Solenoid valve anomaly detection device and method and computer readable storage medium

Abstract

The present invention provides a solenoid valve abnormality detection apparatus and method, and a computer readable storage medium, the solenoid valve abnormality detection apparatus according to an embodiment of the present invention for detecting an abnormality of a solenoid valve of an electronically controlled suspension system (ECS) disposed in a vehicle, the solenoid valve abnormality detection apparatus comprising: a memory for storing one or more instructions; and a processor executing the one or more instructions, the processor inputting input data representing a state of the electrically controlled suspension system to an artificial neural network model by executing the one or more instructions, and a detection unit that acquires an estimated value indicating a physical quantity output by the electrically controlled suspension system and output by the artificial neural network model, and detects whether the solenoid valve is abnormal or not by comparing the estimated value with a measured value of the physical quantity.

Description

Technical Field

[0001] The present invention relates to an electromagnetic valve abnormality detection device and method, and a computer-readable storage medium. More specifically, it relates to an electromagnetic valve abnormality detection device and method for an electronically controlled suspension (ECS) system, and a non-temporary computer-readable storage medium storing a program for executing the method. The electromagnetic valve abnormality detection device and method are configured in the electronically controlled suspension system of a vehicle and can detect the performance degradation of the electromagnetic valve as an actuating element for adjusting the damping force. Background Art

[0002] The Electronically Controlled Suspension (ECS) system automatically controls the suspension according to the road surface and driving conditions of the vehicle. The electronically controlled suspension system can ensure both the safety and ride quality of the vehicle.

[0003] The electronically controlled suspension system includes a solenoid valve. The solenoid valve of the electronically controlled suspension system is an element that semi-actively drives the output (damping force) of the electronically controlled suspension system. If the solenoid valve of the electronically controlled suspension system fails, it may cause a heterogeneous ride feel.

[0004] Therefore, in the case of a vehicle equipped with an electronically controlled suspension system, in order to maintain safety and ride quality, it is necessary to predict (prognostic) the output drop of the solenoid valve in advance, rather than detecting the output drop of the solenoid valve configured in the electronically controlled suspension system afterwards. However, in the existing approach related to the fail-safety of the solenoid valve of the electronically controlled suspension system, the focus is more on the follow-up measures after the failure occurs, rather than on preventing the failure in advance. In addition, the existing electronically controlled suspension system uses the fail-safety based on the solenoid valve current, but there is a limitation in the lack of direct response to the output drop.

[0005] Prior Art Literature

[0006] Patent Literature

[0007] Korean authorized invention patent No. 2429549 "Electronic control suspension device and damping force control method thereof", authorization announcement date is 2022.08.01 Summary of the invention

[0008] Problem that the invention aims to solve

[0009] The present invention is used to solve the above-mentioned problem. The purpose of the present invention is to provide a solenoid valve abnormality detection device and method for an electronically controlled suspension system and a non-temporary computer-readable storage medium storing a program for executing the method, which is configured in the electronically controlled suspension (ECS) system of a vehicle and can predict in advance the performance degradation of the solenoid valve that adjusts the output of the electronically controlled suspension system.

[0010] Another object of the present invention is to provide a solenoid valve abnormality detection device and method for an electronically controlled suspension system and a non-temporary computer-readable storage medium storing a program for executing the method, which can utilize the controller area network (CAN) signal in the vehicle to effectively detect the output drop of the solenoid valve of the electronically controlled suspension system in advance.

[0011] The technical problems of the present invention are not limited to the technical problems mentioned above, and ordinary technicians in the technical field to which the present invention belongs can clearly understand other technical problems not mentioned through the following description.

[0012] Means used to solve problems

[0013] According to one embodiment of the present invention, there is provided a solenoid valve abnormality detection device for detecting an abnormality of a solenoid valve of an electronically controlled suspension system (ECS) configured in a vehicle, the device comprising: a memory for storing one or more instructions; and a processor for executing the one or more instructions, wherein the processor inputs input data representing a state of the electronically controlled suspension system into an artificial neural network model by executing the one or more instructions, obtains an estimated value of a physical quantity representing an output of the electronically controlled suspension system output by the artificial neural network model, and detects whether the solenoid valve is abnormal by comparing the estimated value with a measured value of the physical quantity.

[0014] According to the solenoid valve abnormality detection device of one aspect of the present invention, the input data may be composed of a signal that can be acquired through a controller area network (CAN) of the vehicle.

[0015] According to the solenoid valve abnormality detection device of one aspect of the present invention, the physical quantity may be a damping force of the electronically controlled suspension system.

[0016] According to the solenoid valve abnormality detection device of one aspect of the present invention, the input data may include a vertical acceleration of a wheel of the vehicle and a vertical acceleration of a body of the vehicle.

[0017] According to the solenoid valve abnormality detection device of one embodiment of the present invention, the input data may also include any one or more of the wheel speed of the vehicle, the steering angle of the vehicle, the steering angular velocity of the vehicle, the displacement of the accelerator pedal of the vehicle, the displacement of the brake pedal of the vehicle, and the lateral acceleration of the vehicle.

