Abnormality detection device and method for hydraulic generation motor and computer readable storage medium

Abstract

The invention discloses a hydraulic generation motor anomaly detection device and method and a computer readable storage medium. An abnormality detection device for a hydraulic pressure generation motor, according to one aspect of the present invention, detects an abnormality of a hydraulic pressure generation motor that generates a hydraulic pressure for executing braking in accordance with an input displacement of a brake pedal of a vehicle by driving a pump, comprises: a memory for storing one or more commands; and a processor that executes the one or more instructions. The processor inputs input data to an artificial neural network model by executing the one or more instructions and acquires an estimated value of an output variable, and detects whether an abnormality has occurred in the hydraulic generating motor using a difference between the estimated value of the output variable and a measured value of the output variable, the artificial neural network model receives the input data and outputs an estimated value of the output variable related to an output value of the hydraulic generating motor.

Description

Technical Field

[0001] The present invention relates to a hydraulic generating motor abnormality detection device, method and a computer-readable storage medium storing a program for executing the method, and more specifically, to a hydraulic generating motor abnormality detection device, a hydraulic generating motor abnormality detection method and a non-temporary computer-readable storage medium storing a program for executing the hydraulic generating motor abnormality detection method, wherein the hydraulic generating motor generates hydraulic pressure by driving a pump, and the hydraulic pressure is used to perform braking corresponding to the input displacement of a brake pedal of a vehicle. Background Art

[0002] A braking system must be provided on a vehicle for braking. Recently, in order to improve the stability and efficiency of braking, an integrated braking (Integrated Dynamic Brake, IDB) system has been proposed.

[0003] The integrated brake system includes: a pedal displacement sensor for detecting the input depth (displacement) of the brake pedal; a motor that operates by a control signal generated based on the detection information of the pedal displacement sensor; and a pump having a slave cylinder that generates the hydraulic pressure required in braking by converting the rotational force of the motor into linear motion.

[0004] When a motor in the integrated brake system as described above fails, a braking force corresponding to the input depth of the brake pedal cannot be generated. As a result, a fatal problem in terms of vehicle safety may occur.

[0005] Generally speaking, the output torque of a hydraulic pressure generating motor may decrease due to demagnetization and corrosion of heat insulating materials. The output torque of such a motor shows a tendency to gradually decrease. Therefore, in order to improve the safety of the braking system, it is preferable to predict (prognostics) the abnormality of the motor that generates the hydraulic pressure for braking in advance. However, conventional methods related to abnormality detection of the motor that generates the hydraulic pressure required for braking still remain at the post-fault diagnosis and subsequent measures.

[0006] Prior art literature

[0007] Patent Literature

[0008] Korean Patent No. 2257873 "Pump for integrated brake device", granted on May 24, 2021 Summary of the invention

[0009] Problem that the invention aims to solve

[0010] The present invention aims to solve the above-mentioned problems, and the purpose of the present invention is to provide a hydraulic generating motor abnormality detection device, a hydraulic generating motor abnormality detection method and a non-temporary computer-readable storage medium storing a program for executing the hydraulic generating motor abnormality detection method, which can detect the performance degradation of the hydraulic generating motor used to drive the pump before a failure occurs to generate hydraulic pressure for performing braking corresponding to the input displacement of the vehicle brake pedal.

[0011] Another object of the present invention is to provide a hydraulic generating motor abnormality detection device, a hydraulic generating motor abnormality detection method and a non-temporary computer-readable storage medium storing a program for executing the hydraulic generating motor abnormality detection method, which utilizes the controller area network (CAN, Controller Area Network) signal in the vehicle to effectively predict in advance the reduction in the output torque of the hydraulic generating motor that generates hydraulic pressure for braking.

[0012] Problems of the present invention are not limited to the above-mentioned problems, and other unmentioned problems will be clearly understood by those skilled in the art through the following description.

[0013] Means used to solve problems

[0014] According to one aspect of the present invention, there is provided a hydraulic generating motor abnormality detection device, which detects abnormality of a hydraulic generating motor, wherein the hydraulic generating motor generates hydraulic pressure by driving a pump, and the hydraulic pressure is used to perform braking corresponding to the input displacement of a brake pedal of a vehicle, wherein the device comprises: a memory for storing one or more instructions; and a processor for executing the one or more instructions. The processor inputs input data into an artificial neural network model by executing the one or more instructions and obtains an estimated value of the output variable, and detects whether the hydraulic generating motor is abnormal by using the difference between the estimated value of the output variable and the measured value of the output variable, wherein the artificial neural network model receives the input data and outputs an estimated value of the output variable related to the output value of the hydraulic generating motor.

[0015] According to the hydraulic pressure generating motor abnormality detection device of one aspect of the present invention, the input data may include any one or more of the speed of the vehicle, the lateral acceleration of the vehicle, the input displacement of the brake pedal, or the wheel speed of the vehicle.

[0016] According to the hydraulic pressure generating motor abnormality detection device of one aspect of the present invention, the output variable may include a fluid pressure generated by the pump.

[0017] According to the hydraulic pressure generating motor abnormality detection device of one aspect of the present invention, the input data may be acquired through a controller area network (CAN) of the vehicle.

[0018] According to a hydraulic generating motor abnormality detection device according to one aspect of the present invention, the artificial neural network model can be composed of a generative adversarial network (GAN), which includes: a generator, which receives the input data and generates an estimated value of the output variable; and a discriminator, which receives the input data and the measured value of the output variable and outputs a measured related identification value.

[0019] According to the hydraulic pressure generating motor abnormality detection device of one aspect of the present invention, the discriminator may further receive the input data and the estimated value of the output variable and output an estimation-related discriminant value.

