Brake system and method of controlling a brake system

Abstract

The invention relates to a brake system and a method of controlling the brake system. The brake system includes a sensor module including a motor current sensor and a force sensor, electromechanical brake units respectively mounted on wheels of a vehicle and including a motor, and a controller configured to control one or more of the electromechanical brake units; the controller is configured to: predict a state of the motor of the electromechanical brake unit based on a current signal of the motor detected by the motor current sensor, when at least one predicted state of the motor indicates that at least one of the motors is faulty, determine whether the motor is faulty; determining a fault level of the at least one failed motor based on sensor data obtained from the sensor module, calculating a requested torque for each of the wheels based on the determined fault level of the at least one failed motor, the torque of each of the wheels is controlled based on the calculated requested torque of each of the wheels.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority from Korean Patent Application No. 10-2024-0046031 filed on April 4, 2024, in the Korean Intellectual Property Office, the disclosure of which is incorporated herein by reference in its entirety. Technical Field

[0003] The present invention relates to a braking system and a control method thereof, wherein the braking system predicts a failure of a motor to control the torque of each wheel. Background Art

[0004] Typically, when pedal force is applied to the pedal according to the driver's braking intention, the vehicle controls the vehicle braking system to obtain the driver's requested torque according to the displacement of the brake pedal, calculates the requested torque of each wheel based on the obtained driver's requested torque, and then predicts the requested torque obtained by calculating the torque of each wheel.

[0005] In vehicles, electromechanical brakes (EMBs), which use motors to control braking, are installed in the wheels. The EMB is a key component directly linked to vehicle safety accidents. However, even if its performance degrades, the EMB continues to operate normally until a failure occurs, leading to reduced vehicle safety.

[0006] Therefore, there is a need for a brake system and a control method thereof to prevent dangerous situations that may occur in a vehicle environment in advance by predicting performance degradation due to EMB motor abnormality in advance rather than after diagnosis. Summary of the Invention

[0007] An aspect of the present invention is to provide a braking system for predicting failure of at least one motor in advance and a control method thereof.

[0008] According to one aspect of the present invention, a braking system includes: a sensor module including a motor current sensor and a force sensor, an electromechanical brake unit, each of which is mounted on a wheel of a vehicle and includes a motor, and a controller configured to control one or more of the electromechanical brake units; the controller is configured to: predict a state of the motor of the electromechanical brake unit based on a current signal of the motor detected by the motor current sensor, determine a fault level of at least one of the motors that has failed based on sensor data obtained from the sensor module when at least one predicted state of the motor indicates that at least one of the motors has failed, calculate a requested torque for each of the wheels based on the determined fault level of at least one of the motors that has failed, and control the torque of each of the wheels based on the calculated requested torque for each of the wheels.

[0009] The controller may be configured to use a machine learning model trained to predict the state of the motor to obtain prediction data representing the predicted motor state.

[0010] The state of the motor may include a normal state, a bearing fault state, a rotor imbalance state, or a shaft misalignment state.

[0011] The current signal of the motor may include two current signals having different phases among three-phase current signals of the motor.

[0012] The controller can be configured to control to use the two current signals with different phases to generate the remaining phase current signals in the three-phase current signal, perform direct quadrature (DQ) transformation on the three-phase current signal to obtain a direct quadrature axis current signal, and generate input data for a machine learning model based on the three-phase current signal and the direct quadrature axis current signal.

[0013] The controller can be configured to obtain prediction result data representing the predicted motor state using a machine learning model trained to predict the motor state based on input data, wherein the input data includes frequency data and amplitude data of the first phase current signal in the three-phase current signal, frequency data and amplitude data of the second phase current signal, and frequency data of the quadrature-axis (Q-axis) current signal.

[0014] The prediction result data may include multiple labels and a predicted probability value for each of the multiple labels, the multiple labels including a label corresponding to the normal state, a label corresponding to the bearing fault state, a label corresponding to the rotor imbalance state, and a label corresponding to the shaft offset state.

[0015] The controller may be configured to determine, among a plurality of labels, a label corresponding to a predetermined threshold probability value or above other than a predicted probability value corresponding to a label corresponding to a normal state as the motor state, the predicted probability value being equal to or higher than the threshold probability value.

[0016] The threshold probability value may be calculated by applying a predetermined weight to a predicted probability value corresponding to a label corresponding to a normal state.

[0017] The sensor data may include a signal associated with a braking force measured by the force sensor, and the controller may be configured to determine the fault level based on the signal associated with the measured braking force and the direct quadrature axis current signal.

[0018] If the direct-axis (D-axis) current value and the Q-axis current value of the current signal of one motor increase to be equal to or higher than a predetermined current value and the measured braking force decreases to be lower than a predetermined braking force, the controller can predict the state of the one motor as one of a bearing fault state and a rotor imbalance state.

[0019] If the D-axis current value and the Q-axis current value of the current signal of one motor decrease below predetermined current values ​​and the measured braking force decreases below predetermined braking force, the controller may predict the state of the one motor as an axis offset state.

[0020] The fault levels include a higher fault level and a lower fault level, the lower fault level being lower than the higher fault level, and the controller may be configured to control the display to output information indicating the determined fault level of at least one of the motors that has failed, if the determined fault level of at least one of the motors that has failed is the lower fault level between the higher fault level and the lower fault level.

[0021] The controller may be configured to determine whether the measured braking force reaches the target braking force within a predetermined first threshold time, and if the measured braking force does not reach the target braking force within the first threshold time, determine the fault level as the lower fault level between the higher fault level and the lower fault level.

[0022] The controller may be configured to determine whether a braking response time when the measured braking force reaches the target braking force is within a predetermined second threshold time, and if the braking response time when the measured braking force reaches the target braking force is within the second threshold time, determine the fault level as a lower fault level between a higher fault level and a lower fault level.

