a drive train is used in order to make the drive train particularly efficient. Fig. 1 shows a drive train 1 according to an embodiment of the invention. In the example shown, the drive train 1 is used in a passenger car and includes a motor 2 with a motor shaft 3, a motor control 4, a torque sensor 5 with a microcontroller, a data transmission system 6 and a transmission 7 with an output shaft 8. The motor shaft 3 of the motor 2 is connected to the gearbox 7 and thus transmits the engine torque generated by the engine 2 to the gearbox 7. The gearbox 7 is a Harmonie Drive gearbox with an elastic transmission element that is characterized by a high gear ratio and rigidity. The elastic transmission element is also known as a flexspine. The transmission 7 also has an output shaft 8, the torque of the output shaft 8 being passed on to the wheels of the passenger car.

The torque sensor 5 comprises a strain gauge, wherein the strain gauge consists of a semiconductor and measures a voltage applied to 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 5 is arranged on an elastic transmission element of the transmission 7. The torque sensor 5 is thus embedded in the transmission 7. The torque sensor 5 is connected to the engine control 4 via a data transmission system 6. The engine control 4 measures the signal values relating to the engine 2 and transmits the signal values to the torque sensor 5.

The method for determining the torque is implemented in the form of a computer program on the microcontroller of the torque sensor 5. After the torque has been determined, the torque is transmitted to the engine control 4 via the data transmission system 6.

Fig. 2 shows a method according to an embodiment of the invention in a schematic representation. The method according to the embodiment includes the following steps: First step S1.1: The torque sensor 5 measures a voltage and passes the voltage on to the microcontroller.

Second step S1 .2: The motor control 4 passes on a signal value to the microcontroller.

Third step S1.3: A mathematical model 9 implemented on the microcontroller determines the torque based on the voltage and the signal value.

It is understood that the order of the steps S1.1, S1.2 and S1.3 mentioned can vary depending on the application. For example, the signal value can first be passed on from the motor control 4 to the microcontroller and then the voltage can be measured by the torque sensor 5 and passed on to the microcontroller. A simultaneous execution of the first step S1.1 and the second step S1.2 is also possible.

After determining the torque, the torque can be transmitted to the engine control 4 in a further step, not shown in the exemplary embodiment.

3A shows the mathematical model 9 according to an embodiment of the invention. Various signal values are transmitted to the mathematical model 9 as input data E, whereby in the example shown, the input data E is a voltage II measured by the torque sensor 5, a temperature of the transmission 0, a position of the motor Q and a direction of rotation of the motor

acts. The mathematical model 9 calculates based on the four signal values II, 0, Q and

a torque T of the output shaft 8 of the drive train 1. The calculated torque T is provided as a signal value for further use. The mathematical model 9 is an artificial neural network that was trained using training data before use in the drive train 1. The calculation of the torque T by means of the artificial neural network and the components of the drive train 1 are optimally adjusted to the hard real-time requirements required in engine technology. The calculation time is in the usual range for engine control 4 applications.

3B shows the motor control 4 and the torque sensor 5 according to a further embodiment of the invention. The engine control 4 and the torque sensor 5 are connected to one another via a data transmission system 6 and can therefore send each other the signal values Q, I, T and Z. In the example shown, the Motor control 4 the torque sensor 5 the position of the motor Q, the direction of rotation of the motor

and a current strength of the motor I. The torque sensor 5 transmits the torque T of the output shaft 8 of the drive train 1 and a status Z of the torque sensor 5 to the engine control 4.

In the example shown, the data transmission system 6 is an EtherCAT system. The data transmission system 6 thus advantageously meets the real-time requirements. The engine control 4 and a torque sensor 5 can send or exchange the signal values Q, I, T and Z synchronously. Furthermore, the signal values and I can be transmitted from the engine control 4 to the torque sensor 5 and used there directly, for example to calculate the torque T. The torque T can then be transmitted directly back to the engine control 4.

Fig. 4 shows a method for plausibility checking according to an embodiment of the invention. All method steps S2.1, S2.2, S2.3, S2.4 and S2.5 are carried out on a microcontroller belonging to the torque sensor 5.