[0020] According to a hydraulic generating motor abnormality detection device according to one aspect of the present invention, the artificial neural network model is constructed by alternately learning the generator and the discriminator with each other, and the input data and the measured values ​​of the output variables used in the learning may be data obtained when the vehicle and the hydraulic generating motor are in normal conditions.

[0021] According to the hydraulic pressure generating motor abnormality detection device according to one aspect of the present invention, the generator may be composed of a multivariate transformer.

[0022] According to a hydraulic generating motor abnormality detection device according to one aspect of the present invention, the processor can input error data related to the difference between the estimated value of the output variable and the measured value of the output variable into the abnormality detection module and determine whether the hydraulic generating motor has an abnormality by executing one or more instructions.

[0023] According to the hydraulic pressure generating motor abnormality detection device of one aspect of the present invention, the abnormality detection module may use a One-Class Support Vector Machine (OCSVM) algorithm.

[0024] According to a hydraulic generating motor abnormality detection device according to one aspect of the present invention, there are multiple data sets, the data sets include the input data, the measured values ​​of the output variables and the estimated values ​​of the output variables, and the error data may include: the average value and standard deviation of the error between the measured values ​​of the output variables obtained from each of the multiple data sets and the estimated values ​​of the output variables; the maximum absolute error between the measured values ​​of the output variables obtained from each of the multiple data sets and the estimated values ​​of the output variables; and the discrimination value of the discriminator for the input data and the measured values ​​of the output variables included in the multiple data sets.

[0025] According to another aspect of the present invention, a method for detecting abnormality of a hydraulic generating motor is provided, wherein the hydraulic generating motor generates hydraulic pressure by driving a pump, and the hydraulic pressure is used to perform braking corresponding to the input displacement of a brake pedal of a vehicle, wherein the method comprises: a step in which a processor inputs input data into an artificial neural network model by executing one or more instructions and obtains an estimated value of an output variable, the artificial neural network model receives the input data and outputs an estimated value of the output variable related to the output value of the hydraulic generating motor; and a step in which the processor detects whether an abnormality occurs in the hydraulic generating motor by using the difference between the estimated value of the output variable and the actual measured value of the output variable.

[0026] According to the hydraulic pressure generating motor abnormality detection method according to one aspect of the present invention, the input data may include any one or more of the vehicle speed, the vehicle lateral acceleration, the brake pedal input displacement, or the vehicle wheel speed.

[0027] According to the hydraulic pressure generating motor abnormality detection method according to one aspect of the present invention, the output variable may include a fluid pressure generated by the pump.

[0028] According to a hydraulic generating motor abnormality detection method according to one aspect of the present invention, the artificial neural network model can be composed of a generative adversarial network (GAN), which includes: a generator, which receives the input data and generates an estimated value of the output variable; and a discriminator, which receives the input data and the measured value of the output variable and outputs a measured related identification value.

[0029] According to a method for detecting an abnormality of a hydraulic generating motor according to one aspect of the present invention, the artificial neural network model is constructed by alternately learning the generator and the discriminator with each other, and the data used in the learning may be data obtained when the vehicle and the hydraulic generating motor are in a normal state.

[0030] According to a hydraulic generating motor abnormality detection method according to one aspect of the present invention, the step of detecting whether an abnormality occurs in the hydraulic generating motor may include: a step in which the processor inputs the input data and the measured value of the output variable into the discriminator and obtains the measured related identification value generated by the discriminator; and a step in which the processor inputs the measured related identification value and error data including a numerical value related to the difference between the estimated value of the output variable and the measured value of the output variable into an abnormality detection module and obtains the output of the abnormality detection module.

[0031] According to the hydraulic pressure generating motor abnormality detection method according to one aspect of the present invention, the abnormality detection module may use a One-Class Support Vector Machine (OCSVM) algorithm.

[0032] According to a hydraulic generating motor abnormality detection method according to one aspect of the present invention, there are multiple data sets, the data sets include the input data, the measured values ​​of the output variables and the estimated values ​​of the output variables, and the error data may include: the average value and standard deviation of the error between the measured values ​​of the output variables obtained from each of the multiple data sets and the estimated values ​​of the output variables; the maximum absolute error between the measured values ​​of the output variables obtained from each of the multiple data sets and the estimated values ​​of the output variables; and the discrimination value of the discriminator for the input data and the measured values ​​of the output variables included in the multiple data sets.

[0033] According to still another aspect of the present invention, there is provided a non-transitory computer-readable storage medium storing a program, the program including at least one instruction for executing the hydraulic pressure generating motor abnormality detection method.

[0034] Effects of the Invention

[0035] According to the above configuration, a hydraulic generating motor abnormality detection device, method and a non-temporary computer-readable storage medium storing a program for executing the method according to one aspect of the present invention are based on an artificial intelligence-based digital twin algorithm to predict in advance the performance degradation of the hydraulic generating motor used to drive a pump before a failure occurs, wherein the pump generates hydraulic pressure for performing braking corresponding to the input displacement of a vehicle brake pedal.

[0036] Furthermore, according to an aspect of the present invention, a hydraulic generating motor abnormality detection device, method, and a non-temporary computer-readable storage medium storing a program for executing the method can utilize the vehicle's controller area network (CAN) signal to detect performance degradation of the hydraulic generating motor without any additional separate sensor.

[0037] The effects of the present invention are not limited to the above-mentioned effects, and should be understood to include all effects that can be inferred from the structure of the present invention described in the detailed description of the present invention or the claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 is a diagram schematically showing a configuration of an integrated brake system of a vehicle.