[0018] According to the electromagnetic valve abnormality detection device of one embodiment of the present invention, the artificial neural network model may include: a generative adversarial network (GAN), including a generator that receives the input data to generate the estimated value.

[0019] According to the electromagnetic valve abnormality detection device of one aspect of the present invention, the artificial neural network model may further include: a discriminator that receives measurement data including the input data and the measurement value and outputs a discrimination value for the measurement data.

[0020] According to the electromagnetic valve abnormality detection device of one aspect of the present invention, the generator may be composed of a multivariate transformer.

[0021] According to the solenoid valve abnormality detection device of one embodiment of the present invention, the processor can input error data associated with the difference between the estimated value and the measured value into the abnormality detection model by executing the one or more instructions to determine whether the solenoid valve is abnormal.

[0022] According to the electromagnetic valve abnormality detection device of one aspect of the present invention, the abnormality detection model may use a One-Class Support Vector Machine (OCSVM) algorithm.

[0023] According to an electromagnetic valve abnormality detection device of one embodiment of the present invention, there may be multiple data sets including the input data, the measurement values ​​and the inferred values, and the error data may include the average and standard error of the errors between the measurement values ​​and the inferred values ​​obtained for each of the multiple data sets, the maximum absolute error between the measurement values ​​and the inferred values ​​of the multiple data sets, and the identification value of the identifier of the measurement data of the multiple data sets.

[0024] According to the electromagnetic valve abnormality detection device of one aspect of the present invention, the discriminator may additionally receive estimated data including the input data and the estimated value and additionally output a discrimination value for the estimated data.

[0025] According to the solenoid valve abnormality detection device of one embodiment of the present invention, the artificial neural network model can be constructed by the generator and the discriminator alternately performing learning with each other, and the input data and the measurement values ​​used in the learning can be obtained when the vehicle and the solenoid valve are in normal conditions.

[0026] According to the solenoid valve abnormality detection device of one aspect of the present invention, the estimated value may track the measured value of the physical quantity acquired when the vehicle and the solenoid valve are in a normal state.

[0027] According to another embodiment of the present invention, there is provided a solenoid valve abnormality detection method for detecting an abnormality of a solenoid valve of an electronically controlled suspension system (ECS) configured in a vehicle, comprising: a step in which a processor inputs input data representing a state of the electronically controlled suspension system into an artificial neural network model, and obtains an estimated value of a physical quantity representing an output of the electronically controlled suspension system output by the artificial neural network model; and a step in which the processor detects whether the solenoid valve is abnormal by comparing the estimated value with a measured value of the physical quantity.

[0028] According to the solenoid valve abnormality detection method according to one aspect of the present invention, the physical quantity may be a damping force of the electronically controlled suspension system.

[0029] According to the solenoid valve abnormality detection method of one aspect of the present invention, the input data may include a vertical acceleration of a wheel of the vehicle and a vertical acceleration of a body of the vehicle.

[0030] According to the solenoid valve abnormality detection method of one embodiment of the present invention, the input data may also include any one or more of the wheel speed of the vehicle, the steering angle of the vehicle, the steering angular velocity of the vehicle, the displacement of the accelerator pedal of the vehicle, the displacement of the brake pedal of the vehicle, and the lateral acceleration of the vehicle.

[0031] According to a solenoid valve abnormality detection method of one embodiment of the present invention, the artificial neural network model can be a generative adversarial network (GAN), and the generative adversarial network can include: a generator, which receives the input data to generate the estimated value; and a discriminator, which receives measurement data including the input data and the measurement value and outputs a discrimination value for the measurement data.

[0032] According to the electromagnetic valve abnormality detection method of one aspect of the present invention, in the step of acquiring the estimated value of the physical quantity, the processor may input the input data to the generator and acquire the estimated value generated by the generator.

[0033] According to a solenoid valve abnormality detection method of one embodiment of the present invention, the step of detecting whether the solenoid valve is abnormal or not may include: the processor inputs measurement data including the input data and the measurement value of the physical quantity into the discriminator, and obtains the identification value for the measurement data generated by the discriminator; and the processor inputs error data including the identification value for the measurement data and a numerical value associated with the difference between the inferred value and the measured value into an abnormality detection model, and obtains the output of the abnormality detection model.

[0034] According to a solenoid valve abnormality detection method in one aspect of the present invention, there may be multiple data sets including the input data, the measurement values ​​and the inferred values, and the error data may include the average and standard errors between the measurement values ​​and the inferred values ​​obtained for each of the multiple data sets, the maximum absolute error between the measurement values ​​and the inferred values ​​of the multiple data sets, and the identification value of the identifier of the measurement data of the multiple data sets.

[0035] According to a solenoid valve abnormality detection method of one embodiment of the present invention, the artificial neural network model can be constructed by the generator and the discriminator alternately performing learning with each other, and the input data and the measurement values ​​used in the learning can be obtained when the vehicle and the solenoid valve are in normal conditions.

