DEVICE AND METHOD FOR DETECTING ANOMALIES OF THE MOTOR OF AN ELECTRIC COLUMN SERVO SYSTEM (EPS), AND COMPUTER-READABLE STORAGE MEDIUM FOR STORED A PROGRAM FOR EXECUTING THE METHOD

Abstract

Disclosed are a device and a method for detecting an anomaly of a motor of an electric column power steering system (EPS) and a non-transitory computer-readable storage medium that stores a program for carrying out the method.The device for detecting the motor anomaly of the column EPS is a device for detecting an anomaly of a motor of a column EPS, which detects an anomaly of a motor of a column EPS that provides an additional steering force for a steering column of an EPS system of a vehicle, and comprises a memory configured to store one or more instructions, and a processor configured to execute the one or more instructions, wherein the processor executes the one or more instructions to input operation data relating to the steering wheel operation by a driver of the vehicle and state data indicating a state of the vehicle into an artificial neural network model, to obtain an estimated output value of the motor from the artificial neural network model, to compare the estimated value with an output measurement of the motor and to detect an anomaly of the motor.

Description

CROSS-REFERENCE TO A RELATED APPLICATION

[0001] This application claims priority to and the benefit of Korean Patent Application No. 10-2023-0106181, filed on August 14, 2023, and Korean Patent Application No. 10-2024-0021747, filed on February 15, 2024, the disclosures of which are incorporated herein by reference in their entirety. BACKGROUND 1. Field of technology

[0002] The present disclosure relates to an apparatus and method for detecting an abnormality of a motor of an electric column power steering (EPS) system, and a non-transitory computer-readable storage medium storing a program for carrying out the method, and more particularly, to an apparatus and method for detecting an abnormality of a motor of a column EPS capable of detecting a deterioration in the performance of a motor that is arranged on a steering column connected to a steering wheel and provides an auxiliary steering force when a driver operates the steering wheel, and a non-transitory computer-readable storage medium storing a program for carrying out the method. 2. Discussion of the state of the art

[0003] Electric power steering (EPS) systems use a motor to assist the driver with steering. Among EPS systems, a column EPS system is designed to apply auxiliary steering force generated by a motor to a steering column connected to a vehicle's steering wheel.

[0004] If a motor that generates auxiliary steering force in a column-mounted EPS system fails, the driver will have difficulty steering the vehicle. This can result in steering that is not as desired, which can lead to accidents.

[0005] In this way, the motor of a pillar EPS system is a very important element for vehicle safety. To improve the safety of vehicles equipped with a pillar EPS system, it is therefore necessary to predict a performance decline of a pillar EPS motor before the pillar EPS motor fails.

[0006] However, conventional approaches to motor failsafe operation in a column EPS focus on post-failure diagnosis and follow-up measures rather than preventing failure in advance. Furthermore, conventional column EPS systems use motor current-based failsafe operation, which can only inadequately respond to a power degradation.

[0007] (Patent Document) Korean Patent Registration No. 2098049, “METHOD FOR CONTROLLING AN EPS THAT CURSES MOTOR DRIVING POWER IN CASE OF ELECTRIC OVERCURRENT”, registered on April 1, 2022 DEMOLITION

[0008] The present disclosure relates to an apparatus and method for detecting an abnormality of a motor of an electric column power steering (EPS) system capable of predicting a performance deterioration of a motor disposed in a column EPS system of a vehicle and generating an auxiliary steering force, and a non-transferable computer-readable storage medium storing a program for performing the method.

[0009] The present disclosure also aims to provide an apparatus and a method for detecting an abnormality of a motor of a column EPS capable of efficiently detecting in advance a decrease in the performance of a motor of a column EPS using signals of a controller area network (CAN) in a vehicle, and a non-transferable computer-readable storage medium storing a program for performing the method.

[0010] The objects of the present disclosure are not limited to the objects described above, and other objects not mentioned will become apparent to one skilled in the art to which the present disclosure belongs from the following description.

[0011] According to one aspect of the present disclosure, an apparatus for detecting an abnormality of a motor of a column EPS that provides auxiliary steering force to a steering column of an EPS system of a vehicle is provided, comprising a memory configured to store one or more instructions and a processor configured to execute the one or more instructions, wherein the processor executes the one or more instructions to input operation data related to steering wheel operation by a driver of the vehicle and state data indicative of a state of the vehicle into an artificial neural network model, obtain an estimated output value of the output of the motor from the artificial neural network model, compare the estimated value with an output measurement value of the motor, and detect the abnormality of the motor.

[0012] The operation data may include at least one of a steering angle, a steering angular velocity, or a steering torque.

[0013] The state data may include at least one of a speed of the vehicle, a lateral acceleration of the vehicle, a yaw rate of the vehicle, and a wheel speed of the vehicle.

[0014] The process and status data can consist of signals that can be retrieved via a CAN of the vehicle.

[0015] The artificial neural network model may include a generative adversarial network (GAN) with a generator that receives the process data and the state data and produces the estimated value.

[0016] The artificial neural network model may further include a discriminator that receives measurement data including the process data, the state data, and the measurement value and outputs a discrimination value for the measurement data.

[0017] The generator can be composed of a multivariant transformer.

[0018] The processor may execute the one or more instructions to input fault data related to a difference between the estimated value and the measured value into an anomaly detection model and determine whether the engine is abnormal.

[0019] The anomaly detection model can use a one-class support vector machine (OCSVM) algorithm.