[0039] Figure 2 is a diagram showing a hydraulic pressure generating motor abnormality detection device according to one embodiment of the present invention.

[0040] Figure 3 : is a diagram showing a model for acquiring an estimated value or for detecting abnormality in a hydraulic pressure generating motor abnormality detection device according to one embodiment of the present invention.

[0041] Figure 4 is a diagram showing a detailed configuration of an artificial neural network model.

[0042] Figure 5 is a diagram showing a detailed configuration of a generator of an artificial neural network model.

[0043] Figure 6 1 is a diagram showing the operation of the hydraulic pressure generating motor abnormality detection device according to one embodiment of the present invention.

[0044] Figure 7 is a sequence diagram of a method for detecting abnormality of a hydraulic pressure generating motor according to an embodiment of the present invention.

[0045] Figure 8 The present invention is a detailed sequence diagram of steps for detecting whether an abnormality occurs in a hydraulic pressure generating motor according to a method for detecting an abnormality of a hydraulic pressure generating motor according to an embodiment of the present invention.

[0046] Description of Reference Numerals

[0047] 100: Hydraulic generator motor abnormality detection device

[0048] 110: Memory

[0049] 120: Processor DETAILED DESCRIPTION

[0050] Below, embodiments of the present invention will be described in detail in conjunction with the accompanying drawings so that those skilled in the art can easily implement these embodiments. The present invention can be implemented in a variety of different forms and is not limited to the embodiments described herein. In order to clearly explain the present invention, parts that are not related to the description are omitted in the accompanying drawings, and the same or similar parts are represented by the same reference numerals throughout the specification.

[0051] The words and terms used in this specification and claims should not be interpreted as limited to the commonly used meanings or meanings in dictionaries, but should be interpreted as having meanings and concepts consistent with the technical scope of the present invention, based on the principle that inventors can define terms and concepts in order to explain their own inventions in the best manner.

[0052] In this specification, terms such as "including" or "having" should be understood as describing the existence of features, numbers, steps, actions, constituent elements, parts or their combinations recorded in the specification, without excluding the existence or additional possibilities of one or more other features or numbers, steps, actions, constituent elements, parts or their combinations in advance.

[0053] Figure 1 is a diagram schematically showing a configuration of an integrated brake system of a vehicle.

[0054] The integrated braking system 200 is configured in a vehicle and performs braking of the vehicle. The integrated braking system 200 is configured to drive a pump 270 by an electric control signal generated according to the input depth (displacement) of the detected brake pedal 210, thereby supplying hydraulic pressure required for braking to a brake 280 configured in each wheel of the vehicle. The integrated braking system 200 provides fast braking response performance and also easily implements a by-wire function of the vehicle.

[0055] Reference Figure 1 The integrated braking system 200 may include a brake pedal 210 , a pedal simulator 220 , a pedal displacement sensor 230 , a vehicle sensor unit 240 , an electronic control unit 250 , a hydraulic pressure generating motor 260 , a pump 270 , a brake 280 , and a pressure sensor 290 .

[0056] When the driver's pedal force is input to the brake pedal 210, the brake pedal 210 is displaced to one side. The pedal simulator 220 provides a reaction force to the brake pedal 210. Thus, the driver can feel an appropriate pedal feeling.

[0057] The pedal displacement sensor 230 detects an input depth of the brake pedal 210 , that is, detects an input displacement of the brake pedal 210 . The input displacement of the brake pedal detected by the pedal displacement sensor 230 may be transmitted to the electronic control unit 250 .

[0058] The vehicle sensor unit 240 may detect one or more information related to the state of the vehicle. For example, the vehicle sensor unit 240 may include a vehicle speed sensor 241, a wheel speed sensor 242, and an acceleration sensor 243. The speed of the vehicle detected by the vehicle speed sensor 241, the wheel speed detected by the wheel speed sensor 242, and the lateral acceleration of the vehicle measured by the acceleration sensor 243 may be transmitted to the electronic control unit 250.

[0059] The electronic control unit 250 generates a control signal for controlling the hydraulic pressure generating motor 260 based on one or more input data. At this time, the one or more input data may include any one or more of the input displacement of the brake pedal, the speed of the vehicle, the wheel speed, or the lateral acceleration of the vehicle.

[0060] The hydraulic pressure generating motor 260 receives the control signal from the electronic control unit 250 and is driven in response to the control signal. The rotational driving force of the hydraulic pressure generating motor 260 is transmitted to the pump 270.

[0061] The pump 270 has a driving mechanism that converts the rotational force of the hydraulic pressure generating motor 260 into a linear motion. The pump 270 generates hydraulic pressure so that the brake hydraulic pressure corresponding to the operation of the hydraulic pressure generating motor 260 is supplied to the brake 280 arranged on the wheel side of the vehicle.

[0062] The brake 280 performs braking by hydraulic pressure supplied according to the operation of the pump 270. In this regard, a hydraulic circuit for supplying hydraulic pressure may be disposed between the pump 270 and the brake 280.

[0063] The pressure sensor 290 is used to detect the fluid pressure generated by the pump 270. The pressure sensor 290 may be disposed inside the pump 270 or in the hydraulic circuit. The fluid pressure detected by the pressure sensor 290 may be transmitted to the electronic control unit 250 as feedback.

[0064] Figure 2 1 is a diagram showing a hydraulic pressure generating motor abnormality detection device according to an embodiment of the present invention. Figure 3 : is a diagram showing a model for acquiring an estimated value or for detecting abnormality in a hydraulic pressure generating motor abnormality detection device according to one embodiment of the present invention.