[0023] The controller can be configured to determine that the requested torque value of the wheel corresponding to at least one of the motors of the electromechanical brake unit that is predicted to fail is lower than a predetermined threshold requested torque value, and to determine that one or more requested torque values ​​of the wheel corresponding to another motor or other motors of the motors of the electromechanical brake unit that are not predicted to fail are equal to or higher than a predetermined threshold requested torque value.

[0024] According to one aspect of the present invention, a method for controlling a braking system may include the following steps: predicting a state of a motor of an electromechanical brake unit based on a current signal of the motor detected by a motor current sensor, when at least one predicted state of the motor indicates that at least one of the motors has failed, determining a fault level of at least one of the failed motors based on sensor data obtained from a sensor module including the motor current sensor, calculating a requested torque for each wheel based on the determined fault level of at least one of the failed motors, and controlling the torque of each wheel based on the calculated requested torque for each of the wheels.

[0025] The method may also include the following steps: controlling to use the two current signals with different phases to generate the remaining phase current signal in the three-phase current signal; performing a direct orthogonal transformation on the three-phase current signal to obtain a direct orthogonal axis current signal; and generating input data for a machine learning model based on the three-phase current signal and the direct orthogonal axis current signal.

[0026] The method may also include the following steps: using a machine learning model trained to predict the motor state based on input data to obtain prediction result data representing the predicted motor state; the input data includes frequency data and amplitude data of the first phase current signal in the three-phase current signal, frequency data and amplitude data of the second phase current signal, and frequency data of the quadrature-axis (Q-axis) current signal.

[0027] The step of predicting the state of the motor of the electromechanical brake unit may include the following steps: determining a tag having a predicted probability value corresponding to the tag corresponding to the normal state among the multiple tags as the motor state; the predicted probability value is equal to or higher than a threshold probability value.

[0028] The step of determining a fault level of the at least one motor that has failed may comprise the step of determining the fault level based on a signal associated with the measured braking force and the direct quadrature axis current signal.

[0029] The method may further comprise the step of displaying information indicative of the determined fault level of at least one of the motors that has failed.

[0030] The effects of the present invention are not limited to the above-mentioned effects, and other effects not mentioned above will be apparently understood by those skilled in the art from the following description.

[0031] The above-described objects to be achieved by the present invention, means for achieving the objects, and effects of the present invention do not specify essential features of the claims, and therefore, the scope of the claims is not limited to the contents of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The above and other aspects, features and other advantages of the present invention will be more clearly understood from the following detailed description taken in conjunction with the accompanying drawings, in which:

[0033] Figure 1 shows a braking system included in a vehicle according to an exemplary embodiment;

[0034] Figure 2 shows a control configuration of a braking system according to an exemplary embodiment;

[0035] Figure 3 shows functional modules of a controller included in a brake control unit according to an exemplary embodiment;

[0036] Figure 4 An example diagram is shown for explaining the prediction work of a machine learning model according to an exemplary embodiment;

[0037] Figure 5 An example diagram for explaining the learning operation of a machine learning model according to an exemplary embodiment is shown;

[0038] Figure 6 and Figure 7 An example diagram for explaining the operation of a control module when a motor of a first wheel is predicted to fail according to an exemplary embodiment is shown;

[0039] Figure 8 and Figure 9 An example diagram for explaining the operation of a control module when a failure is predicted for motors of a first wheel and a third wheel according to an exemplary embodiment is shown;

[0040] Figure 10 shows a torque control operation of a controller included in a brake control unit according to an exemplary embodiment;

[0041] Figure 11 illustrates the operation of a controller for predicting the motor state of each wheel according to an exemplary embodiment; and

[0042] Figure 12 An operation of the controller is shown when a fault level of at least one motor is determined to be a second fault level according to an exemplary embodiment. DETAILED DESCRIPTION

[0043] Hereinafter, exemplary embodiments of the present invention will be described with reference to the following drawings and exemplary embodiments. For descriptive purposes, the proportions of components shown in the drawings are different from the actual proportions, and thus the proportions are not limited to the proportions shown in the drawings.

[0044] Throughout the specification, the same reference numerals denote the same elements. The specification does not describe all elements of the exemplary embodiments, but may omit redundant descriptions of general aspects of the technical field to which the present invention pertains or exemplary embodiments. The terms "unit," "module," "component," and "block" used in the specification may be implemented by software or hardware, and, depending on the exemplary embodiments, multiple "units," "modules," "components," and "blocks" may be implemented by one component, or one "unit," "module," "component," and "block" may include multiple components.

[0045] Throughout the specification, when a part is described as being “connected” to another part, the part may be directly connected to the other part or may be indirectly connected to the other part, and the indirect connection includes connection to a wireless communication network therebetween.

[0046] In addition, unless explicitly described to the contrary, the word “comprise” will be understood to imply the inclusion of stated elements but not the exclusion of any other elements.

[0047] Throughout the description of the present invention, when a component is "on" another component, the component may be adjacent to the other component, or a third component may be disposed between the two components.

[0048] Terms such as first and second may be used to distinguish one component from another component, and the components are not limited by the above terms.

[0049] Singular forms may include plural forms unless the context clearly indicates otherwise.

[0050] In each step, reference symbols are used for convenience of description, so that the reference symbols do not describe the order of each step, and unless a specific order is clearly described in the context, the steps may be performed in an order different from the specified order.

[0051] Hereinafter, working principles and exemplary embodiments of the present invention will be described with reference to the accompanying drawings.

[0052] Figure 1 A braking system included in a vehicle is shown according to an exemplary embodiment.

[0053] Reference Figure 1 , a vehicle 1 includes: a body that forms its exterior and accommodates a driver and / or luggage; a chassis that includes parts of the vehicle 1 other than the body; and wheels 2 that rotate to enable the vehicle to move.

[0054] The vehicle 1 includes a brake system 40 that generates a braking torque to stop the vehicle 1 and includes an electromechanical brake unit 100 and a brake control unit (BCU) 200.