[0020] There may be a plurality of data sets including the operation data, the state data, the measured value, and the estimated value, and the error data may include a mean and a standard deviation of the errors between the measured value and the estimated value obtained from each of the plurality of data sets, a maximum absolute error between the measured value and the estimated value of the plurality of data sets, and a discrimination value of the discriminator for the measured data of the plurality of data sets.

[0021] The discriminator can additionally receive estimation data containing the process data, the state data and the estimated value, and additionally output the discrimination value for the estimation data.

[0022] The artificial neural network model can be constructed by the generator and the discriminator alternately performing learning, and the operation data, state data and measured value used during learning can be obtained when the engine is in a normal state.

[0023] According to another aspect of the present disclosure, there is provided a method for detecting an abnormality of a motor of a column EPS, in which an abnormality of a motor of a column EPS that provides auxiliary steering force to a steering column of an EPS system of a vehicle is detected, which comprises inputting operation data related to steering wheel operation by a driver of the vehicle and state data indicative of a state of the vehicle into an artificial neural network model and obtaining an estimated output value of the motor from the artificial neural network model, and comparing the estimated value with an output measurement value of the motor by the processor and detecting the abnormality of the motor.

[0024] The operation data may include at least one of a steering angle, a steering angular velocity, or a steering torque.

[0025] The state data may include at least one of a speed of the vehicle, a lateral acceleration of the vehicle, a yaw rate of the vehicle, and a wheel speed of the vehicle.

[0026] The artificial neural network model may comprise a GAN including a generator that receives the operation data and the state data and generates the estimated value, and a discriminator that receives the measurement data including the operation data, the state data, and the measurement value and outputs a discrimination value for the measurement data.

[0027] To obtain the estimated value, the processor can input the operation data and the state data into the generator and obtain the estimated value generated by the generator.

[0028] Detecting the abnormality of the engine may include inputting, by the processor, measurement data, including the operation data, the state data, and the measured value, to the discriminator and obtaining the discrimination value for the measurement data generated by the discriminator, and inputting, by the processor, error data, including the discrimination value for the measurement data and numerical values ​​related to a difference between the estimated value and the measured value, to an anomaly detection model and obtaining an output of the anomaly detection model.

[0029] There may be a plurality of data sets including the operation data, the state data, the estimated value, and the measured value, and the error data may include a mean and a standard deviation of the errors between the measured value and the estimated value obtained from each of the plurality of data sets, a maximum absolute error between the measured value and the estimated value of the plurality of data sets, and a discrimination value of the discriminator for the measured data of the plurality of data sets.

[0030] The artificial neural network model can be constructed by the generator and the discriminator alternately performing learning, and the operation data, state data and measured value used during learning can be obtained when the engine is in a normal state.

[0031] According to another aspect of the present disclosure, a non-transferable, computer-readable storage medium is provided having stored therein a program including at least one instruction for performing the column EPS motor abnormality detection method. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The above and other objects, features and advantages of the present disclosure will become more apparent to those skilled in the art by describing exemplary embodiments thereof with reference to the accompanying drawings. Fig. 1 is a diagram schematically showing the configuration of an electric column power steering (EPS) system of a vehicle; Fig. 2 is a block diagram showing the configuration of an apparatus for detecting an abnormality of a motor of a column EPS according to an embodiment of the present disclosure; Fig. 3 is a diagram illustrating models used to obtain an estimated value or detect an abnormality in the column EPS motor abnormality detection apparatus according to the embodiment of the present disclosure; Fig. 4 is a diagram showing a detailed configuration of an artificial neural network model; Fig. 5 is a diagram showing a detailed configuration of a generator of the artificial neural network model; Fig. 6 is a diagram showing the operation of the column EPS motor abnormality detecting apparatus according to the embodiment of the present disclosure; Fig. 7 is a flowchart of a method for detecting an abnormality of a motor of a column EPS according to an embodiment of the present disclosure; and Fig. 8 is a detailed flowchart of an abnormality detection process of the column EPS motor abnormality detection method according to the embodiment of the present disclosure. DETAILED DESCRIPTION OF EXEMPLARY EMBODIMENTS

[0033] In the following, embodiments of the present disclosure will be described in detail so that a person skilled in the art to which the present disclosure belongs can easily practice the embodiments. The present disclosure may be implemented in many different forms and is not limited to the embodiments described herein. To clearly describe the present disclosure, parts that are not related to the description are omitted from the accompanying drawings, and the same or similar components are denoted by the same reference numerals throughout the description.

[0034] The words and terms used in the specification and claims are not to be understood merely in their ordinary or dictionary meaning, but should be interpreted in their meaning and concept consistent with the technical spirit of the present disclosure, in accordance with the principle that inventors may define terms and concepts to best describe their disclosure.

[0035] It should be noted that in the description, terms such as "comprise" or "have" are intended to specify the presence of features, a respective number, steps, operations, components, parts or combinations thereof described in the description, but are not intended to exclude the possibility of the presence or addition of one or more other features, a respective different number, one or more other steps, operations or components, parts or combinations thereof.

[0036] Fig. 1 is a diagram schematically showing the configuration of an electric column power steering (EPS) system of a vehicle.

[0037] Fig. 1 shows, by way of example, a column EPS system 10, operation data relating to the steering wheel operation by a driver of a vehicle equipped with the column EPS system 10, and state data relating to a driving state of the vehicle.

[0038] Referring to Fig. 1, the column EPS system 10 of the vehicle may include a driver-operated steering wheel 11, a steering column 12 connected to the steering wheel 11 and transmitting a steering force to the wheels of the vehicle, a steering angle sensor 13, a steering torque sensor 14, and a vehicle condition measurement module 15.