[0036] According to another aspect of the present invention, there is provided a computer-readable storage medium, which is a non-transitory computer-readable storage medium storing a program including at least one instruction for executing the solenoid valve abnormality detection method.

[0037] Effects of the Invention

[0038] According to the above-mentioned structure, a solenoid valve abnormality detection device and method for an electronically controlled suspension (ECS) system according to one embodiment of the present invention and a non-temporary computer-readable storage medium storing a program for executing the method, which is based on a digital twin algorithm based on artificial intelligence, can predict performance degradation in advance before a failure occurs in a solenoid valve configured in an electronically controlled suspension system.

[0039] In addition, according to an embodiment of the present invention, a solenoid valve abnormality detection device and method for an electronically controlled suspension system 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 predict performance degradation of the solenoid valve of the electronically controlled suspension system without the need for additional sensors.

[0040] 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 derived from the constitution of the invention described in the detailed description of the present invention or the appended claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 The diagram schematically shows the configuration of an electronically controlled suspension (ECS) system of a vehicle.

[0042] Figure 2 1 is a diagram showing a structure of a solenoid valve abnormality detection device of an electronically controlled suspension (ECS) system according to an embodiment of the present invention.

[0043] Figure 3 The diagram shows a model used for obtaining an estimated value or detecting abnormality in a solenoid valve abnormality detection device for an electronically controlled suspension system according to an embodiment of the present invention.

[0044] Figure 4 It is a diagram showing the detailed structure of the artificial neural network model.

[0045] Figure 5 This is a diagram showing the detailed structure of a generator of an artificial neural network model.

[0046] Figure 6 1 is a diagram showing the operation of the solenoid valve abnormality detection device of the electronically controlled suspension system according to an embodiment of the present invention.

[0047] Figure 7 4 is a flow chart of a method for detecting abnormality of a solenoid valve of an electronically controlled suspension system according to an embodiment of the present invention.

[0048] Figure 8 The invention is a detailed flowchart of the steps of detecting whether a solenoid valve is abnormal or not in a method for detecting abnormality of a solenoid valve of an electronically controlled suspension system according to an embodiment of the present invention.

[0049] Description of Reference Numerals

[0050] 100: electromagnetic valve abnormality detection device; 110: memory; 120: processor DETAILED DESCRIPTION

[0051] Hereinafter, embodiments of the present invention are described in detail with reference to the accompanying drawings so that a person skilled in the art can easily implement the present invention. The present invention can be implemented in various forms and is not limited to the embodiments described herein. In order to clearly explain the present invention, parts not related to the description are omitted in the accompanying drawings, and the same reference numerals are given to the same or similar components throughout the specification.

[0052] The words and terms used in this specification and the appended claims should not be interpreted as being limited to their usual or dictionary meanings, but in order to describe one's invention in the best way, they should be interpreted as meanings and concepts that are consistent with the technical ideas of the present invention in accordance with the principle that the inventor is able to define terms and concepts.

[0053] In this specification, the terms "include" or "have" are used to illustrate the existence of features, numbers, steps, actions, constituent elements, parts or their combinations recorded in the specification, rather than excluding the existence or additional possibilities of one or more other features, numbers, steps, actions, constituent elements, parts or their combinations in advance.

[0054] Figure 1 The diagram schematically shows the configuration of an electronically controlled suspension (ECS) system of a vehicle.

[0055] The electronically controlled suspension system 10 of the vehicle is provided in the vehicle to actively control the suspension device according to the road surface state and driving conditions. The electronically controlled suspension system 10 can also change the vehicle body height according to the road surface state and driving conditions, thereby ensuring driving safety and riding feeling at the same time.

[0056] Reference Figure 1 The electronically controlled suspension system 10 may include: a shock absorber 11 having a solenoid valve 11 a (solenoid valve); a solenoid driving unit 12 driving a solenoid (solenoid) of the solenoid valve 11 a; and an electronic controller (ECU) 13 controlling the solenoid driving unit 12 .

[0057] Figure 2 1 is a diagram showing the structure of a solenoid valve abnormality detection device of an electronically controlled suspension (ECS) system according to an embodiment of the present invention. Figure 3 The diagram shows a model used for obtaining an estimated value or detecting abnormality in a solenoid valve abnormality detection device for an electronically controlled suspension system according to an embodiment of the present invention.

[0058] According to an embodiment of the present invention, a solenoid valve abnormality detection device 100 for an electronically controlled suspension system detects abnormality of a solenoid valve 11a configured in an electronically controlled suspension system of a vehicle. More specifically, the solenoid valve abnormality detection device 100 for an electronically controlled suspension system is configured in an electronically controlled suspension system of a vehicle and can detect abnormality of a solenoid valve used to adjust a damping force.

[0059] According to an embodiment of the present invention, the electromagnetic valve abnormality detection device 100 of the electronically controlled suspension system can predict the functional degradation of the electromagnetic valve before a fault occurs in the electromagnetic valve by using a digital twin algorithm for the vehicle based on artificial intelligence. The digital twin algorithm can virtually construct the electronically controlled suspension system of the vehicle.