[0055] Among the front wheel FR, the front wheel FL, the rear wheel RR, and the rear wheel RL of the vehicle 1, the electromechanical brake unit 100 that generates a braking force can be provided, but the present application is not limited thereto, and the electromechanical brake unit 100 can be provided only in the rear wheels. In the present exemplary embodiment, an example in which the electromechanical brake unit 100 is provided in all of the front and rear wheels will be described.

[0056] The brake control unit 200 includes a controller ECU (to be described below Figure 2 that generates a control signal to control the operation of the electromechanical brake unit 100 depending on a braking intention according to the manipulation of the brake pedal P by the driver, thereby generating a braking force.

[0057] The electromechanical brake unit 100 provided in the wheel 2 can include an electromechanical brake 110 and an electronic control unit 120 that controls the operation of the electromechanical brake 110 according to the control signal of the controller 210. The electromechanical brake 110 can include a bracket having a pair of pads provided to pressurize a brake disc rotating together with the wheel 2 of the vehicle 1, a caliper housing slidably mounted in the bracket to operate the pair of pads, a piston mounted in the caliper housing to move forward and backward, an actuator that generates and provides a driving force to move the piston, a power conversion unit that receives a rotational driving force provided from the actuator, converts it into linear motion, and transmits it to the piston to achieve axial forward and backward movement of the piston, and a detection unit that measures adhesion or fastening force between the brake disc 13 and the brake pad 10.

[0058] Specifically, the actuator can be provided to include a reduction gear having a motor and a plurality of reduction gears, and generate a driving force using power supplied from a power device provided in the vehicle 1. The actuator is connected to the other end of the main shaft to transmit the generated driving force to the rotational motion of the main shaft. The actuator can be mounted outside the caliper housing, and the reduction gear can be applied in various structures, such as a planetary gear assembly or a worm structure, to reduce the power of the motor and provide it to the main shaft. The motor can rotate the main shaft to allow the nut to move forward and backward, thereby pressurizing or depressurizing the piston.

[0059] The motor can include a rotating shaft rotatably provided. The motor can include a rotor connected to the rotating shaft and a stator fixed to the housing. For example, the rotor can include a permanent magnet in which N poles and S poles are alternately provided along an outer surface, and the stator can include a plurality of teeth provided along the outer surface of the rotor and a plurality of coils surrounding each of the plurality of teeth.

[0060] The rotor can rotate by magnetic interaction with the stator and can provide the rotation of the rotor to the rotating shaft. The motor can receive a driving current from the brake control unit 200. The plurality of coils included in the stator can form a magnetic field that rotates around the rotor through the driving current, and the rotor can rotate by magnetic interaction between the magnetic fields of the rotor and the stator.

[0061] According to various exemplary embodiments, the motor is a three-phase motor and, for example, may be a three-phase EMB motor, but is not limited thereto. The three-phase EMB motor may include a three-phase coil (u-phase, v-phase, w-phase) mounted on the stator side and a permanent magnet magnetized on the rotor side. The drive circuit of the EMB motor causes current to flow to each phase of the coil of the stator of the three-phase EMB motor, and the rotor of the motor rotates by a magnetic field based on the current provided from the drive circuit. In order to continuously rotate the rotor of the motor in one direction, a switching element that detects the position of the rotor (the magnitude of the rotor magnetic field) and switches the direction of current flow in each phase of the coil according to the detected rotor position may be turned on or off in sequence.

[0062] In the present exemplary embodiment, it has been described that the electromechanical brake unit 100 is a caliper-type EMB, but is not limited thereto, and the electromechanical brake unit 100 may be a drum-type EMB.

[0063] Figure 2 A control configuration of a brake system according to an exemplary embodiment is shown.

[0064] like Figure 2 As shown, the brake system 40 may include a sensor module S, an electromechanical brake unit 100 arranged to correspond to the wheel 2 of the vehicle 1, and a brake control unit 200 arranged to control the operation of the electromechanical brake unit 100. Figure 2 The illustrated configuration in each of the sensor module S, the brake control unit 200 , and the electromechanical brake unit 100 does not correspond to an essential configuration, and some of them may be omitted.

[0065] The brake control unit 200 may control the braking of each wheel 2 based on data obtained through the sensor module S.

[0066] The sensor module S may include a brake pedal sensor 130 , a motor current sensor 140 , a force sensor 150 , and a wheel speed sensor 160 .

[0067] The brake pedal sensor 130 can detect the distance traveled by the brake pedal P, the travel speed, and / or the pedal force applied to the brake pedal P according to the driver's braking intention, and provide a detection signal corresponding to the detected travel distance, travel speed, and / or pedal force to the brake control unit 200. The brake control unit 200 can determine the driver's braking request based on the sensing signal of the brake pedal sensor 130.

[0068] The motor current sensor 140 can measure the current value of each phase (U, V, and W) of the motor. For example, the motor current sensor 140 can measure the current of two of the U, V, and W terminals of the motor. The motor current sensor 140 can be directly connected to the controller 210 via a hard wire or connected to the controller 210 via a communication network, and provides an electrical signal (motor current signal) corresponding to the measured current value to the brake control unit 200.

[0069] The force sensor 150 is provided corresponding to each electromechanical brake unit 100 to measure (or sense) adhesion or fastening force between the brake disc and the brake pad, and provides a sensing signal corresponding to the measured adhesion or fastening force to the brake control unit 200 .

[0070] The wheel speed sensor 160 can measure the rotational speed of each wheel 2 provided in the vehicle 1. The wheel speed sensor 160 is provided in each of the plurality of wheels and can measure the rotational speed of each of the plurality of wheels. For example, the wheel speed sensor 160 may include a Hall sensor that detects a magnetic field and its changes, or a coil that detects a change in a magnetic field.

[0071] The wheel speed sensor 160 may provide the brake control unit 200 with a sensing signal (speed signal) corresponding to the measured rotation speed.