[0039] The steering angle sensor 13 measures a steering angle θ steering and a steering angle velocity θ̇ wsteering of the steering wheel 11. That is, the steering angle information D received from the steering angle sensor 13 steering includes the steering angle θ steering and the steering angle velocity θ̇ wsteering .

[0040] The steering torque sensor 14 measures a steering torque T steering More specifically, the steering torque sensor 14 can measure a torque acting on the steering column 12, which depends on the operation of the steering wheel 11.

[0041] The steering angle information D measured by the steering angle sensor 13 steering , ie the steering angle θ steering and the steering angle velocity θ̇ wsteering , and the steering torque T measured by the steering torque sensor 14 steering can be defined as the process data related to the steering wheel operation by the driver.

[0042] The vehicle condition measurement module 15 comprises one or more sensors that measure information about the condition of the vehicle. The information D measured by the vehicle condition measurement module 15 vehicle can reach a speed V x of the vehicle, a lateral acceleration a y of the vehicle, a yaw rate θ̇ z of the vehicle and a wheel speed Vwheel of the vehicle. The wheel speed V wheel of the vehicle for each individual wheel of the vehicle.

[0043] The information from the vehicle condition measurement module D vehicle can be defined as the status data that indicates the condition of the vehicle.

[0044] The steering angle θ measured by the steering angle sensor 13 steering and the steering angle velocity θ̇ wsteering , the steering torque T measured by the steering torque sensor 14 steering and the speed V measured by the vehicle condition measuring module 15 x of the vehicle, the lateral acceleration a y , the yaw rate θ̇ z and the wheel speed V wheel are transmitted via a Controller Area Network (CAN) 16 to an electronic control unit (ECU) 17.

[0045] The ECU 17 receives the data transmitted via CAN 16 and performs calculations to generate a command for a motor 18, which generates additional torque to assist the steering. The command is transmitted to the motor 18. In this case, the command can be supplied to the motor 18 in the form of current. In this case, the ECU 17 can perform feedback control for the motor 18.

[0046] A torque input via the steering wheel 11 and a torque generated by the motor 18 are transmitted to a steering mechanism 19 connected to the wheels of the vehicle. The steering mechanism 19 is driven by the torque input via the steering wheel 11 and the torque generated by the motor 18. When the steering mechanism 19 is driven, the direction of the vehicle's wheels is adjusted and the vehicle is steered.

[0047] Fig. 2 is a diagram showing the configuration of an apparatus for detecting an abnormality of a motor of a column EPS according to an embodiment of the present disclosure. Further, Fig. 3 is a diagram illustrating models used to obtain an estimated value or detect an abnormality in the column EPS motor abnormality detection apparatus according to the embodiment of the present disclosure.

[0048] A column EPS motor abnormality detection device 100 according to the embodiment of the present disclosure detects an abnormality of a motor 18 that generates auxiliary steering force in the column EPS of the vehicle. Specifically, the column EPS motor abnormality detection device 100 can detect a power loss of the motor 18 or an abnormality of the motor 18.

[0049] The column EPS motor abnormality detection device 100 according to an embodiment of the present disclosure can detect a deterioration of the motor 18 before a failure of the motor 18 occurs by using an artificial intelligence-based digital twin algorithm for a vehicle. The digital twin algorithm can virtually construct the column EPS system of the vehicle.

[0050] For example, the column EPS motor abnormality detection device 100 according to the embodiment of the present disclosure may be installed in the ECU 17 of the Fig. 1. Of course, this is an example, and the column EPS motor abnormality detection device 100 may be located elsewhere.

[0051] Referring to Fig. 2, the column EPS motor abnormality detection apparatus 100 according to the embodiment of the present disclosure may include a memory 110 and a processor 120.

[0052] Memory 110 stores one or more instructions. The one or more instructions can configure a program for detecting an abnormality in the column EPS motor. The one or more instructions stored in memory 110 can be executed by processor 120.

[0053] Memory 110 may include a hardware device configured to store the one or more instructions. Memory 110 may, for example, comprise a storage medium such as read-only memory (ROM), random access memory (RAM), flash memory, etc. Furthermore, memory 110 may comprise a magnetic medium such as a floppy disk or magnetic tape, an optical medium such as a compact disc read-only memory (CD-ROM) or a digital video disc (DVD), a magneto-optical medium such as a floppy disk, etc.

[0054] Processor 120 executes the one or more instructions. For example, processor 120 may be a hardware unit that performs calculations and control. Processor 120 may include at least one arithmetic logic unit (ALU) and one register.

[0055] The processor 120 executes the one or more instructions to input operation data related to steering wheel operation by a driver of the vehicle and state data indicative of a state of the vehicle into an artificial neural network model 20, and obtains an estimated output value of the motor 18 output from the artificial neural network model 20.

[0056] The process data can contain at least one steering angle θ steering , a steering angle velocity θ̇ steering and a steering torque T steering include.

[0057] As described above, the steering angle θ steering and the steering angle velocity θ ̇steering by the steering angle sensor 13 and the steering torque T steering measured by the steering torque sensor 14.

[0058] The status data can contain at least one speed V x of the vehicle, a lateral acceleration a y, a yaw rate θ̇ z and a wheel speed V wheel of the vehicle. If the vehicle has multiple wheels, the wheel speed can be determined for each of the multiple wheels.

[0059] All of the above-described process and status data can be retrieved via a CAN 200. In other words, the process data and status data can be formed from signals that can be retrieved via the vehicle's CAN 200. Accordingly, according to the present disclosure, a motor abnormality can be efficiently detected without the use of an additional sensor related to detecting a motor abnormality of the EPS column.