[0060] Reference Figure 2 According to an embodiment of the present invention, a solenoid valve abnormality detection device 100 may include a memory 110 and a processor 120 .

[0061] The memory 110 stores one or more instructions that can be executed by the processor 120 .

[0062] 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), digital video disks (DVD), and magneto-optical media such as floppy disks.

[0063] 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 one or more arithmetic logic units (ALUs) and registers.

[0064] By executing the one or more instructions, the processor 120 inputs input data representing the state of the electronically controlled suspension system 10 to the artificial neural network model 20 and obtains an estimated value of a physical quantity representing the output of the electronically controlled suspension system output by the artificial neural network model 20 .

[0065] In one embodiment of the present invention, the input data may include the vertical acceleration of the wheels of the vehicle and the vertical acceleration of the body of the vehicle. In addition, the input data may also include: the wheel speed of the vehicle, the steering angle of the vehicle, the steering angular velocity of the vehicle, the displacement of the accelerator pedal of the vehicle, the displacement of the brake pedal of the vehicle, and the lateral acceleration of the vehicle.

[0066] The physical quantity may be the damping force of the electronically controlled suspension system 10. That is, the artificial neural network model 20 may output an estimated value of the damping force. The damping force of the electronically controlled suspension system may be determined based on the operation of the solenoid valve 11a.

[0067] In addition, the input data may be composed of a signal that can be obtained through the vehicle's controller area network (CAN) 200. The above-mentioned example of the input data can be obtained through the CAN of a general vehicle. Therefore, according to the present invention, there is no need to use an additional sensor related to abnormality detection of the solenoid valve.

[0068] In association with this, the memory 110 may be directly or indirectly connected to the CAN 200 of the vehicle. The memory 110 may receive input data from the CAN 200. Therefore, the processor 120 may acquire the input data from the memory 110.

[0069] The artificial neural network model 20 receives the input data and outputs the estimated value of the physical quantity. In one embodiment of the present invention, the artificial neural network model 20 may include a generative adversarial network (GAN).

[0070] In one embodiment of the present invention, the artificial neural network model 20 may be a deep learning-based neural twin model for an electronically controlled suspension system of a vehicle. In other words, the artificial neural network model 20 may function as a virtual twin model of the electronically controlled suspension system of the vehicle.

[0071] Figure 4 It is a diagram showing the detailed structure of the artificial neural network model.

[0072] Reference Figure 4 , the artificial neural network model 20 may include a generator 21 and a discriminator 22.

[0073] The artificial neural network model 20 is constructed by alternately performing learning by the generator 21 and the discriminator 22, and the data used in the learning can be obtained when the vehicle and the solenoid valve are in a normal state. Thus, the estimated value can track the measured value of the physical quantity obtained when the vehicle and the solenoid valve are in a normal state.

[0074] The generator 21 receives the input data to generate the inference value. 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 GAN based on a multivariate transformer.

[0075] In one embodiment of the present invention, the generator 21 can receive as the input data the vertical acceleration of the vehicle's wheels, the vertical acceleration of the vehicle's body, the vehicle's wheel speed, the vehicle's steering angle, the vehicle's steering angular velocity, the vehicle's accelerator pedal displacement, the vehicle's brake pedal displacement and the vehicle's lateral acceleration.

[0076] The generator 21 can output the estimated value of the damping force of the electronically controlled suspension system by calculating the input data. The estimated value tracked by the estimation generator 21 is the measured value of the damping force outputted in accordance with the input data when the electronically controlled suspension system is in a normal state.

[0077] Figure 5 This is a diagram showing the detailed structure of a generator of an artificial neural network model.

[0078] Reference Figure 5 , the generator 21 may include an encoder 21a and a decoder 21b.

[0079] The encoder 21a and the decoder 21b constitute a network for estimating the damping force of the electronically controlled suspension system. In other words, the generator 21 may be composed of an encoder 21a-decoder 21b structure network trained in an end-to-end manner to estimate the damping force.

[0080] The encoder 21a may be designed to extract high-level features from input data using a transformer block and linear operations. In this case, the transformer block may include an embedding layer, a self-attention layer, and a feed-forward layer.

[0081] First, the embedding layer in the transformer block can encode spatiotemporal information into input sequences using a trainable embedding matrix.

[0082] Next, deep features are extracted through a self-attention layer and refined using a feed-forward layer.

[0083] At this time, the self-attention layer can use a scaled dot-product attention mechanism. The self-attention layer can be the main building block in the variable-based deep neural network (DNN) architecture that allows the model to process and understand dependencies within the input sequence.

[0084] The scaled dot product attention first applies layer normalization to the input sequence and processes its results into three intrinsic vectors of query vector Q, key vector K and value vector V through linear operation.