[0072] The brake control unit 200 may provide a control signal to the electromechanical brake unit 100 according to a driver's brake request to allow the vehicle 1 to brake.

[0073] According to the disclosed exemplary embodiment, the controller 210 of the brake control unit 200 may include a processor 211 and a memory 212 .

[0074] The processor 211 may provide a control signal to control the operation of the structures included in the electromechanical brake unit 100 according to the driver's braking request.

[0075] The memory 212 may store or memorize programs and data to implement operations of controlling the structures included in the electromechanical brake unit 100 .

[0076] The memory 212 provides stored programs and data to the controller 210 and may store temporary data generated during the operation of the controller 210. For example, the memory 212 may include a volatile memory such as a static random access memory (S-RAM) or a dynamic random access memory (D-RAM), and a nonvolatile memory such as a read-only memory (ROM), an erasable programmable read-only memory (EPROM), or a flash memory.

[0077] The processor 211 may be electrically connected to the sensor module S.

[0078] The processor 211 processes the electrical signals received from the brake pedal sensor 130 , the motor current sensor 140 , the force sensor 150 , and the wheel speed sensor 160 , and may provide a brake control signal to the electromechanical brake unit 100 based on the processed electrical signals.

[0079] According to the disclosed exemplary embodiment, processor 211 obtains a motor current signal of the motor of electromechanical brake unit 100 from motor current sensor 140 and can determine the motor status based on the obtained motor current signal. Processor 211 can control the operation of electromechanical brake unit 100 based on the determined motor status.

[0080] Hereinafter, an operation of determining the motor status in the controller 210 according to the disclosed exemplary embodiment will be described in detail.

[0081] Figure 3 Functional modules of a controller included in a brake control unit according to an exemplary embodiment are shown.

[0082] Reference Figure 3 Controller 210 may functionally include multiple modules. Each module may be a hardware module (e.g., ASIC or FPGA) included in processor 211 or a software module (e.g., application or data) stored in memory 212. In this exemplary embodiment, an example in which a driver's braking request is present will be described.

[0083] like Figure 3 As shown, the controller 210 may include a state determination module 300 , a fault level determination module 310 , and a control module 320 .

[0084] The state determination module 300 receives a motor current signal from the motor current sensor 140 and can determine a motor state based on the received motor current signal. Here, a current signal of one motor can include a signal representing a current value measured in two or more of three phases (at least two of U, V, and W) of a motor (e.g., a three-phase motor). To determine a state of one motor, the state determination module 300 can use a machine learning model 305 trained to predict a motor state based on a motor current signal.

[0085] The state determination module 300 determines a plurality of prediction factors for predicting a motor state by converting a motor current signal obtained from the motor current sensor 140, and can use the determined prediction factors as input data of the machine learning model 305.

[0086] For example, when a two-phase current signal obtained from the motor current sensor 140 includes a U-phase current signal (e.g., a second-phase signal) and a V-phase current signal (e.g., a first-phase signal), the state determination module 300 can estimate a remaining phase current (W-phase current) signal using the two-phase current signal and convert a three-phase current signal into two-phase signals. In other words, the state determination module 300 performs a direct-quadrature (DQ) transformation on a U-W-V phase current signal to obtain a DQ-axis current signal.

[0087] The state determination module 300 converts UV-phase and DQ-axis current signals into harmonic signals using a fast Fourier transform (FFT), and can determine input data for predicting a motor state based on at least one of the peaks of the harmonic signals that is above a predetermined threshold. When all of the peaks of the harmonic signals obtained by the fast Fourier transform are used as input data, the use of memory increases, and thus, to prevent the increased use of memory, at least one of the peaks that is above a threshold can be used.

[0088] To determine the input data, the state determination module 300 can perform an analysis of variance on at least one of the peaks of the harmonic signals that is above a predetermined threshold. For example, the analysis of variance can be an analysis of variance (ANOVA) or a one-way ANOVA, but is not limited thereto.

[0089] The state determination module 300 can determine data for predicting a motor state among data (e.g., frequency data and amplitude data) related to a plurality of current signals (UV-phase and DQ-axis current signals) that are input data through an analysis of variance on at least one peak. For example, the state determination module 300 can obtain F-statistics and P-values of data (frequency data and amplitude data of UV-phase current signals and frequency data and amplitude data of DQ-axis current signals) through the analysis of variance, as shown in Table 1.

[0090] [Table 1]

[0091]

[0092] As shown in Table 1, when the P value of each data is obtained, the state determination module 300 can determine the frequency data and amplitude data of the U-phase current signal, the frequency data and amplitude data of the V-phase current signal, and the frequency data of the Q-axis current signal that are below a predetermined threshold value (for example, 0.05) as input data.

[0093] Will refer to Figure 4 The operation of the state determination module 300 is described illustratively as predicting the motor state using a machine learning model 305 trained to predict the motor state based on the input data determined as described above.

[0094] Figure 4 An example diagram is shown for explaining the prediction work of a machine learning model according to an exemplary embodiment.

[0095] Reference Figure 4 , shows a machine learning model 305 that is trained to predict the state of the motor using the above-mentioned input data 400 as input and outputting prediction result data 410.

[0096] The state determination module 300 may input input data (i.e., frequency data and amplitude data of the UV phase current signal and frequency data of the Q axis current signal) 400 to the machine learning model 305 and obtain prediction result data 410 using the machine learning model 305. Here, the prediction result data 410 may include a label and a predicted probability value of at least one of a normal state, a first fault state (bearing fault), a second fault state (rotor imbalance), and a third fault state (shaft offset).

[0097] According to an exemplary embodiment, the machine learning model 305 may be based on various algorithms for predicting or estimating data. Here, the various algorithms may be machine learning-based algorithms for predicting or estimating data. For example, the machine learning model 305 may be based on K-NN, Naive Bayes, Support Vector Machine (SVM), Decision Tree, Multilayer Perceptron (MLP), Random Forest, or a combination thereof, but is not limited thereto, and may be based on an artificial neural network model such as a Convolutional Neural Network (CNN), a Recurrent Neural Network (RNN), and / or a Long Short-Term Memory (LSTM).