[0060] The memory 110 can be connected directly or indirectly to the vehicle's CAN 200. The memory 110 can receive the process data and the status data from the CAN 200. Accordingly, the processor 120 can obtain the process data and the status data from the memory 110.

[0061] The artificial neural network model 20 receives the operation data and the state data and outputs the estimated output value of the motor 18. In one embodiment of the present disclosure, the artificial neural network model 20 may include a generative adversarial network (GAN).

[0062] In one embodiment of the present disclosure, the artificial neural network model 20 may be a deep learning-based neural twin model for the vehicle's pillar EPS system. In other words, the artificial neural network model 20 may serve as a virtual twin model of the vehicle's EPS system.

[0063] Fig. Figure 4 is a diagram showing a detailed configuration of the artificial neural network model.

[0064] Referring to Fig. 4, the artificial neural network model 20 may comprise a generator 21 and a discriminator 22.

[0065] The artificial neural network model 20 can be constructed by the generator 21 and the discriminator 22, which learn alternately. In this case, the data used for learning can be obtained when the motor 18 of the column EPS system 10 is in a normal state. Accordingly, the estimated value output by the artificial neural network model 20 can follow an output measurement of the motor obtained when the motor 18 is in a normal state.

[0066] The generator 21 receives the process data and the state data and generates the estimated output value of the motor 18. The generator 21 may include a neural network. The generator 21 may be configured as a multivariate transformer. That is, the artificial neural network model 20 may be a GAN based on a multivariate transformer.

[0067] The generator 21 can calculate the operation data and the state data and output the estimated output value of the motor 18 of the column EPS system. The output value estimated by the generator 21 follows an initial value of the motor output to correspond to the operation data and the state data when the column EPS system is in a normal state.

[0068] Fig. Figure 5 is a diagram showing a detailed configuration of the generator of the artificial neural network model.

[0069] As in Fig. 5, the generator 21 may comprise an encoder 21a and a decoder 21b.

[0070] Encoder 21a and decoder 21b form a network to obtain an estimated output value of the motor's power. In other words, generator 21 may be formed from a structured network of encoder 21a and decoder 21b, which is continuously trained to estimate the estimated value.

[0071] The encoder 21a may be designed to extract high-level features from the process data and the state data using a transformer block and a linear operation.

[0072] In this case, the transformer block may contain an embedding layer, a self-attention layer, and a feedforward layer.

[0073] First, the embedding layer in the transformer block can encode spatio-temporal information in input sequences using a trainable embedding matrix.

[0074] Next, deep features can be extracted by the self-attention layer, and refinement can be done using the feedforward layer.

[0075] In this case, the self-attention layer can use a scaled dot-product attention mechanism. The self-attention layer can be a key component in transformer-based deep neural network (DNN) architectures, enabling the model to process and understand dependencies within the input sequence.

[0076] The scaled dot product attention can first apply layer normalization to the input sequence and process the layer normalization results through linear operations into three unique vectors: a query (Q), a key (K), and a value (V).

[0077] An attention value AS can then be calculated for each position in the input sequence based on the derived vectors. An example calculation of the attention value AS can be given as follows: Formula 1. AS=σ(Q⋅K‖xinput‖dim)⋅V (σ() denotes a softmax operation, ||x input || dim denotes a feature dimension of an input, and - denotes a dot product)

[0078] An attention score quantifies the relative importance of each position in the input sequence relevant to a particular query. This means that higher values ​​are assigned to more appropriate positions.

[0079] An output of the transformer block can be fed to two different linear operations to calculate a query, a key vector and an initial input value of the decoder 21b.

[0080] The decoder 21b may estimate the estimated value using the obtained values ​​in an autoregressive manner, similar to sequence-by-sequence recurrent neural networks.

[0081] Like the encoder 21a, the decoder 21b may also include a transformer block and linear operations. Furthermore, the decoder 21b may include a cross-attention layer within the transformer block.

[0082] The cross-attention layer may perform scaled dot-product attention using the query and key vector of the encoder 21a to learn a dynamic correlation between input signals and estimated signals.

[0083] The discriminator 22 receives measurement data including the operation data, the state data and the output measurement value of the motor and outputs a discrimination value for the measurement data.

[0084] Discriminator 22 is a classifier designed to determine conditional probability. Discriminator 22 determines the probability that an example is genuine or fake, given a set of input features. The discrimination values ​​of discriminator 22 are classification labels. The discrimination value of discriminator 22 is, for example, 0 or 1.

[0085] Meanwhile, the discriminator 22 may additionally receive estimated data containing the process data, the state data, and the estimated output value of the motor, and additionally output a discrimination value for the estimated data. Such a process may be performed when the generator 21 and the discriminator 22 perform a learning process.

[0086] In this way, the discriminator 22 can output the discrimination value for the measurement data during an anomaly detection operation, while the discriminator 22 can output the discrimination value for the estimation data during a learning operation.

[0087] The discriminator 22 can be configured as a discriminator network used for adversarial training of a GAN. In other words, the generator 21 and the discriminator 22 can be trained using an adversarial training method to improve estimation performance and create a GAN model.

[0088] More specifically, the discriminator 22 can perform an optimization alternately with the generator 21 to solve a Wasserstein min-max problem such as formula 2. minG maxD EXD,Real[D(XD,Real)]−EXD,Fake[D(XD,Fake)] (x D,Real denotes a real data set consisting of the process data, the state data and the output measurement value of the engine, and x D,Fake refers to a virtual data sample consisting of the process data, the state data, and the output estimate of the engine.)