[0065] The hydraulic pressure generating motor abnormality detection device 100 according to one embodiment of the present invention is used to detect abnormality of the hydraulic pressure generating motor 260, which generates brake hydraulic pressure corresponding to the input displacement of the vehicle brake pedal 210 by driving the pump 270. More specifically, the hydraulic pressure generating motor abnormality detection device 100 can detect a decrease in the output torque of the hydraulic pressure generating motor 260 before the hydraulic pressure generating motor 260 is in a fault state.

[0066] According to an embodiment of the present invention, the hydraulic generating motor abnormality detection device 100 can predict the function reduction of the hydraulic generating motor 260 before the hydraulic generating motor 260 fails by using an artificial intelligence-based digital twin algorithm. The digital twin algorithm can virtually construct an integrated braking system of the vehicle.

[0067] Reference Figure 2 , a hydraulic pressure generating motor abnormality detection device 100 according to one embodiment of the present invention may include a memory 110 and a processor 120 .

[0068] The memory 110 stores one or more instructions that can be executed by the processor 120 .

[0069] The memory 110 may include a hardware device configured to store and execute program instructions. For example, the memory 110 may include storage media such as a read-only memory (ROM), a random access memory (RAM), a flash memory, etc. In addition, the memory 110 may also include magnetic media such as floppy disks and magnetic tapes, optical recording media such as compact disk read only memory (CD-ROM, Compact Disk Read Only Memory), digital video disks (DVD, Digital Video Disk), and magneto-optical media such as floptical disks (Magneto-Optical Media), etc.

[0070] The processor 120 executes the one or more instructions. For example, the processor 120 may be a hardware unit that performs calculations and controls in a computer. The processor 120 may include at least one arithmetic logic unit (ALU) and a register.

[0071] The processor 120 inputs the input data into the artificial neural network model 20 and obtains the estimated value of the output variable by executing the one or more instructions, wherein the artificial neural network model 20 receives the input data and outputs the estimated value of the output variable related to the output value of the hydraulic generating motor 260.

[0072] The input data may include any one or more of the speed of the vehicle, the lateral acceleration of the vehicle, the input displacement of the brake pedal, or the wheel speed of the vehicle. For example, the input data may include the speed of the vehicle, the lateral acceleration of the vehicle, the input displacement of the brake pedal, and the wheel speed of the vehicle. At this time, the wheel speed of the vehicle may be acquired for each of a plurality of wheels of the vehicle.

[0073] The input data may be obtained from a Controller Area Network (CAN) 300 of the vehicle. For example, the input data may be transmitted to the memory 110 via the CAN 300 .

[0074] As described above, the input data may be acquired through the vehicle's controller area network 300. Therefore, according to the present invention, there is no need to use an additional sensor to detect abnormality of the hydraulic pressure generating motor.

[0075] In this regard, the memory 110 may be directly or indirectly connected to the controller area network 300 of the vehicle. The memory 110 may receive the input data from the controller area network 300. Thus, the processor 120 may obtain the input data from the memory 110.

[0076] The output variable may include the fluid pressure generated by the pump 270. In one embodiment of the present invention, the output variable may be the fluid pressure.

[0077] The artificial neural network model 20 receives the input data and outputs an estimated value of an output variable related to an output value of the hydraulic pressure generating motor 260. More specifically, the artificial neural network model 20 may receive the input data and output an estimated value of the fluid pressure generated by the pump.

[0078] The estimated value of the output variable is obtained for comparison with the measured value of the output variable. At this time, the measured value of the output variable can be obtained by a pressure sensor configured in the pump or the hydraulic circuit. That is, the measured value of the output variable can be the fluid pressure detected by the pressure sensor.

[0079] In one embodiment of the present invention, the artificial neural network model 20 may be composed of a generative adversarial network (GAN).

[0080] In one embodiment of the present invention, the artificial neural network model 20 may be a deep learning-based neural twin model for the vehicle's integrated braking system 200. In other words, the artificial neural network model 20 may function as a virtual twin model of the integrated braking system.

[0081] Figure 4 is a diagram showing a detailed configuration of an artificial neural network model.

[0082] Reference Figure 4 , the artificial neural network model 20 may include a generator 21 and a discriminator 22.

[0083] The artificial neural network model 20 can be constructed by alternately performing learning with each other by the generator 21 and the discriminator 22. At this time, the data used when performing the learning can be data obtained when the vehicle and the integrated braking system are in a normal state. Thus, the estimated value of the output variable can track the measured value of the output variable obtained when the vehicle and the integrated braking system are in a normal state.

[0084] The generator 21 receives the input data and generates an estimated value of the output variable. The generator 21 may include a neural network. The generator 21 may be composed of a multivariate transformer. That is, the artificial neural network model 20 may be a generative adversarial network (GAN) based on a multivariate transformer.

[0085] In one embodiment of the present invention, the generator 21 may receive the input data. As described above, the input data may include any one or more of the vehicle speed, the vehicle lateral acceleration, the brake pedal input displacement, or the vehicle wheel speed.

[0086] The generator 21 may calculate the input data and output an estimated value of the output variable. In this case, the output variable may be the fluid pressure generated by the pump. In one embodiment of the present invention, when the integrated braking system of the vehicle is in a normal state, the estimated value of the output variable estimated by the generator 21 tracks the measured value of the output variable output in response to the input data.

[0087] Figure 5 is a diagram showing a detailed configuration of a generator of an artificial neural network model.

[0088] Reference Figure 5, the generator 21 may include an encoder 21a and a decoder 21b.

[0089] The encoder 21a and the decoder 21b constitute a network for obtaining the estimated value of the output variable. In other words, the generator 21 can be composed of a network of an encoder 21a-decoder 21b structure trained in an end-to-end manner to obtain the estimated value of the output variable.