[0098] In order to determine the motor state based on the prediction result data 410, the state determination module 300 can determine a label corresponding to a predetermined threshold probability value or more (rather than a predicted probability value corresponding to a label corresponding to a normal state) among multiple labels as a motor state. In other words, the state determination module 300 assigns (or applies) a weight to the predicted probability value corresponding to the label corresponding to the normal state to calculate the threshold probability value, and can determine the label corresponding to the generated threshold probability value or more as the motor state. For example, when the predicted probability value corresponding to the label corresponding to the normal state is 30%, the state determination module 300 applies a weight of 15% to the predicted probability value corresponding to the label corresponding to the normal state to generate a threshold probability value of 45%. The state determination module 300 determines the label corresponding to the threshold probability value or more of 45% as the motor state, so that when there is no label corresponding to the threshold probability value or more of 45%, the state determination module 300 can determine the motor state as the normal state.

[0099] When the predicted probability value corresponding to the bearing fault among the multiple fault labels is greater than 45%, the state determination module 300 may determine the motor state as a bearing fault.

[0100] When the motor state is determined as described above, the state determination module 300 may provide motor state data representing the determined motor state to the fault level determination module 310 .

[0101] Figure 5 An example diagram for explaining the learning operation of a machine learning model according to an exemplary embodiment is shown.

[0102] Reference Figure 5 , the learning data 500 used to train the machine learning model 305 may include: input data 510 for each of the normal state, the first fault state, the second fault state and the third fault state; and the correct answer state label 520, used as the correct answer data (i.e., the benchmark true value).

[0103] The state determination module 300 can train the machine learning model 305 to input the input data 510 for training the model obtained to correspond to each state of the motor into the machine learning model 305, predict the motor state based on the input data 510, and output a predicted state label 530 representing the predicted motor state.

[0104] Specifically, the state determination module 300 may calculate a loss 540 of the predicted state label 530, and train the machine learning model 305 to minimize the calculated loss 540 while updating the weights of the machine learning model 305. Here, the loss 540 may refer to the difference between the predicted state label 530, which is the output value of the machine learning model 305, and the correct answer state label 520 corresponding to the ground truth value. To calculate the loss, the state determination module 300 may use a loss function such as mean square error (MSE) loss, perceptual loss, structural similarity index (SSIM) loss, and / or VGG loss, but is not limited thereto, and may use various loss functions to allow the difference between two values ​​to converge to a minimum threshold, such as "0."

[0105] Refer again Figure 3 When the state determination module 300 determines that the motor state is a fault, the fault level determination module 310 may determine the fault level based on the above-mentioned DQ current signal together with the braking force (clamping force).

[0106] To determine the motor performance level, the fault level determination module 310 may collect various motor-related data and determine factors associated with motor performance degradation from the collected data. For example, the fault level determination module 310 may perform a Pearson correlation analysis between the collected data and motor performance degradation to obtain a Pearson correlation coefficient, as shown in Table 2.

[0107] [Table 2]

[0108] grade Analytical factors Correlation coefficient Remark 1 Braking force (clamping force) 1 2 Q-axis current 0.887 The main factors affecting motor torque 3 D-axis current 0.881 The main factors affecting motor speed

[0109] As shown in Table 2, when the correlation coefficient of each factor is obtained, the fault level determination module 310 may determine the braking force, the Q-axis current signal, and the D-axis current signal associated with motor performance degradation.

[0110] The fault level determination module 310 can determine a fault level corresponding to at least one of a first fault state, a second fault state, and a third fault state based on a signal associated with the measured braking force, a Q-axis current signal, and a D-axis current signal obtained by the force sensor 150. For example, when the fault state of the motor is a "bearing fault," the D-axis current and the Q-axis current increase, but the braking force decreases. In this case, if the difference between the obtained D-axis current value and the D-axis current value corresponding to the normal value is equal to or higher than a predetermined threshold, the difference between the Q-axis current value and the Q-axis current value corresponding to the normal value is equal to or higher than a predetermined threshold, and the difference between the obtained braking force and the braking force corresponding to the normal value is lower than a predetermined threshold, the fault level determination module 310 can determine the fault level of the bearing fault as a first fault level (e.g., a higher fault level).

[0111] According to various exemplary embodiments, when the motor's fault state is "rotor imbalance," the D-axis current and Q-axis current increase, but the braking force decreases. In this case, if the difference between the obtained D-axis current value and the D-axis current value corresponding to the normal value is equal to or greater than a predetermined threshold, the difference between the obtained Q-axis current value and the Q-axis current value corresponding to the normal value is equal to or greater than a predetermined threshold, and the difference between the obtained braking force and the braking force corresponding to the normal value is lower than a predetermined threshold, the fault level determination module 310 may determine the rotor imbalance fault level as a first fault level.

[0112] According to various exemplary embodiments, when the motor fault state is "axle offset," all D-axis current, Q-axis current, and braking force are reduced. In this case, if the difference between the obtained D-axis current value and the D-axis current value corresponding to the normal value is lower than a predetermined threshold, the difference between the obtained Q-axis current value and the Q-axis current value corresponding to the normal value is lower than a predetermined threshold, and the difference between the obtained braking force and the braking force corresponding to the normal value is lower than a predetermined threshold, the fault level determination module 310 may determine the fault level of the axis offset as the first fault level.