[0085] Next, based on the derived vector, the attention score AS for each position of the input sequence can be calculated. The calculation formula of the attention score AS can be given as shown in the following formula 1:

[0086] Formula 1

[0087]

[0088] (σ(·) represents the softmax operation, ||x input || dim represents the feature dimension of the input, (·) represents the dot product)

[0089] The attention score quantifies the relative importance of each position in the input sequence associated with a given query vector. That is, more appropriate positions will be assigned higher scores.

[0090] In order to calculate the query vector, key vector and initial input value of the decoder 21b, the output of the transformer block can be supplied to two different linear operations.

[0091] The decoder 21b may use the acquired values ​​to infer the damping force of the electronically controlled suspension system in an auto-regressive manner similar to Sequence-to-Sequence Recurrent Neural Networks.

[0092] Similar to the encoder 21a, the decoder 21b may include a transformer block and a linear operation. In addition, the decoder 21b may further include a cross-attention layer in the transformer block.

[0093] The cross-attention layer can use the query vector and key vector of encoder 21a to perform scaled dot-product attention to learn the dynamic correlation between the input signal and the inferred signal.

[0094] The discriminator 22 receives the measurement data including the input data and the measurement value, and outputs a discriminator value for the measurement data.

[0095] In addition, the discriminator 22 may also additionally receive the inferred data including the input data and the inferred value, and additionally output the discriminant value for the inferred data. Such a process may be completed when the generator 21 and the discriminator 22 perform learning.

[0096] In this way, the discriminator 22 can output a discriminant value for the measurement data during the operation for abnormality detection, and at the same time, can output a discriminant value for the inferred data during the operation for learning.

[0097] The discriminator 22 may be composed of a discriminator network used for adversarial training of GAN. In other words, the generator 21 and the discriminator 22 may improve the inference performance and perform training in a manner of performing adversarial training to set the GAN model.

[0098] In more detail, the discriminator 22 may perform optimization alternately with the generator 21 to solve a Wasserstein min-max problem such as Formula 2.

[0099] Formula 2

[0100]

[0101] (x D，Real represents a real data set consisting of input data and measured values ​​of the damping force of the electronically controlled suspension system, x D，Fake represents a virtual data sample consisting of input data and estimated values ​​of the damping force of the electronically controlled suspension system)

[0102] In the artificial neural network model 20, the generator 21 is trained to deceive theD，Real and x D，Fake The generator 21 can thus generate an estimated value for tracking the damping force of the vehicle electronically controlled suspension system in a normal state.

[0103] In association with this, the loss function for GAN training can be defined as the following formula 3:

[0104] Formula 3

[0105]

[0106]

[0107] (N is the batch size, L G is the loss function of the generator, L D is the loss function of the discriminator)

[0108] The first term of the loss of generator 21 is equivalent 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 GAN loss function L GAN .

[0109] In one embodiment of the present invention, in order to reflect the mechanical characteristics of the electronically controlled suspension system into the loss function, a damping force function derived from a quarter car model may be considered. The quarter car model is commonly used in passenger car design and is a suspension model for considering input from one wheel of a vehicle.

[0110] The quarter car model can be defined as follows:

[0111] Formula 4

[0112]

[0113] (m b The quality of the car body, is the vertical acceleration of the vehicle body, is the vertical speed of the vehicle body, is the up-down speed of the vehicle's wheels, (x b -x w ) is the displacement between the vehicle body and wheels, k s is the given elastic modulus)

[0114] From such a quarter car model, the damping force of the vehicle's electronically controlled suspension system can be derived as shown in Formula 5:

[0115] Formula 5

[0116]

[0117] ( is the vertical speed of the vehicle body, is the up and down speed of the vehicle's wheels, m b The quality of the car body, is the vertical acceleration of the vehicle body, k s is the given elastic modulus, (x b -x w ) is the displacement between the vehicle's body and wheels)

[0118] At this time, the damping force calibrated by reflecting the uncertainty of the model and the system can be derived according to the following formula 6:

[0119] Formula 6

[0120]

[0121] ( is the input data, m b The quality of the car body, is the vertical acceleration of the vehicle body, (x b -x w ) is the displacement between the vehicle body and wheels, k s is the given elastic modulus)

[0122] Calibrated damping force F e will become the estimated value of the damping force of the electronically controlled suspension system. In addition, in Formula 6, δ1 and δ2 represent model-related uncertainties, and δ3 represents system uncertainty. In addition, the input data It is the data that the vehicle's ECU can receive or measure via CAN.

[0123] The processor 120 detects whether the solenoid valve is abnormal by comparing the estimated value with the measured value of the physical quantity. The processor 120 can judge whether the solenoid valve is abnormal by executing the one or more instructions and inputting the error data associated with the difference between the estimated value and the measured value into the abnormality detection model 30. For example, the abnormality detection model 30 can use a one-class support vector machine (OCSVM) algorithm.

[0124] There may be multiple data sets including the input data, the measured values, and the inferred values. The error data may include the average and standard error of the error between the measured values ​​and the inferred values ​​obtained in each of the multiple data sets, the maximum absolute error between the measured values ​​and the inferred values ​​of the multiple data sets, and the identification value of the discriminator for the measured data of the multiple data sets.