[0089] In the artificial neural network model 20, the generator 21 is trained to deceive the discriminator 22, which uses the spatial-temporal features of x D,Real and x D,Fake Accordingly, the generator 21 can generate the estimated value to follow the performance of the column EPS system motor in a normal state.

[0090] In this context, a loss function for GAN training can be defined as follows: Formula 3. LG=1N∑|yG−y^G|+0.001×1N∑[−D(XD,Fake)], LD=1N∑[D(XD,Fake)−D(XD,Real)]+10×1N∑[(‖∇XD,FakeD(XD,Fake)‖2−1)2] (N denotes the size of a batch, L G denotes the loss function of the generator and L D the loss function of the discriminator)

[0091] The first term of a loss of the generator 21 corresponds to a loss L MAEa mean absolute error of the supervised training. The remaining term of the loss of the generator 21 together with a loss of the discriminator 22 forms a loss function L GAN of the GAN.

[0092] In one embodiment of the present disclosure, the following may be considered to represent the dynamic properties of the column EPS system in a loss function.

[0093] First, the power of the engine in the column EPS system can be calculated according to formula 4. Tmotor=TLast−TSteering (T motor denotes the power of the engine, T Load denotes a force acting on a tire and T Steering refers to a steering force of the driver)

[0094] Meanwhile, T Load in Formula 4 as a function of an input signal of the ECU by the Dugoff tire model. In particular, T Loadin Formula 4 as a function shown in Formula 5 below. TLoad=f(ϕECU)=(δ1sinθswn+δ2cosθswn)+δ3 (O ECU denotes the input signal of the vehicle's ECU, θ sw denotes a steering angle of the vehicle and n denotes a given value) In formula 5, δ1 and δ2 represent the model-related uncertainty and δ3 the uncertainty of the system. Meanwhile, the input signal Ø ECU a signal that the vehicle's ECU can receive or measure via the CAN. The motor output value can be calculated by subtracting a driver's steering force (torque) from T Load which is obtained by the function in formula 5.

[0095] The dynamic properties of the column-EPS system can be defined using Equation 4 and Equation 5. The dynamic properties of the column-EPS system can be considered as a physically based loss function of the artificial neural network model 20.

[0096] Through the method described above, the processor 120 can obtain the estimated output value of the motor of the column EPS system.

[0097] Furthermore, the processor 120 compares the estimated value obtained in this way with a real output measured value of the engine. The processor 120 can compare the estimated value with the measured value and detect an anomaly of the engine 18.

[0098] The measured value can be the command generated by the ECU 17, which is shown above in Fig. 1. The value estimated using the artificial neural network model 20 can be considered the actual output of the motor 18. Therefore, the processor 120 can detect an abnormality of the motor 18 by comparing the estimated value with the measured value.

[0099] Processor 120 may execute one or more instructions to input error data related to a difference between the estimated value and the measured value into an anomaly detection model 30 and determine whether the column EPS motor is abnormal. Anomaly detection model 30 may, for example, use a one-class support vector machine (OCSVM) algorithm.

[0100] There may be a plurality of data sets containing the process data, the state data, the measured value, and the estimated value. The error data may include a mean and standard deviation of the errors between the measured value and the estimated value obtained from each of the plurality of data sets, a maximum absolute error between the measured value and the estimated value of the plurality of data sets, and a discrimination value of a discriminator for the measured data of the plurality of data sets.

[0101] As described above, the measurement data can include the process data, the state data and the measured value.

[0102] The multiple data sets containing the measured value and the estimated value can be treated as a batch. For example, 100 to 150 data sets (as a specific example, 128 data sets) can be treated as a batch, and the error data can be determined for each batch.

[0103] The failure data obtained for a batch can be input to the anomaly detection model 30. Furthermore, the degree of the engine's anomaly (engine power degradation) can be determined based on an output (characteristic) of the anomaly detection model 30 that receives the failure data.

[0104] An F1 score according to formula 6 can be considered as a metric for anomaly detection. F1=2×TP2×TP+FP+FN (TP stands for true positive, FP for false positive and FN for false negative, where positive indicates a deterioration in engine performance)

[0105] As described above, the discriminator 22 can calculate not only the discrimination value for the measurement data contained in the data set, but also the discrimination value for the estimated data contained in the data set. The estimated data can include the process data, the state data, and the estimated value. The discrimination value of the discriminator 22 for the estimated data of the data set can be provided as feedback to the generator 21 of the artificial neural network model 20 and used for training.

[0106] So far, the configuration of the column EPS motor abnormality detection device 100 according to the embodiment of the present disclosure has been described in detail. The operation of the column EPS motor abnormality detection device 100 will be described below.

[0107] Fig. 6 is a diagram showing an operation of the column EPS motor abnormality detecting apparatus according to the embodiment of the present disclosure.

[0108] Referring to Fig. 6, the column EPS motor abnormality detection apparatus 100 according to the embodiment of the present disclosure may operate as follows.

[0109] First, the memory 110 stores an input signal Ø received from the vehicle's EPS system ECU and an output measurement value F m of the column EPS motor.

[0110] The input signal includes Ø ECU Operation data and state data. In one embodiment of the present disclosure, the input signal Ø ECUa steering angle, a steering angular velocity, a steering torque, a speed of the vehicle, a lateral acceleration of the vehicle, a yaw rate of the vehicle and a wheel speed of the vehicle.

[0111] Meanwhile, a variety of input signals Ø ECU and a variety of output measurements F m of the column EPS motor as a batch. That is, in one embodiment of the present disclosure, the data processing and calculations may be performed in units of data batches.