[0113] Furthermore, the fault level determination module 310 may determine whether the measured braking force reaches the target braking force within a predetermined first threshold time, and if the measured braking force does not reach the target braking force within the first threshold time, may determine the fault level to be a second fault level (e.g., a lower fault level). According to various exemplary embodiments, the fault level determination module 310 may determine whether the braking response time when the measured braking force reaches the target braking force falls within a predetermined second threshold time, and if the braking response time when the measured braking force reaches the target braking force exceeds the second threshold time, may determine the fault level to be the second fault level. According to various exemplary embodiments, when the driver requests posture control, the fault level determination module 310 may determine whether the braking force reaches the driver's requested braking force, and if the braking force does not reach the driver's requested braking force, may determine the fault level to be the second fault level. According to various exemplary embodiments, the fault level determination module 310 may determine the second fault level by considering both the time to reach the target braking force and the braking response time, but is not limited thereto, and the criteria for determining the second fault level may vary depending on the performance of the vehicle. According to various exemplary embodiments, the second fault level may have a fault level lower than the first fault level.

[0114] The control module 320 may generate a control signal for controlling the operation of the electromechanical brake unit 100 based on the fault level determined by the fault level determination module 310 , and transmit the generated control signal to the electromechanical brake unit 100 .

[0115] Specifically, the control module 320 may determine a target braking force based on the driver's braking request and calculate a requested torque for each wheel 2 to achieve the determined target braking force. In other words, when a failure of a motor provided in at least one of the multiple wheels is predicted, the control module 320 may calculate a requested torque for each wheel 2 based on the failure level of the motor predicted to fail. In other words, the control module 320 may redistribute the requested torque based on the performance level of the other motors according to the performance level of the motor predicted to fail. According to various exemplary embodiments, the controller 210 may determine that the requested torque value for the wheel corresponding to at least one motor predicted to fail in the electromechanical brake unit is lower than a predetermined threshold requested torque value. Furthermore, the controller 210 may determine that one or more requested torque values ​​for the wheels corresponding to another motor not predicted to fail or another motor in the electromechanical brake unit are equal to or higher than a predetermined threshold requested torque value.

[0116] Will refer to Figures 6 to 9 Operation of a control module for calculating a requested torque for each wheel based on a fault level of a motor is described.

[0117] Figure 6 and Figure 7 is an exemplary diagram for explaining the operation of a control module when a motor of a first wheel is predicted to fail according to an exemplary embodiment. Figure 8 and Figure 9 An example diagram for explaining the operation of a control module when motors of first and third wheels are predicted to fail according to an exemplary embodiment is shown.

[0118] Reference Figure 6 If a fault of the first wheel is predicted and the fault level of the first wheel is the first fault level, the control module 320 may calculate a maximum torque value that the first wheel can generate when the first fault level is reached, and determine the calculated maximum torque value as the requested torque value of the first wheel. Figure 6 As shown in (a) of FIG. 3 , the control module 320 may reduce the maximum allowable value Mx of the torque that can be generated by the first wheel to a predetermined first threshold value Thd1, and as shown in FIG. Figure 6As shown in (b), (c), and (d) of FIG, the maximum permissible value Mx of torque that can be generated by the remaining wheels (the second, third, and fourth wheels) can be increased to a predetermined second threshold value Thd2. Because the torque that can be generated by the first wheel is lower than that which can be generated under normal conditions, the control module 320 reduces the maximum permissible torque value that can be generated by the first wheel and allows the remaining wheels to generate torque values ​​higher than those under normal conditions. Therefore, even if a failure of the motor provided in the first wheel is predicted, the electromechanical brake unit of each wheel can be controlled to generate the target braking force according to the driver's braking request.

[0119] Reference Figure 7 If a fault of the first wheel is predicted and the fault level of the first wheel is the first fault level, the control module 320 may calculate a maximum torque value that the first wheel can generate when the fault level is the second fault level, and determine the calculated maximum torque value as the requested torque value of the first wheel. Figure 7 As shown in (a) of FIG. 3 , the control module 320 may reduce the maximum allowable value Mx of the torque that can be generated by the first wheel to a third threshold value Thd3 that is lower than the first threshold value Thd1, and as shown in FIG. Figure 7 As shown in (b), (c), and (d) of FIG. , the maximum permissible torque value Mx that can be generated by the remaining wheels can be increased to a fourth threshold value Thd4 that is higher than the second threshold value Thd2. Because the torque that can be generated by the first wheel is lower than the torque that can be generated at the first fault level, the control module 320 reduces the maximum permissible torque value that can be generated by the first wheel to a value lower than the maximum permissible torque value at the first fault level and increases the maximum permissible torque values ​​of the remaining wheels to a value higher than the maximum permissible torque value at the first fault level. Therefore, even if the performance of the motor provided in the first wheel is degraded, the electromechanical brake unit of each wheel can be controlled to generate the target braking force according to the driver's braking request.

[0120] Reference Figure 8 If the failure of the first wheel and the third wheel is predicted and the failure level of the first wheel and the third wheel is the first failure level, the control module 320 may calculate the maximum torque value that the first wheel and the third wheel can generate when the failure level is the first, and determine the calculated maximum torque value as the requested torque value of the first wheel and the third wheel. In this case, Figure 8 As shown in (a) and (c) of FIG. 3 , the control module 320 may reduce the maximum allowable value Mx of the torque that the first wheel and the third wheel can generate to a predetermined fifth threshold value Thd5, and as shown in FIG. Figure 8As shown in (b) and (d) of FIG. , the maximum permissible value Mx of torque that can be generated by the remaining wheels (the second and fourth wheels) can be increased to a predetermined sixth threshold value Thd6. Because the torque that can be generated by the first and third wheels is lower than that which can be generated under normal conditions, the control module 320 reduces the maximum permissible torque value that can be generated by the first and third wheels and allows the remaining wheels to generate a torque value higher than that under normal conditions. Therefore, even if a failure of the motors provided in the first and third wheels is predicted, the electromechanical brake unit of each wheel can be controlled to generate the target braking force according to the driver's braking request.