[0125] In more detail, a plurality of data sets including the measured values ​​and the estimated values ​​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 may be acquired for each batch.

[0126] The error data acquired for one lot may be input to the abnormality detection model 30. In addition, the degree of abnormality (degradation in performance of the electromagnetic valve) of the electromagnetic valve may be detected based on the output (feature) acquired from the abnormality detection model 30 receiving the error data.

[0127] Regarding the anomaly detection metric, we can consider the F1 score given by the following formula 7:

[0128] Formula 7

[0129]

[0130] (TP is true positive, FP is false positive, FN is false negative, positive indicates that the performance of the solenoid valve is reduced)

[0131] In addition, as described above, the discriminator 22 can calculate not only the discriminant value of the measured data included in the data set, but also the discriminant value of the inferred data included in the data set. The discriminant value of the inferred data of the data set by the discriminator 22 can be fed back to the generator 21 of the artificial neural network model 20 for training.

[0132] So far, the structure of the electromagnetic valve abnormality detection device 100 for an electronically controlled suspension system according to an embodiment of the present invention has been described in detail. Next, the operation of the electromagnetic valve abnormality detection device 100 for an electronically controlled suspension system will be described in detail.

[0133] Figure 6 1 is a diagram showing the operation of the solenoid valve abnormality detection device of the electronically controlled suspension system according to an embodiment of the present invention.

[0134] Reference Figure 6The solenoid valve abnormality detection device 100 of the electronically controlled suspension system according to an embodiment of the present invention may operate as follows.

[0135] First, the memory 110 stores input data obtained from the electronically controlled suspension system of the vehicle. and the measured value of the damping force of the electronically controlled suspension system F m .

[0136] As mentioned above, input data It may include the vertical acceleration of the vehicle's wheels, the vertical acceleration of the vehicle's body, the vehicle's wheel speed, the vehicle's steering angle, the vehicle's steering angular velocity, the vehicle's accelerator pedal displacement, the vehicle's brake pedal displacement and the vehicle's lateral acceleration.

[0137] In addition, multiple input data can be and the measured value of the damping force of the electronically controlled suspension system F m That is, in one embodiment of the present invention, data processing and calculation can be performed in units of data batches.

[0138] Then, the processor 120 may execute one or more instructions to convert the input data The artificial neural network model 20 is inputted to obtain the estimated value F of the damping force of the electronically controlled suspension system outputted by the artificial neural network model 20. e In more detail, the processor 120 may convert the input data The estimated value F of the damping force of the electronically controlled suspension system output by the generator 21 is obtained. e .

[0139] In addition, the processor 120 may execute more than one instruction to convert the input data and the measured value of the damping force F m The input data is input to the artificial neural network model 20, and the identification value DS is obtained. In more detail, the processor 120 can include the input data and the measured value of the damping force F m The measurement data is input to the discriminator 22 of the artificial neural network model 20, and the discrimination value DS output by the discriminator 22 for the measurement data is obtained.

[0140] Next, the processor 120 outputs the error data D Error . Error data D Error The data can be output in different batches.

[0141] As described above, there are a plurality of data sets including the input data, the measured values, and the estimated values, and the error data D Error The average μ of the errors between the measured values ​​and the estimated values ​​obtained from each of the multiple data sets may be included. E and standard error σ E , the maximum absolute error Max between the measured values ​​and the estimated values ​​of the multiple data sets E , a discriminator for the measurement data obtained from the plurality of said data sets.

[0142] Finally, the processor 120 executes one or more instructions to convert the error data D 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.

[0143] Based on the received error data D Error Based on the output (features) acquired by the abnormality detection model 30, the processor 120 can detect the abnormality level (performance degradation of the solenoid valve) of the solenoid valve of the electronically controlled suspension system.

[0144] The above has described in detail the solenoid valve abnormality detection device 100 for an electronically controlled suspension system according to an embodiment of the present invention. The following will describe a solenoid valve abnormality detection method for an electronically controlled suspension system.

[0145] Figure 7 4 is a flow chart of a method for detecting abnormality of a solenoid valve of an electronically controlled suspension system according to an embodiment of the present invention.

[0146] The solenoid valve abnormality detection method S100 of an electronically controlled suspension system according to an embodiment of the present invention detects abnormality of a solenoid valve 11a configured in an electronically controlled suspension system of a vehicle. More specifically, the solenoid valve abnormality detection method S100 of an electronically controlled suspension system can detect abnormality of a solenoid valve configured in an electronically controlled suspension system of a vehicle and adjusting damping force.

[0147] Reference Figure 7 The solenoid valve abnormality detection method S100 of the electronically controlled suspension system according to an embodiment of the present invention may be performed as follows.

[0148] First, the processor 120 inputs input data representing the state of the electronically controlled suspension system into the artificial neural network model 20, and obtains an estimated value of a physical quantity representing an output of the electronically controlled suspension system output by the artificial neural network model (step S110).