[0112] Next, the processor 120 executes one or more instructions to convert the input signal Ø ECU into the artificial neural network model 20 and an estimated output value F eof the column EPS motor output by the artificial neural network model 20. More specifically, the processor 120 may receive the input signal Ø ECU into the generator 21 of the artificial neural network model 20 and the estimated output value F e of the column EPS motor output from the generator 21.

[0113] Furthermore, the processor 120 executes one or more instructions to convert the input signal Ø ECU and the output measured value F m of the column EPS motor into the artificial neural network model 20 and obtain a discrimination value DS. More specifically, the processor 120 can process measurement data including the input signal Ø ECU and the output measured value F mof the column EPS motor into the discriminator 22 of the artificial neural network model 20 and obtain the discrimination value DS for the measurement data output by the discriminator 22.

[0114] The processor 120 then outputs the error data D Error The error data D Error can be output for each data batch.

[0115] As described above, there are a variety of data sets, including the operation data, the state data, the measured value and the estimated value, and the error data D Error can have an average value µ E and a standard deviation σ E of errors between the measured value and the estimated value obtained from each of the plurality of data sets, a maximum absolute error Max Ebetween the measured value and the estimated value of the plurality of data sets and the discrimination value DS of the discriminator 22 for the measured data obtained from the plurality of data sets.

[0116] Finally, the processor 120 may execute the one or more instructions to obtain the error data D Error to input the anomaly detection model 30 and obtain an output of the anomaly detection model 30. As described above, the anomaly detection model 30 may include an OCSVM algorithm.

[0117] Based on the output of the network model 30, which contains the error data D Error receives, the processor 120 can detect the extent of the abnormality (deterioration of motor performance) of the column EPS motor.

[0118] As described above, the column EPS motor abnormality detection apparatus 100 according to the embodiment of the present disclosure has been described in detail. A method for detecting an abnormality of a column EPS motor will be described below.

[0119] Fig. 7 is a flowchart of a method for determining an abnormality of a motor of a column EPS according to an embodiment of the present disclosure.

[0120] In a column EPS motor abnormality detection method S100 according to the exemplary embodiment of the present disclosure, an abnormality of the motor 18 disposed in the column EPS system of the vehicle is detected. More specifically, in the column EPS motor abnormality detection method S100, an abnormality of the motor disposed in the column EPS system of the vehicle to provide auxiliary steering force to a steering column may be detected.

[0121] Referring to Fig. 7, the process S100 for detecting the abnormality of the motor of the column EPS according to the embodiment of the present disclosure may be performed as follows.

[0122] First, the processor 120 inputs operation data related to the driver's steering wheel operation and state data indicating a state of the vehicle into the artificial neural network model 20 and obtains an estimated output value of the engine output from the artificial neural network model 20 (S110).

[0123] The process data may include at least one of a steering angle, a steering angular velocity, or a steering torque. Furthermore, the state data may include at least one of a vehicle speed, a vehicle lateral acceleration, a vehicle yaw rate, and a vehicle wheel speed.

[0124] For example, the operation data may include a steering angle, a steering angular velocity, and a steering torque, and the state data may include a speed of the vehicle, a lateral acceleration of the vehicle, a yaw rate of the vehicle, and a wheel speed of the vehicle.

[0125] In one embodiment of the present disclosure, the artificial neural network model 20 may comprise a GAN. In other words, the artificial neural network model 20 may be a deep learning-based neural twin model for the vehicle's pillar EPS system.

[0126] The artificial neural network model 20 may include a generator 21 and a discriminator 22. The generator 21 receives the process data and the state data and generates the estimated value. The generator 21 may be configured as a multivariate transformer. Furthermore, the discriminator 22 receives measurement data, including the process data, the state data, and the measured value, and outputs a discrimination value for the measurement data.

[0127] In the process S110 of determining the estimated value of the engine, the processor 120 may input the process data and the state data to the generator 21 and obtain the estimated value generated by the generator 21.

[0128] Meanwhile, the artificial neural network model 20 is constructed by the generator 21 and the discriminator 22, which performs learning. The data used during learning can be obtained when the motor 18 of the column EPS system 10 is in a normal state. Accordingly, the estimated value can follow the motor's output measurement value obtained when the motor 18 is in a normal state.

[0129] The processor 120 then compares the estimated value with the measured value and detects an anomaly of the motor 18 (S120).

[0130] Processor 120 may execute one or more instructions to input error data related to a difference between the estimated value and the measured value to anomaly detection model 30 and determine whether engine 18 is abnormal. Anomaly detection model 30 may, for example, use an OCSVM algorithm.

[0131] Fig. 8 is a detailed flowchart of an abnormality detection process of the column EPS motor abnormality detection method according to the embodiment of the present disclosure.

[0132] Referring to Fig. 8, the engine abnormality detection process S120 can be performed as follows.

[0133] First, the processor 120 inputs measurement data including the operation data, the state data, and the output measurement value of the engine to the discriminator 22 and obtains a discrimination value for the measurement data generated by the discriminator 22 (S121).

[0134] Next, the processor 120 inputs error data including the discrimination value for the measured data and numerical values ​​related to a difference between the estimated value and the measured value to the anomaly detection model 30 and obtains an output of the anomaly detection model 30 (S122).

[0135] There may be a plurality of data sets containing the input data, the measured value, and the estimated value. The error data may include a mean and standard deviation of the errors between the measured value and the estimated value obtained from each of the plurality of data sets, a maximum absolute error between the measured value and the estimated value of the plurality of data sets, and a discrimination value of the discriminator for the measured data of the plurality of data sets.