[0121] Reference Figure 9 If the failure of the first wheel and the third wheel is predicted and the failure level of the first wheel and the third wheel is the second failure level, the control module 320 may calculate the maximum torque value that the first wheel and the third wheel can generate when the second failure level is reached, and determine the calculated maximum torque value as the requested torque value of the first wheel and the third wheel. In this case, Figure 9 As shown in (a) and (c) of FIG. 3 , the control module 320 may reduce the maximum allowable value Mx of the torque that can be generated by the first wheel and the third wheel to a seventh threshold value Thd7 that is lower than the fifth threshold value Thd5, and as shown in FIG. Figure 9 As shown in (b) and (d) of FIG. , the maximum permissible value Mx of the torque that can be generated by the remaining wheels can be increased to an eighth threshold value Thd8 that is higher than the sixth threshold value Thd6. Because the torque that can be generated by the first and third wheels is lower than the torque that can be generated at the first fault level, the control module 320 reduces the maximum permissible torque value that can be generated by the first and third wheels to a value lower than the maximum permissible torque value at the first fault level and increases the maximum permissible torque value of the remaining wheels to a value higher than the maximum permissible torque value at the first fault level. Therefore, even if the performance of the motors provided in the first and third wheels is reduced, the electromechanical brake unit of each wheel can be controlled to generate the target braking force according to the driver's braking request.

[0122] Figure 10 A torque control operation of a controller included in a brake control unit according to an exemplary embodiment is shown.

[0123] Reference Figure 10 , the controller 210 can predict the state of at least one motor based on the motor current signal obtained from the motor current sensor 140 (step S1000). In other words, the controller 210 can predict the state of the motor of the electromechanical brake unit based on the current signal of the motor detected by the motor current sensor 140. Figure 11 The specific operation of the controller 210 for predicting the motor state will be described.

[0124] Figure 11 The operation of a controller for predicting the motor state of each wheel is shown according to an exemplary embodiment.

[0125] Reference Figure 11 , the controller 210 obtains two phase current signals from the motor current sensor 140 of the three phase current signals of at least one motor (step S1100). For example, the controller 210 may obtain the UV phase current signal from the UVW phase current signal through FFT. The motor current signal may include two current signals with different phases from the three phase current signals of the motor.

[0126] The controller 210 may generate input data for a machine learning model 305 trained to predict the motor state based on the obtained two-phase current signals (step S1110). Specifically, the controller 210 may use the UV current signal to generate a W-phase current signal, and perform a DQ transformation on the UVW-phase current signal to generate a DQ-axis current signal. In other words, the controller 210 may use two current signals with different phases to generate the remaining phase current signals in the three-phase current signal, and may perform a DQ transformation on the three-phase current signal to obtain the DQ-axis current signal.

[0127] The controller 210 may determine the frequency data and amplitude data of the U-phase current signal, the frequency data and amplitude data of the V-phase current signal, and the frequency data of the Q-axis current signal among the UV-phase current signal and the DQ-axis current signal as input data for predicting the motor state through ANOVA analysis.

[0128] The controller 210 can use the machine learning model 305 to predict the motor state based on the generated input data (step S1120). Figure 4 As described above, the controller 210 may input the determined input data 400 into the machine learning model 305, and use the machine learning model 305 to obtain prediction result data 410, which indicates the predicted motor state among the normal state, the first fault state, the second fault state and the third fault state of the motor.

[0129] Refer again Figure 10, the controller 210 determines whether at least one motor is predicted to fail (step S1010), and if at least one motor is predicted to fail, the controller 210 may determine the failure level of the at least one motor based on the sensor data obtained from the sensor module S (step S1020). In other words, when at least one predicted state of the motor indicates that at least one motor has failed, the controller 210 may determine the failure level of the at least one failed motor based on the sensor data obtained from the sensor module. Here, the sensor module S may include a motor current sensor 140 and a force sensor 150, and the sensor data may include a two-phase current signal obtained by the motor current sensor 140 and a signal (braking force measurement signal) representing the braking force measured by the force sensor 150 (i.e., a signal associated with the braking force measured by the force sensor 150).

[0130] Specifically, the controller 210 may perform a DQ transformation on the two-phase current signals to obtain DQ axis current signals, and may determine the fault level of at least one motor by taking into account the correlation between the DQ axis current signals and the braking force measurement signals and the motor performance degradation. For example, if the difference between each signal and the normal signal is equal to or higher than a predetermined first threshold difference, the controller 210 may determine the fault level to be a first fault level. According to various exemplary embodiments, the controller 210 may determine whether the target braking force, the braking response, and the requested braking force are achieved, and if at least one of them is not met, the controller 210 may determine the fault level to be a second fault level.

[0131] As described above, when the fault level of at least one motor is determined, the controller 210 may calculate the requested torque of each wheel based on the determined fault level (step S1030), and control the torque of each wheel based on the calculated requested torque (step S1040). In other words, the controller 210 may calculate the requested torque of each wheel based on the determined fault level of at least one motor that has failed. To calculate the requested torque of each wheel, as described above in Figures 6 to 9 As described in the foregoing, the controller 210 may reduce the maximum allowable torque value of the wheel corresponding to the motor predicted to fail according to the failure level of at least one motor (i.e., the first failure level or the second failure level) to a value lower than the reference value (normal value), and may increase the maximum allowable torque value of the wheel corresponding to the motor in a normal state to a value higher than the reference value. Therefore, even if a failure occurs in at least one motor, the target braking force according to the driver's braking request may be generated. According to various exemplary embodiments, the reference Figure 12 The operation of the controller is described when the fault level of at least one motor is determined to be the second fault level.

[0132] Figure 12An operation of the controller is shown when a fault level of at least one motor is determined to be a second fault level according to an exemplary embodiment.

[0133] Reference Figure 12 ,from Figure 10 (a) advance to Figure 12 The controller 210 of (a) determines whether the fault level is above the second fault level (step S1200). If the determined fault level of at least one motor in which the fault occurs is above the second fault level, the controller may output information indicating the fault level of the motor through a display device provided in the vehicle (step S1210), and Figure 12 (b) advance to Figure 10 (b). In other words, the controller 210 may control the display to output information indicating the determined fault level of at least one motor that has failed. For example, the controller 210 may output the information through a human-machine interface (HMI) provided in the vehicle. According to various exemplary embodiments, the controller 210 may output the information in a voice through an audio device to allow the driver to recognize it, but is not limited thereto, and various methods that allow the driver to recognize it may be used.