[0149] The input data may include the vertical acceleration of the wheels of the vehicle and the vertical acceleration of the body of the vehicle. In addition, the input data may also include any one or more of the wheel speed of the vehicle, the steering angle of the vehicle, the steering angular velocity of the vehicle, the displacement of the accelerator pedal of the vehicle, the displacement of the brake pedal of the vehicle, and the lateral acceleration of the vehicle.

[0150] The physical quantity may be a damping force of the electronically controlled suspension system. The damping force of the electronically controlled suspension system may be determined according to the operation of the solenoid valve.

[0151] For example, the input data may consist of the vertical acceleration of the vehicle's wheels, the vertical acceleration of the vehicle's body, the vehicle's wheel speed, the vehicle's steering angle, the vehicle's steering angular velocity, the vehicle's accelerator pedal displacement, the vehicle's brake pedal displacement and the vehicle's lateral acceleration.

[0152] 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 neural twin model based on deep learning of the electronically controlled suspension system of the vehicle.

[0153] The artificial neural network model 20 may include a generator 21 and a discriminator 22. The generator 21 receives the input data to generate the inference value. The generator 21 may be composed of a multivariate transformer. In addition, the discriminator 22 receives the measurement data including the input data and the measurement value, and outputs a discriminator value for the measurement data.

[0154] In the step S110 of acquiring the estimated value of the physical quantity, the processor 120 may input the input data to the generator 21 and acquire the estimated value generated by the generator 21 .

[0155] In addition, the artificial neural network model 20 is constructed by performing learning by 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 solenoid valve are in a normal state. Thus, the estimated value can track the measured value of the physical quantity obtained when the vehicle and the solenoid valve are in a normal state.

[0156] Next, the processor 120 detects whether the solenoid valve is abnormal by comparing the estimated value with the measured value of the physical quantity (step S120 ).

[0157] The processor 120 can determine whether the solenoid valve is abnormal by executing one or more instructions and inputting the error data associated with the difference between the estimated value and the measured value into the abnormality detection model 30. For example, the abnormality detection model 30 can use a one-class support vector machine (OCSVM) algorithm.

[0158] Figure 8 The invention is a detailed flowchart of the steps of detecting whether a solenoid valve is abnormal or not in a method for detecting abnormality of a solenoid valve of an electronically controlled suspension system according to an embodiment of the present invention.

[0159] Reference Figure 8 The step S120 of detecting whether the solenoid valve is abnormal or not can be performed as follows.

[0160] First, the processor 120 inputs the measurement data including the input data and the measurement value of the physical quantity to the discriminator 22 , and acquires the discrimination value of the measurement data generated by the discriminator 22 (step S121 ).

[0161] Next, the processor 120 inputs error data including a discrimination value for the measurement data and a numerical value associated with a difference between the estimated value and the measurement value into the abnormality detection model 30 and acquires an output of the abnormality detection model 30 (step S122 ).

[0162] There may be multiple data sets including the input data, the measured values, and the inferred values. The error data may include the average and standard error of the error between the measured values ​​and the inferred values ​​obtained for each of the multiple data sets, the maximum absolute error between the measured values ​​and the inferred values ​​of the multiple data sets, and the identification value of the discriminator for the measured data of the multiple data sets.

[0163] In more detail, a plurality of data sets including the measured values ​​and the estimated values ​​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 may be acquired for each batch.

[0164] The error data acquired for one lot may be input to the abnormality detection model 30. In addition, the level of abnormality (degradation in performance of the electromagnetic valve) of the electromagnetic valve may be detected based on the output (feature) acquired from the abnormality detection model 30 receiving the error data.

[0165] Regarding the anomaly detection metric, the F1 score as described above may be considered.

[0166] In addition, the present invention additionally provides a non-transitory computer-readable storage medium storing a program for executing the solenoid valve abnormality detection method of the electronically controlled suspension system. Specifically, the present invention may provide a non-transitory computer-readable storage medium storing a program including at least one instruction for executing the solenoid valve abnormality detection method of the electronically controlled suspension system.

[0167] At this time, the instructions may include not only the machine code generated by the compiler, but also the high-level language code executable by the computer.

[0168] The recording medium may also include 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) and digital video disk (DVD), magneto-optical media (Magneto-Optical Media) such as floppy disks, hardware devices such as read-only memory (ROM), random access memory (RAM), flash memory, etc., which are constructed in a manner of storing and executing program instructions.

[0169] The simulation results show that according to the present invention, the maximum output comparison of the damping force of the electronically controlled suspension system can be predicted by using the performance of the F1 score of 0.85 or more. In other words, according to the present invention, the deterioration level of the solenoid valve configured in the electronically controlled suspension system of the vehicle can be accurately predicted before the solenoid valve completely fails.

[0170] According to the present invention, it is possible to predict in advance before a failure occurs in the solenoid valve of the electronically controlled suspension system, thereby enabling advance measures to be taken before a failure occurs in the solenoid valve, thereby enabling the failure of the solenoid valve to be prevented.