[0090] The encoder 21a may be designed to extract high-level features from input data using a transformer block and linear operations. At this time, the transformer block may include an embedding layer, a self-attention layer, and a feed-forward layer.

[0091] First, the embedding layer in the Transformer block can use a trainable embedding matrix to encode spatiotemporal information into the input sequences.

[0092] Next, deep features can be extracted through a self-attention layer, and refinement can be performed using a feed-forward layer.

[0093] At this time, the self-attention layer can utilize a scaled dot-product attention mechanism. The self-attention layer can be a major component in the Transformer-based Deep Neural Network (DNN) architecture, which enables the model to process and understand dependencies in the input sequence.

[0094] The scaled inner product attention first applies layer normalization to the input sequence and processes the result into three intrinsic vectors of query (Q) vector, key (K) vector and value (V) vector through linear operation.

[0095] Next, an attention score AS for each position in the input sequence may be calculated based on the derived vector. The calculation formula of the attention score AS may be as follows:

[0096] [Formula 1]

[0097]

[0098] Among them, σ(·) represents the softmax operation, ∥x input ∥ dim represents the feature level of the input, and (·) represents the inner product.

[0099] The attention score quantifies the relative importance of each position in the input sequence with respect to a given query. That is, higher scores are assigned to more appropriate positions.

[0100] The output of the transformer block can be fed into two different linear operations to compute the query, key vector and initial input value for decoder 21b

[0101] The decoder 21b may use the acquired values ​​to estimate one or more estimated values ​​in an autoregressive manner similar to a Sequence-to-Sequence Recurrent Neural Networks.

[0102] Similar to the encoder 21a, the decoder 21b may include a transformer block and a linear operation. In addition, the decoder 21b may also include a cross-attention layer within the transformer block.

[0103] The cross-attention layer can use the query vector and key vector of encoder 21a to perform scaled inner product attention to learn the dynamic correlation between the input signal and the estimated signal.

[0104] The discriminator 22 receives the input data and the measured value of the output variable and outputs a discriminant value. The discriminant value output by the discriminator 22 for the input data and the measured value of the output variable may be defined as a measured-related discriminant value.

[0105] The discriminator 22 is a classifier whose purpose is to determine conditional probabilities. The discriminator 22 determines the probability that an example is true or false based on input features. The discriminator 22's discriminator value may be a classification label. For example, the discriminator 22's discriminator value may be 0 or 1.

[0106] On the other hand, the discriminator 22 can also receive the input data and the estimated value of the output variable and output the discrimination value. The discrimination value of the discriminator 22 output for the input data and the estimated value of the output variable can be defined as the estimation-related discrimination value. This process can be implemented when the generator 21 and the discriminator 22 are learning.

[0107] As described above, the discriminator 22 outputs the discrimination value for the input data and the measured value of the output variable during the operation for detecting abnormality. On the other hand, the discriminator 22 may output the discrimination value for the input data and the estimated value of the output variable during the operation for learning.

[0108] The discriminator 22 may be formed by a discriminator network used in adversarial training of a generative adversarial network (GAN). In other words, the generator 21 and the discriminator 22 may be trained by adversarial training to improve estimation performance and establish a generative adversarial network model.

[0109] More specifically, the discriminator 22 may perform optimization alternately with the generator 21 to solve a Wasserstein min-max problem such as Equation 2.

[0110] [Formula 2]

[0111]

[0112] Among them, x D,Real represents the actual data set consisting of input data and measured values, x D,Fake represents a virtual data sample consisting of the input data and the estimated values.

[0113] In the artificial neural network model 20, the generator 21 is trained to deceive the D,Real and x D,Fake The generator 21 can thereby generate an estimate of the integrated braking system for tracking a normal state.

[0114] In this regard, the loss function for generating adversarial network training can be defined as the following formula 3.

[0115] [Formula 3]

[0116]

[0117] Where N is the batch size, L G is the loss function of the generator, L D is the loss function of the discriminator.

[0118] The first term of the loss of generator 21 corresponds to the mean absolute error loss L for supervised training MAE The remaining items of the loss of the generator 21 together with the loss of the discriminator 22 constitute the generative adversarial network loss function L GAN .

[0119] Furthermore, the processor 120 detects whether the hydraulic pressure generating motor is abnormal by using the difference between the estimated value of the output variable and the measured value of the output variable. As described above, the output variable may be the fluid pressure generated by the pump that operates according to the operation of the hydraulic pressure generating motor. In other words, the processor 120 may detect whether the hydraulic pressure generating motor is abnormal by comparing the estimated value of the fluid pressure with the measured value. At this time, the measured value may be measured by the pressure sensor 290.

[0120] The processor 120 can input the error data related to the difference between the estimated value of the output variable and the measured value of the output variable to the abnormality detection module 30 by executing the one or more instructions, and determine whether the hydraulic generating motor is abnormal. For example, the abnormality detection model 30 can use a one-class support vector machine (OCSVM) algorithm.

[0121] There may be a plurality of data sets, the data sets including the input data, the measured values ​​of the output variables, and the estimated values ​​of the output variables. The error data may include the mean value and standard deviation of the error between the measured values ​​of the output variables and the estimated values ​​of the output variables obtained from each of the plurality of data sets, the maximum absolute error between the measured values ​​of the output variables and the estimated values ​​of the output variables obtained from each of the plurality of data sets, and the discriminator value output for the input data and the measured values ​​of the output variables included in the plurality of data sets.