[0136] More specifically, the plurality of data records containing the measured value and the estimated value can be treated as a batch. For example, 100 to 150 data records (as a specific example, 128 data records) can be treated as a batch, and the error data can be determined for each batch.

[0137] The failure data obtained for a batch can be input to the anomaly detection model 30. Furthermore, the degree of the engine's anomaly (engine power degradation) can be determined based on an output (characteristic) of the anomaly detection model 30 that receives the failure data.

[0138] The F1 score described above can be used as a metric for detecting anomalies.

[0139] Meanwhile, the present disclosure additionally provides a non-transitory computer-readable storage medium storing a program for executing the column EPS motor abnormality detection method. More specifically, the present disclosure may provide a non-transitory computer-readable storage medium storing a program including at least one instruction for executing the column EPS motor abnormality detection method.

[0140] In this case, the instruction may include not only the machine code generated by a compiler, but also the high-level language code executable by a computer.

[0141] The non-transitory, computer-readable storage medium may be a magnetic medium such as a hard disk, a floppy disk, or a magnetic tape, an optical medium such as a CD-ROM or a DVD, a magneto-optical medium such as a floppy disk, or a hardware device configured to store and execute program instructions, such as a ROM, a RAM, a flash memory, etc.

[0142] As a result of a simulation, it was determined that, according to the present disclosure, a decrease of up to 10% in the maximum power of the EPS system motor can be predicted with an F1 value of 0.85 or higher. In other words, according to the present disclosure, the degree of motor degradation can be accurately predicted before the motor in the vehicle's column EPS system completely fails.

[0143] According to the present disclosure, it is possible to predict an abnormality of the motor of the column EPS system in advance, before a failure of the motor of the column EPS system occurs. This makes it possible to take proactive measures before a motor failure occurs and prevent the motor failure and the resulting risks.

[0144] In this way, the present disclosure enables the estimation of the system's future operation and remaining service life, as well as prognostic diagnostics suitable for predictive maintenance applications. This allows proactive measures and maintenance to be effectively controlled before a failure of the vehicle's column EPS motor occurs.

[0145] According to the above configuration, in the apparatus and method for detecting the abnormality of the motor of the column EPS system and the non-transitory computer-readable storage medium storing a program for performing the method according to one aspect of the present disclosure, it is possible to predict deterioration in the performance of the motor before a failure of the motor arranged in the column EPS system occurs based on an artificial intelligence-based digital twin algorithm.

[0146] Furthermore, in the apparatus and method for detecting the abnormality of the motor of the column EPS system and the non-transitory computer-readable storage medium storing a program for executing the method according to one aspect of the present disclosure, it is possible to predict deterioration in the performance of the motor of the column EPS using signals of a CAN of the vehicle without a separate additional sensor.

[0147] It should be understood that the effects of the present disclosure are not limited to the effects described above, and include all effects that can be derived from a configuration of the invention described in detailed descriptions or claims of the present disclosure.

[0148] Although embodiments of the present disclosure have been described, the spirit of the present disclosure is not limited to the embodiments set forth in the description. A person skilled in the art who understands the spirit of the present disclosure can easily propose other embodiments by supplementing, changing, removing, or adding components within the scope of this spirit, but these embodiments are also within the spirit of the present disclosure. QUOTES CONTAINED IN THE DESCRIPTION

[0000] This list of documents submitted by the applicant was generated automatically and is included solely for the convenience of the reader. This list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions. Cited patent literature

[0000] KR 10-2023-0106181

[0001] KR 10-2024-0021747

[0001] KR 2098049

[0007]