In a first step S2.1, the microcontroller receives the signal values Q,

and I from the motor control 4 via the data transmission system 6. The signal values include the current strength of the motor I. In a second step S2.2, the microcontroller calculates the torque T using a mathematical model 9 implemented on the microcontroller and based on the signal values

and I. Furthermore, in the second step S2.2, the microcontroller determines an estimated torque based on the current strength of the motor I. The microcontroller additionally determines an upper torque limit and a lower torque limit based on the estimated torque. In a third step S2.3, the microcontroller evaluates the torque T determined using the mathematical model 9 based on the upper torque limit and the lower torque limit.

If the torque T determined using the mathematical model 9 is outside the torque limits, ie greater than the value defined by the upper torque limit, or smaller than the value defined by the lower torque limit, the microcontroller sets S2 in a further step .4 the status Z to “Error”. If the torque T determined using the mathematical model 9 is within the torque limits, ie smaller than the value defined by the upper torque limit and at the same time greater than the value defined by the lower torque limit, the microcontroller sets in a further step S2.5 status Z to “plausible”.

The status Z and the torque T can be transmitted to the engine control 4 in a further step using the data transmission system 6. The method shown in Fig. 4 is therefore used to check the technical plausibility of the torque T determined using the mathematical model 9. If the status Z is output as an “error”, the torque T is not used for any further calculations in the engine control 6.

Although embodiments of the invention have been described primarily for use in passenger vehicles, it should be mentioned that the principles disclosed here can nevertheless be implemented in other areas of application, for example in trucks, trucks, motorcycles, axle drives in robots and drives in flight simulators.

The algorithms or computer programs necessary for the above functions can expediently be implemented in whole or in part in the torque sensor 5, in the microcontroller, in the motor control 4 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.

Claims

PATENT CLAIMS

1 . A method for determining a torque (T) of a device (1) to be driven, in particular a drive train, the method having the following method steps:

Acquiring input data (E), the input data (E) comprising: a voltage (II) measured by a torque sensor (5); and at least one signal value (Q,

I) an engine control;

Determining the torque (T) using a mathematical model (9) based on the voltage (II) and the at least one signal value (Q,

I).

2. The method according to claim 1, wherein the at least one signal value (Q,

I) the motor control (4) comprises a motor position (Q) and/or a direction of rotation (^P).

3. The method according to any one of claims 1 to 2, wherein the input data (E) further comprises a temperature (0).

4. The method according to any one of claims 1 to 3, wherein the mathematical model (9) comprises a finite element model.

5. The method according to one of claims 1 to 3, wherein the mathematical model (9) comprises a machine learning model, in particular an artificial neural network.

6. The method according to claim 5, wherein the machine learning model is trained using at least one training data set, the at least one training data set comprising: a voltage (U) measured by a torque sensor (5); at least one signal value (Q,

I) an engine control; and a reference torque.

7. The method according to one of claims 1 to 6, wherein the mathematical model (9) is implemented on a microcontroller of the torque sensor (5).

8. The method according to one of claims 1 to 7, wherein the torque (9) determined by means of the mathematical model (9) is transmitted to a control loop, the control loop regulating a current strength of a motor (I) and / or a speed.

9. The method according to one of claims 1 to 8, wherein the method for plausibility checking has the following further method steps:

Determining an estimated torque based on a current of a motor (I);

determining an upper torque limit and lower torque limit based on the estimated torque;

Evaluating the torque (T) determined using the mathematical model (9) based on the upper torque limit and the lower torque limit.

10. A device for data processing, in particular a microcontroller of a torque sensor (5), comprising means for carrying out the method according to any one of claims 1 to 9.

11. Computer program comprising instructions which, when the program is executed by a computer, cause it to carry out the method according to any one of claims 1 to 10.

12. A machine learning model for determining a torque (T) of a drive train (1), the machine learning model having:

Input means for receiving: a voltage (U) measured by a torque sensor (5); and at least one signal value from an engine control (Q,

I); and Means for determining the torque (T) based on the voltage (II) and the at least one signal value (Q,

I).

13. A training data set for training the machine learning model according to claim 12, wherein the at least one training data set comprises: a voltage (II) measured by a torque sensor (5); at least one signal value from an engine control (Q,

I); and a reference torque.