[0122] A plurality of data sets including the input data, the estimated values ​​of the output variables, and the measured values ​​of the output variables may be considered as a batch. For example, 100 to 150 data sets (as a specific example, 128 data sets) may be considered as a batch, and the error data of each batch may be obtained.

[0123] The error data obtained for one batch may be input to the abnormality detection module 30. And, the abnormality level of the hydraulic pressure generating motor (performance degradation of the hydraulic pressure generating motor) may be detected based on the output (feature) obtained from the abnormality detection module 30 that received the error data.

[0124] Regarding the anomaly detection metric, the F1 score given by the following formula 4 can be considered.

[0125] [Formula 4]

[0126]

[0127] Among them, TP is true positive, FP is false positive, FN is false negative, and positive indicates that the performance of the hydraulic generating motor is reduced.

[0128] On the other hand, as described above, the discriminator 22 can calculate not only the discriminant value associated with the input data included in the data set and the measured value of the output variable, but also the discriminant value associated with the estimated value of the input data included in the data set and the output variable. The discriminant value of the discriminator 22 for the estimated value of the data set can be provided as feedback to the generator 21 of the artificial neural network model 20, thereby being used in training.

[0129] So far, the configuration of the hydraulic pressure generating motor abnormality detection device 100 according to one embodiment of the present invention has been described in detail. Next, the operation of the hydraulic pressure generating motor abnormality detection device 100 will be described in detail.

[0130] Figure 6 1 is a diagram showing the operation of the hydraulic pressure generating motor abnormality detection device according to one embodiment of the present invention.

[0131] Reference Figure 6 , the hydraulic generating motor abnormality detection device 100 according to an embodiment of the present invention can perform the following operations.

[0132] First, the memory 110 stores input data and the measured value Pm of the output variable. At this time, the output variable may be the fluid pressure generated by the pump of the integrated brake system. That is, the measured value of the output variable may include the measured value of the fluid pressure.

[0133] As mentioned above, input data The input data may include any one or more of the vehicle speed, the vehicle lateral acceleration, the brake pedal input displacement, or the vehicle wheel speed. A plurality of the measured data of the measured value Pm can be regarded as a batch. That is, in one embodiment of the present invention, the data batch can be used as a unit for data processing and calculation.

[0134] Next, the processor 120 converts the input data into The processor 120 can input the input data into the artificial neural network model 20, and obtain the estimated value Pe of the output variable output by the artificial neural network model 20. More specifically, the processor 120 can The output is input to the generator 21 of the artificial neural network model 20, and the estimated value of the fluid pressure output by the generator 21 is obtained.

[0135] Furthermore, the processor 120 converts the input data into and the measured value Pm of the output variable are input to the artificial neural network model 20, and the identification value DS is obtained. More specifically, the processor 120 can input the data The measured value of the fluid pressure is input to the discriminator 22, and a discriminator value DS is obtained.

[0136] Next, the processor 120 outputs the error data D Error The error data D can be output according to each data batch Error .

[0137] More specifically, there may be a plurality of data sets, the plurality of data sets including the input data, the estimated value Pe of the output variable, and the measured value Pm of the output variable.

[0138] At this time, the error data D Error It may include: the average value μ of the error between the estimated value Pe obtained from each of the plurality of data sets and the measured value Pm E and standard deviation σ E The maximum absolute error Max between the estimated value Pe obtained from each of the plurality of said data sets and the measured value Pm E ; and for input data and the discrimination value DS of the discriminator 22 of the measured value Pm.

[0139] Finally, the processor 120 converts the error data D into Error The anomaly detection model 30 is inputted, and the output of the anomaly detection model 30 is obtained. As described above, the anomaly detection model 30 may include a one-class support vector machine (OCSVM) algorithm.

[0140] Based on the received error data D Error Based on the output (features) obtained by the abnormality detection module 30, the processor 120 can detect the abnormality (performance reduction of the hydraulic generating motor) level of the hydraulic generating motor.

[0141] The above is a detailed description of the hydraulic pressure generating motor abnormality detection device 100 according to one embodiment of the present invention. Next, a hydraulic pressure generating motor abnormality detection method will be described in detail.

[0142] Figure 7 is a sequence diagram of a method for detecting abnormality of a hydraulic pressure generating motor according to an embodiment of the present invention.

[0143] According to an embodiment of the present invention, a hydraulic pressure generating motor abnormality detection method (step S100) detects abnormality of a hydraulic pressure generating motor for driving a pump to generate hydraulic pressure for performing braking corresponding to an input displacement of a brake pedal of a vehicle.

[0144] Reference Figure 7 , the hydraulic pressure generating motor abnormality detection method (step S100) according to one embodiment of the present invention may be performed as follows.

[0145] First, the processor 120 inputs input data into the artificial neural network model and obtains the estimated value of the output variable by executing one or more instructions. The artificial neural network model receives the input data and outputs the estimated value of the output variable related to the output value of the hydraulic generating motor (step S110).

[0146] The input data may include any one or more of the speed of the vehicle, the lateral acceleration of the vehicle, the input displacement of the brake pedal, or the wheel speed of the vehicle. In one embodiment of the present invention, the input data may include the speed of the vehicle, the lateral acceleration of the vehicle, the input displacement of the brake pedal, and the wheel speed of the vehicle. At this time, the wheel speed of the vehicle may be obtained for each of the multiple wheels of the vehicle.

[0147] On the other hand, the input data may be obtained from a controller area network (CAN) 300 of the vehicle.

[0148] The output variable may be the pressure of the fluid generated by the pump. In other words, the estimated value of the output variable may be an estimated value of the fluid pressure.

[0149] The estimated value of the output variable is obtained for comparison with the measured value of the output variable. In this case, the measured value of the output variable may be the measured value of the fluid pressure generated by the pump.