Claims

[1] Device (100) comprising: a memory (110) configured to store one or more instructions; and a processor (120) configured to detect an anomaly of a motor (18) of an electric column power steering system (EPS) configured to provide auxiliary steering force to a steering column (12) of the vehicle, and to execute the one or more instructions comprising: Inputting operation data relating to the steering wheel operation by a driver of the vehicle and state data indicating a state of the vehicle into an artificial neural network model (20) to obtain at least one estimated value of an output of the engine (18) from the artificial neural network model (20), Comparing the estimated value of the output of the motor (18) estimated by the artificial neural network model (20) with at least one measured value of the output of the motor (18) detected by a sensor, and Detecting the abnormality of the motor (18) of the EPS based on the comparison result between the estimated value of the output of the motor (18) estimated by the artificial neural network model (20) and the measured value of the output of the motor (18) detected by the sensor. [2] The device (100) according to claim 1, wherein the operation data relating to the steering wheel operation by the driver includes at least a steering angle, a steering angular velocity and a steering torque. [3] The apparatus (100) according to any one of claims 1 or 2, wherein the state data indicative of the state of the vehicle includes at least one of a speed of the vehicle, a lateral acceleration of the vehicle, a yaw rate of the vehicle, and a wheel speed of the vehicle. [4] Device (100) according to one of claims 1 to 3, wherein the process data and the status data comprise signals available via a Controller Area Network, CAN, (16; 200) of the vehicle. [5] The device (100) according to any one of claims 1 to 4, wherein the artificial neural network model (20) includes a generative adversarial network (GAN) having a generator (21) configured to receive the operation data relating to the steering wheel operation by the driver and the state data indicating the state of the vehicle and to generate the estimated value of the output of the engine (18). [6] The apparatus (100) of claim 5, wherein the artificial neural network model (20) further includes a discriminator (22) configured to receive measurement data including the operation data, the state data, and the measured value of the output of the motor (18) and to output a discrimination value for the measurement data. [7] The apparatus (100) according to claim 6, wherein the processor (120) is configured to input error data relating to a difference between the estimated value of the output of the motor (18) estimated by the artificial neural network model (20) and the measured value of the output of the motor (18) detected by the sensor into an anomaly detection model (30) to detect the anomaly of the motor (18). [8] The apparatus (100) of claim 7, wherein the anomaly detection model (30) uses a one-class support vector machine (OCSVM) algorithm. [9] Device (100) according to one of claims 7 or 8, wherein: the at least one estimated value of the output of the engine (18) estimated by the artificial neural network model (20) comprises a plurality of estimated values, the at least one measured value of the output of the engine (18) detected by the sensor comprises a plurality of measured values, and a plurality of data sets comprises the process data, the state data, the plurality of measured values, and the plurality of estimated values; and the error data input to the anomaly detection model (30) includes a mean and a standard deviation of errors between the plurality of measured values ​​and the plurality of estimated values ​​obtained from the plurality of data sets, a maximum absolute error among the errors between the plurality of measured values ​​and the plurality of estimated values ​​of the plurality of data sets, and a discrimination value of the discriminator (22) for the measured data. [10] The device (100) according to any one of claims 6 to 9, wherein the discriminator (22) is configured to additionally receive estimation data including the operation data relating to the steering wheel operation, the state data indicating the state of the vehicle, and the estimated value of the output of the engine (18) estimated by the artificial neural network model (20), and additionally outputs a discrimination value for the estimation data. [11] Device (100) according to one of claims 6 to 10, wherein: the artificial neural network model (20) comprises the generator (21) and the discriminator (22) which alternately carry out the learning, and the artificial neural network model (20) is configured to perform the learning using the process data, the state data and the measured value obtained when the engine (18) is in a normal state. [12] A method for detecting an abnormality of a motor (18) of an electric column power steering system (EPS) arranged to provide an auxiliary steering force for a steering column (12) of a vehicle, the method comprising: Obtaining, by a processor (120), at least one estimated value of an output of the engine (18) from the artificial neural network model (20) by inputting operation data relating to the steering wheel operation by a driver of the vehicle and state data indicating a state of the vehicle into the artificial neural network model (20); and Detecting the anomaly of the motor (18) by the processor (120) by comparing the estimated value of the output of the motor (18) estimated by the artificial neural network model (20) with at least one measured value of the output of the motor (18) detected by a sensor. [13] The method according to claim 12, wherein the operation data relating to the driver's steering wheel operation includes at least one of a steering angle, a steering angular velocity and a steering torque. [14] The method according to any one of claims 12 or 13, wherein the state data indicative of the state of the vehicle includes at least one of a speed of the vehicle, a lateral acceleration of the vehicle, a yaw rate of the vehicle, and a wheel speed of the vehicle. [15] The method of any one of claims 12 to 14, wherein the artificial neural network model (20) comprises a generative adversarial network (GAN) including: a generator (21) configured to receive the operation data relating to the steering wheel operation by the driver and the status data indicating the status of the vehicle and to generate the estimated value of the output of the motor (18), and a discriminator (22) configured to receive measurement data including the operation data, the state data and the measured value of the output of the motor (18), and to output a discrimination value for the measurement data. [16] The method of claim 15, wherein obtaining the at least one estimated value of the output of the motor (18) from the artificial neural network model (20) comprises inputting the operation data and the state data to the generator (21) and obtaining the estimated value of the output of the motor (18) generated by the generator (21). [17] A method according to any one of claims 15 or 16, wherein detecting the abnormality of the engine (18) comprises: Inputting, by the processor (120), measurement data containing the process data, the state data and the measured value of the output of the motor (18) into the discriminator (22) and receiving, by the processor (120), the discrimination value for the measurement data generated by the discriminator (22); and Inputting, by the processor (120), error data including the discrimination value for the measurement data and a value related to a difference between the estimated value and the measured value into an anomaly detection model (30) and receiving, by the processor (120), an output of the anomaly detection model (30). [18] The method of claim 17, wherein: the at least one estimated value of the output of the engine (18) estimated by the artificial neural network model (20) comprises a plurality of estimated values, the at least one measured value of the output of the engine (18) detected by the sensor comprises a plurality of measured values, and a plurality of data sets comprises the process data, the state data, the plurality of estimated values, and the plurality of measured values; and the error data input to the anomaly detection model (30) includes a mean and a standard deviation of errors between the plurality of measured values ​​and the plurality of estimated values ​​obtained from the plurality of data sets, a maximum absolute error among the errors between the plurality of measured values ​​and the plurality of estimated values ​​of the plurality of data sets, and a discrimination value of the discriminator (22) for the measured data. [19] A method according to any one of claims 15 to 18, wherein: the artificial neural network model (20) comprises the generator (21) and the discriminator (22) which alternately carry out the learning, and the artificial neural network model (20) is configured to perform the learning using the process data, the state data and the measured value obtained when the engine (18) is in a normal state. [20] A non-transitory computer-readable medium configured to store at least one instruction that, when executed by a processor (120), causes the processor (120) to perform operations for detecting an abnormality of a motor (18) of an electric column power steering (EPS) system configured to provide auxiliary steering force to a steering column (12) of a vehicle, the operations comprising: Obtaining at least one estimated value of an output of the engine (18) from the artificial neural network model (20) by inputting operation data relating to the steering wheel operation by a driver of the vehicle and state data indicating a state of the vehicle into the artificial neural network model (20); and Detecting the abnormality of the motor (18) by comparing the estimated value of the output of the motor (18) estimated by the artificial neural network model (20) with at least one measured value of the output of the motor (18) detected by a sensor.

Images