[0171] Thus, the present invention provides prediction of future behavior and remaining useful life of the system and predictive diagnosis suitable for predictive maintenance repair applications. Thus, proactive measures and repairs can be effectively guided before a fault occurs in the solenoid valve of the vehicle's electronically controlled suspension system.

[0172] Although the embodiments of the present invention have been described, the technical concept of the present invention is not limited to the embodiments presented in this specification, and those skilled in the art who understand the technical concept of the present invention can easily propose other embodiments within the scope of the same technical concept by adding, changing, deleting, adding, etc. to the constituent elements. However, this will also be considered to fall within the scope of the present invention.

Claims

1. A solenoid valve abnormality detection device for detecting an abnormality of a solenoid valve configured in an electronically controlled suspension system of a vehicle, in, include: a memory storing one or more instructions; as well as a processor, executing said one or more instructions, The processor inputs input data representing the state of the electronically controlled suspension system into the artificial neural network model by executing one or more instructions, obtains an estimated value of a physical quantity representing the output of the electronically controlled suspension system output by the artificial neural network model, and detects whether the solenoid valve is abnormal by comparing the estimated value with a measured value of the physical quantity.

2. The solenoid valve abnormality detection device according to claim 1, wherein: The input data consists of signals accessible via a controller area network of the vehicle.

3. The solenoid valve abnormality detection device according to claim 1, wherein: The physical quantity is the damping force of the electronically controlled suspension system.

4. The solenoid valve abnormality detection device according to claim 1, wherein: The input data includes a vertical acceleration of a wheel of the vehicle and a vertical acceleration of a body of the vehicle.

5. The solenoid valve abnormality detection device according to claim 4, wherein: The input data further includes any one or more of the wheel speed of the vehicle, the steering angle of the vehicle, the steering angular velocity of the vehicle, the displacement of the accelerator pedal of the vehicle, the displacement of the brake pedal of the vehicle, and the lateral acceleration of the vehicle.

6. The electromagnetic valve abnormality detection device according to claim 1, wherein: The artificial neural network model comprises: A generative adversarial network includes a generator that receives the input data to generate the inference value.

7. The electromagnetic valve abnormality detection device according to claim 6, wherein: The artificial neural network model also includes: The discriminator receives measurement data including the input data and the measurement value, and outputs a discrimination value for the measurement data.

8. The solenoid valve abnormality detection device according to claim 7, wherein: The processor inputs error data associated with a difference between the estimated value and the measured value into an abnormality detection model by executing the one or more instructions to determine whether the solenoid valve is abnormal.

9. The electromagnetic valve abnormality detection device according to claim 8, wherein: There are multiple data sets including the input data, the measured values, and the inferred values. The error data includes the average and standard error of the errors between the measured values ​​and the inferred values ​​obtained for each of the plurality of data sets, the maximum absolute error between the measured values ​​and the inferred values ​​of the plurality of data sets, and the identification value of the identifier for the measured data of the plurality of data sets.

10. A solenoid valve abnormality detection method for detecting an abnormality of a solenoid valve configured in an electronically controlled suspension system of a vehicle, wherein: include: The processor inputs input data representing the state of the electronically controlled suspension system into an artificial neural network model, and obtains an estimated value of a physical quantity representing an output of the electronically controlled suspension system output by the artificial neural network model; as well as The processor detects whether the electromagnetic valve is abnormal by comparing the estimated value with the measured value of the physical quantity.

11. The solenoid valve abnormality detection method according to claim 10, wherein: The input data includes a vertical acceleration of a wheel of the vehicle and a vertical acceleration of a body of the vehicle.

12. The solenoid valve abnormality detection method according to claim 11, wherein: The input data further includes any one or more of the wheel speed of the vehicle, the steering angle of the vehicle, the steering angular velocity of the vehicle, the displacement of the accelerator pedal of the vehicle, the displacement of the brake pedal of the vehicle, and the lateral acceleration of the vehicle.

13. The solenoid valve abnormality detection method according to claim 10, wherein: The artificial neural network model is a generative adversarial network, The generative adversarial network comprises: A generator, receiving the input data to generate the inferred value; and A discriminator receives measurement data including the input data and the measurement value and outputs a discrimination value for the measurement data.

14. The solenoid valve abnormality detection method according to claim 13, wherein: The step of detecting whether the solenoid valve is abnormal comprises: The processor inputs measurement data including the input data and the measurement value of the physical quantity into the discriminator, and acquires a discrimination value for the measurement data generated by the discriminator; and The processor inputs error data including a discrimination value for the measurement data and a numerical value associated with a difference between the inferred value and the measurement value into an abnormality detection model and acquires an output of the abnormality detection model.

15. A computer-readable storage medium, which is a non-transitory computer-readable storage medium, wherein: A program including at least one instruction for executing the electromagnetic valve abnormality detection method according to any one of claims 10 to 14 is stored.

Images