[0150] In one embodiment of the present invention, the artificial neural network model 20 may include a generative adversarial network (GAN). In other words, the artificial neural network model 20 may be a deep learning-based neural twin model for an integrated braking system of a vehicle.

[0151] The artificial neural network model 20 may include a generator 21 and a discriminator 22. The generator 21 receives the input data and generates an estimated value of the output variable. The generator 21 may be composed of a multivariate transformer. Furthermore, the discriminator 22 may receive the input data and the measured value of the output variable and output a measured related discriminant value.

[0152] In the step of acquiring the estimated value of the output variable (step S110 ), the processor 120 may input the input data into the generator 21 , and acquire the estimated value of the output variable generated by the generator 21 .

[0153] On the other hand, the artificial neural network model 20 is constructed by learning the generator 21 and the discriminator 22, and the input data and the measured values ​​used in the learning can be obtained when the vehicle and the integrated braking system are in a normal state. Therefore, the estimated value of the output variable can track the measured value of the output variable obtained when the integrated braking system is in a normal state.

[0154] Next, the processor 120 detects whether an abnormality occurs in the hydraulic pressure generating motor by using the difference between the estimated value of the output variable and the measured value of the output variable (step S120 ).

[0155] The processor 120 inputs the error data related to the difference between the estimated value and the measured value into the abnormality detection module 30 by executing one or more instructions, and determines whether the hydraulic generating motor has an abnormality. For example, the abnormality detection model 30 can use a one-class support vector machine (OCSVM) algorithm.

[0156] Figure 8 The present invention is a detailed sequence diagram of steps for detecting whether an abnormality occurs in a hydraulic pressure generating motor according to a method for detecting an abnormality of a hydraulic pressure generating motor according to an embodiment of the present invention.

[0157] Reference Figure 8 , the step of detecting whether an abnormality occurs in the hydraulic pressure generating motor (step S120) can be performed as follows.

[0158] First, the processor 120 inputs the input data and the measured value of the output variable to the discriminator 22, and obtains the measured correlation discriminant value generated by the discriminator 22 (step S121).

[0159] The identification value output by the identification device 22 for the input data and the measured value of the output variable may be defined as a measured correlation identification value.

[0160] Next, the processor 120 inputs the measured relevant identification value and the error data including the numerical value related to the difference between the estimated value of the output variable and the measured value of the output variable into the abnormality detection module 30, and obtains the output of the abnormality detection module 30 (step S122).

[0161] There may be a plurality of data sets, the plurality of data sets including the input data, the measured values ​​of the output variables and the estimated values ​​of the output variables. The error data may include: the mean value and standard deviation of the error between the measured values ​​of the output variables and the estimated values ​​of the output variables obtained from each of the plurality of data sets; the maximum absolute error between the measured values ​​of the output variables and the estimated values ​​of the output variables obtained from each of the plurality of data sets; and the discriminator value for the input data and the measured values ​​of the output variables included in the plurality of data sets.

[0162] In one embodiment of the present invention, the error data may include: the average value and standard deviation of the error between the estimated value and the measured value of the fluid pressure obtained from each of the multiple data sets; the maximum absolute error between the estimated value and the measured value of the fluid pressure obtained from each of the multiple data sets; and the identification value of the identifier 22 for the input data and the measured value of the fluid pressure included in the multiple data sets.

[0163] A plurality of data sets including the input data, the estimated values ​​of the output variables, and the measured values ​​of the output variables may be considered as a batch. For example, 100 to 150 data sets (as a specific example, 128 data sets) may be considered as a batch, and the error data of each batch may be obtained.

[0164] The error data obtained for one batch may be input to the abnormality detection module 30. And, the abnormality level (performance degradation of the hydraulic generating motor) of the hydraulic generating motor may be detected based on the output (feature) obtained from the abnormality detection module 30 that received the error data.

[0165] On the other hand, the present invention also provides a non-transitory computer-readable storage medium storing a program for executing the hydraulic generating motor abnormality detection method. Specifically, the present invention can provide a non-transitory computer-readable storage medium storing a program including at least one instruction for executing the hydraulic generating motor abnormality detection method.

[0166] At this time, the instruction may include a machine code generated by a compiler, and the instruction may also include a high-level language code executable by a computer.

[0167] The storage medium may include a hardware device configured to store and execute program instructions such as magnetic media (Magnetic Media) such as hard disks, floppy disks and magnetic tapes, optical recording media (Optical Media) such as compact disk read only memory (CD-ROM), digital video disk (Digital Video Disk, DVD), magneto-optical media (Magneto-Optical Media) such as floptical disks (Floptical Disk), read-only memory (ROM), random access memory (RAM), flash memory, etc.

[0168] As for an anomaly detection metric showing the effect of the present invention, the F1 score as described above may be considered.

[0169] As a result of executing the simulation, according to the present invention, it is possible to predict the performance of the F1 score of 0.85 or more until the hydraulic pressure generating motor is reduced to 10% relative to the maximum output. In other words, according to the present invention, before the hydraulic pressure generating motor for driving the pump for generating the brake hydraulic pressure corresponding to the input displacement of the vehicle brake pedal completely fails, it is possible to accurately predict the degradation level of the hydraulic pressure generating motor.

[0170] According to the present invention, a reduction in the output torque of the hydraulic pressure generating motor can be detected sensitively. As a result, preventive measures can be taken before the hydraulic pressure generating motor fails, thereby effectively preventing the hydraulic pressure generating motor from failing.

[0171] As described above, the present invention provides an estimate of the future behavior and remaining useful life of the system and predictive diagnostics suitable for predictive maintenance applications. Thus, preventive measures and maintenance can be effectively prompted before the hydraulic generating motor fails.