[0134] Refer again Figure 10 If at least one motor is not predicted to fail or at least one motor is predicted to be normal, the controller 210 may calculate the requested torque of each wheel according to the driver's braking request (step S1050) and control the torque of each wheel based on the calculated requested torque (step S1040).

[0135] As described above, according to the disclosed exemplary embodiments, motor failure is predicted in advance based on the three-phase current signals of the three-phase motor to prevent safety accidents of the vehicle and minimize damage.

[0136] Furthermore, the torque requested for the wheels corresponding to the predicted motor and the torque requested for the wheels corresponding to the remaining motors are adjusted based on the failure level of the at least one motor predicted to fail. Therefore, even if a failure of at least one motor is predicted, the target braking force according to the driver's brake request can be generated.

[0137] Furthermore, motor failure can be predicted based on the three-phase current signals of the motor current sensor without adding a sensor for predicting motor failure.

[0138] Furthermore, the safety of the diagnostic system can be provided by pre-diagnosis of the motor rather than post-diagnosis.

[0139] In addition, the disclosed exemplary embodiments may be implemented as a recording medium storing computer-executable instructions. The instructions may be stored as program code, and when executed by a processor, a program module is generated to perform the operations of the disclosed exemplary embodiments. The recording medium may be implemented as a computer-readable recording medium.

[0140] Computer-readable recording media include all types of recording media in which instructions that can be deciphered by a computer are stored. For example, recording media may include read-only memory (ROM), random access memory (RAM), magnetic tapes, magnetic disks, flash memory, and optical data storage devices.

[0141] The machine-readable storage medium may be provided as a non-transitory storage medium. Here, "non-transitory" means that the storage medium is a tangible device and does not contain signals (e.g., electromagnetic waves), and the term does not distinguish between situations where data is stored semi-permanently or temporarily on the storage medium. For example, a "non-transitory storage medium" may include a cache that temporarily stores data.

[0142] As described above, the disclosed exemplary embodiments have been described with reference to the accompanying drawings. Those skilled in the art will appreciate that the present invention may be implemented in forms other than the disclosed embodiments without changing the technical spirit or essential features of the present disclosure. The disclosed exemplary embodiments are illustrative and should not be construed as limiting.

Claims

1. A braking system, wherein: include: Sensor module, including motor current sensor and force sensor, an electromechanical brake unit, each mounted on a wheel of the vehicle and comprising a motor, and a controller configured to control one or more of the electromechanical brake units; The controller is configured to: predicting a state of the motor of the electromechanical brake unit based on a current signal of the motor detected by the motor current sensor, determining a fault level of at least one of the motors that has failed based on sensor data obtained from the sensor module when at least one predicted state of the motors indicates that at least one of the motors has failed, calculating a requested torque for each of said wheels based on a determined fault level of at least one of said motors that has failed, The torque of each of the wheels is controlled based on the calculated requested torque of each of the wheels.

2. The braking system according to claim 1, wherein: The state of the motor includes a normal state, a bearing fault state, a rotor imbalance state, or a shaft offset state.

3. The braking system according to claim 2, wherein: The current signal of the motor includes two current signals having different phases among the three-phase current signals of the motor, The controller is configured to control to use the two current signals with different phases to generate remaining phase current signals in the three-phase current signal, perform direct orthogonal transformation on the three-phase current signal to obtain a direct orthogonal axis current signal, and generate input data for a machine learning model based on the three-phase current signal and the direct orthogonal axis current signal.

4. The braking system according to claim 3, wherein: The controller is configured to obtain prediction result data representing the predicted motor state using the machine learning model trained to predict the motor state based on the input data, wherein the input data includes frequency data and amplitude data of the first phase current signal, frequency data and amplitude data of the second phase current signal, and frequency data of the orthogonal axis current signal in the three-phase current signal.

5. The braking system according to claim 4, wherein: The prediction result data includes a plurality of labels and a predicted probability value for each of the plurality of labels, the plurality of labels including a label corresponding to the normal state, a label corresponding to the bearing fault state, a label corresponding to the rotor imbalance state, and a label corresponding to the shaft offset state.

6. The braking system according to claim 5, wherein: The controller is configured to determine, among the plurality of labels, a label having the predicted probability value corresponding to the label corresponding to the normal state, the predicted probability value being equal to or higher than a threshold probability value, as the motor state.

7. The braking system according to claim 4, wherein: the sensor data comprising a signal associated with the braking force measured by the force sensor, The controller is configured to determine the fault level based on a signal associated with the measured braking force and the direct quadrature axis current signal.

8. The braking system according to claim 7, wherein: The fault level includes a higher fault level and a lower fault level, the lower fault level being lower than the higher fault level, The controller is configured to control the display to output information indicating the determined fault level of the at least one motor that has failed, if the determined fault level of the at least one motor that has failed is the lower fault level of the higher fault level and the lower fault level.

9. The braking system according to claim 8, wherein: The controller is configured to: determining whether the measured braking force reaches the target braking force within a predetermined first threshold time, If the measured braking force does not reach the target braking force within the first threshold time, the fault level is determined to be the lower fault level of the higher fault level and the lower fault level.

10. A method of controlling a braking system, wherein: The method comprises the following steps: predicting the state of the motor of the electromechanical brake unit based on a current signal of the motor detected by a motor current sensor, determining a fault level of at least one of the motors that has failed based on sensor data obtained from a sensor module including the motor current sensor when at least one predicted state of the motors indicates that at least one of the motors has failed, calculating a requested torque for each wheel based on the determined fault level of at least one of the motors that has failed, and The torque of each of the wheels is controlled based on the calculated requested torque of each of the wheels.

Images