[0172] Although the embodiments of the present invention have been described, the spirit of the present invention is not limited to the embodiments described in the specification, and those skilled in the art who understand the spirit of the present invention will be able to easily propose other embodiments within the scope of the same spirit by adding, changing, deleting or adding components. However, this is also included in the spirit of the present invention.

Claims

1. A hydraulic pressure generating motor abnormality detection device, which detects abnormality of a hydraulic pressure generating motor, wherein the hydraulic pressure generating motor generates hydraulic pressure by driving a pump, and the hydraulic pressure is used to perform braking corresponding to an input displacement of a brake pedal of a vehicle, characterized in that: include: memory for storing one or more instructions, and A processor, executing the one or more instructions; The processor inputs the input data into the artificial neural network model and obtains the estimated value of the output variable by executing the one or more instructions, and uses the difference between the estimated value of the output variable and the measured value of the output variable to detect whether the hydraulic generating motor has an abnormality. The artificial neural network model receives the input data and outputs the estimated value of the output variable related to the output value of the hydraulic generating motor.

2. The hydraulic pressure generating motor abnormality detection device according to claim 1, wherein: The input data includes one or more of the speed of the vehicle, the lateral acceleration of the vehicle, the input displacement of the brake pedal, or the wheel speed of the vehicle.

3. The hydraulic pressure generating motor abnormality detection device according to claim 1, wherein: The output variable includes a fluid pressure generated by the pump.

4. The hydraulic pressure generating motor abnormality detection device according to claim 1, wherein: The input data is obtained via a controller area network of the vehicle.

5. The hydraulic pressure generating motor abnormality detection device according to claim 1, wherein: The artificial neural network model is composed of a generative adversarial network. The generative adversarial network comprises: a generator that receives the input data and generates an estimate of the output variable; and A discriminator receives the input data and the measured value of the output variable and outputs a measured related discriminant value.

6. The hydraulic pressure generating motor abnormality detection device according to claim 5, wherein: The discriminator also receives the input data and an estimated value of the output variable and outputs an estimated associated discriminant value.

7. The hydraulic pressure generating motor abnormality detection device according to claim 5, wherein: The processor inputs error data related to a difference between an estimated value of the output variable and an actual value of the output variable into an abnormality detection module by executing the one or more instructions and determines whether an abnormality occurs in the hydraulic pressure generating motor.

8. The hydraulic pressure generating motor abnormality detection device according to claim 7, wherein: There are a plurality of data sets, each of which includes the input data, measured values ​​of the output variables, and estimated values ​​of the output variables. The error data includes: a mean and a standard deviation of errors between measured values ​​of the output variable and estimated values ​​of the output variable obtained from each of a plurality of the data sets; a maximum absolute error between a measured value of the output variable and an estimated value of the output variable obtained from each of the plurality of the data sets; and The discriminator has discriminant values ​​for the input data and measured values ​​of the output variables included in the plurality of data sets.

9. A method for detecting abnormality of a hydraulic pressure generating motor, which detects abnormality of a hydraulic pressure generating motor, wherein the hydraulic pressure generating motor generates hydraulic pressure by driving a pump, and the hydraulic pressure is used to perform braking corresponding to an input displacement of a brake pedal of a vehicle, characterized in that: include: The processor executes one or more instructions to input input data into an artificial neural network model and obtain an estimated value of an output variable, wherein the artificial neural network model receives the input data and outputs an estimated value of the output variable related to the output value of the hydraulic generating motor; as well as The processor detects whether an abnormality occurs in the hydraulic pressure generating motor by using a difference between the estimated value of the output variable and the measured value of the output variable.

10. The method for detecting abnormality of a hydraulic pressure generating motor according to claim 9, wherein: The input data includes one or more of the speed of the vehicle, the lateral acceleration of the vehicle, the input displacement of the brake pedal, or the wheel speed of the vehicle.

11. The method for detecting abnormality of a hydraulic pressure generating motor according to claim 9, wherein: The output variable includes a fluid pressure generated by the pump.

12. The method for detecting abnormality of a hydraulic pressure generating motor according to claim 9, wherein: The artificial neural network model is composed of a generative adversarial network. The generative adversarial network comprises: a generator that receives the input data and generates an estimate of the output variable; and A discriminator receives the input data and the measured value of the output variable and outputs a measured related discriminant value.

13. The method for detecting abnormality of a hydraulic pressure generating motor according to claim 12, wherein: The step of detecting whether the hydraulic pressure generating motor is abnormal comprises: The processor inputs the input data and the measured value of the output variable into the discriminator and obtains the measured related discriminant value generated by the discriminator; and The processor inputs the measured correlation identification value and error data including a numerical value related to the difference between the estimated value of the output variable and the measured value of the output variable into an abnormality detection module and obtains an output of the abnormality detection module.

14. The method for detecting abnormality of a hydraulic pressure generating motor according to claim 13, wherein: There are a plurality of data sets, each of which includes the input data, measured values ​​of the output variables, and estimated values ​​of the output variables. The error data includes: a mean and a standard deviation of errors between measured values ​​of the output variable and estimated values ​​of the output variable obtained from each of a plurality of the data sets; a maximum absolute error between a measured value of the output variable and an estimated value of the output variable obtained from each of a plurality of the data sets; and The discriminator outputs a discriminator value for the input data and the measured values ​​of the output variable included in the plurality of data sets.

15. A computer-readable storage medium storing a program, wherein: The program includes at least one instruction for executing the hydraulic pressure generating motor abnormality detection method according to claim 9 .

Images