Method and apparatus for controlling NVH of electric compressor of vehicle

Abstract

The invention relates to a method for predictively controlling NVH of an electric compressor of a vehicle based on NVH characteristics of the electric compressor of the vehicle. The method comprises the steps of obtaining data related to an operating state of the vehicle; generating, by the NVH feature prediction model, an NVH feature of the electric compressor based on the data related to the operating state of the vehicle; generating, by an NVH control model, an NVH control parameter based on the generated NVH feature; operation of the electric compressor is adjusted based on the generated NVH control parameter.

Description

Technical Field

[0001] The present application generally relates to noise management of electric vehicles, and more particularly, to methods and devices for controlling the noise, vibration, and harshness (NVH) of an electric compressor of a vehicle. Background Art

[0002] Electric vehicles generally use an electric compressor to achieve temperature regulation, such as regulating the cabin temperature in hot weather, or cooling the battery during rapid charging. Since there is no masking by the noise of a traditional internal combustion engine, the in-vehicle noise generated by the vibration of the electric compressor is more easily perceived by users.

[0003] An electric compressor generally generates more NVH than a compressor in an internal combustion engine vehicle because the electric compressor bears a greater load when circulating coolant to control the cabin temperature. In addition, an electric compressor generally operates in different speed ranges and excitations, resulting in more NVH. During operation, when cooling capacity is generated, mechanical vibrations of the electric compressor, pipes, or ventilation systems may generate undesirable noise.

[0004] Therefore, it is necessary to effectively eliminate or suppress the NVH of the electric compressor in an electric vehicle. Summary of the Invention

[0005] The following introduction is provided to introduce some selected concepts in a simple form, which will be further described in the detailed description below. This introduction is not intended to highlight the key features or essential features of the claimed subject matter, nor is it intended to limit the scope of the claimed subject matter.

[0006] According to one aspect of the present application, there is provided a method for controlling the NVH of an electric compressor of a vehicle, including: obtaining data related to the operating state of the vehicle; generating the NVH characteristics of the electric compressor by an NVH characteristic prediction model based on the data related to the operating state of the vehicle; generating an NVH control parameter by an NVH control model based on the generated NVH characteristics; and adjusting the operation of the electric compressor based on the generated NVH control parameter.

[0007] According to one aspect of the present application, there is provided a method for training a neural network model for controlling the NVH of an electric compressor of a vehicle, including: generating an NVH control parameter by the NVH control model in the neural network model based on the NVH characteristics of the electric compressor; adjusting the operation of the electric compressor based on the generated NVH control parameter; detecting the NVH characteristics of the adjusted electric compressor; and updating the NVH control model based on the detected NVH characteristics.

[0008] According to one aspect of the present application, there is provided a device for controlling the NVH of an electric compressor of a vehicle, including: a data acquisition module for acquiring data related to the operating state of the vehicle; an NVH characteristic prediction module for generating the NVH characteristics of the electric compressor based on the data related to the operating state of the vehicle; and an NVH control module for generating NVH control parameters based on the generated NVH characteristics, and adjusting the operating parameters of the electric compressor based on the generated NVH control parameters.

[0009] According to one aspect of the present application, there is provided a processing device for the NVH of an electric compressor of a vehicle, including: one or more processors and a memory storing computer-executable instructions, and the instructions, when executed, cause the one or more processors to execute the methods according to various embodiments of the present disclosure.

[0010] According to one aspect of the present application, there is provided a machine-readable storage medium storing executable instructions, and the instructions, when executed, cause one or more processors to execute the methods according to various embodiments of the present disclosure.

[0011] By using the method for controlling the NVH of an electric compressor of a vehicle according to an embodiment of the present disclosure, by training an NVH characteristic prediction model based on historical big data, it is possible to dynamically extract the NVH characteristics of the electric compressor based on the collected empirical data, thereby solving the difficulty of dynamic measurement of NVH characteristics caused by differences in manufacturing, assembly, and dynamic characteristics of the electric compressor, and associating the NVH characteristics with the input data in the big data pool to achieve early prediction of NVH in different scenarios. In addition, by establishing an NVH control model to generate NVH control parameters based on the extracted NVH characteristics, active and early intervention of NVH control is realized, and the control effect of NVH is further improved. Other advantages of the embodiments of the present application will be described below. Description of the Drawings

[0012] By referring to the following drawings, a further understanding of the essence and advantages of the content of the present application can be achieved. In the drawings, similar components or features may have the same reference numerals.

[0013] Figure 1 Shows a schematic diagram of an electric compressor in a vehicle according to an embodiment.

[0014] Figure 2 Shows a schematic diagram of an electric compressor in a vehicle according to an embodiment.

[0015] Figure 3 Shows a schematic diagram of a system for suppressing the NVH of an electric compressor in a vehicle according to an embodiment.

[0016] Figure 4 Shows a schematic diagram of a system for suppressing NVH of an electric compressor in a vehicle according to an embodiment.

[0017] Figure 5 Shows a schematic diagram of a system for suppressing NVH of an electric compressor in a vehicle according to an embodiment.

[0018] Figure 6 Shows a schematic diagram of a process for training an NVH feature prediction model according to an embodiment.

[0019] Figure 7 Shows a schematic diagram of training an NVH control model according to an embodiment.

[0020] Figure 8 Shows a flowchart of a method for controlling NVH of an electric compressor of a vehicle according to an embodiment.

[0021] Figure 9 Shows a flowchart of a method for training a neural network model for controlling NVH of an electric compressor of a vehicle according to an embodiment.

[0022] Figure 10 Shows a block diagram of a device for controlling NVH of an electric compressor of a vehicle according to an embodiment.

[0023] Figure 11 Shows a block diagram of a processing device according to an embodiment. Detailed Description

[0024] The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that the discussion of these embodiments is only to enable those skilled in the art to better understand and thus implement the subject matter described herein, and is not a limitation on the scope of protection, applicability, or examples set forth in the claims. The functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the content of this application. Each example may omit, substitute, or add various processes or components as needed. For example, the methods described may be performed in a different order than described, and each step may be added, omitted, or combined. Additionally, features described relative to some examples may be combined in other examples.

[0025] Figure 1 and 2 Shows a schematic diagram of an electric compressor in a vehicle according to an embodiment. The same numbers in the figure represent the same or corresponding components.

[0026] As Figure 1As shown, vehicle 100 can be an electric vehicle, which includes a battery 110, an electric compressor 120, a processing module 130, and a sensor 140. It can be understood that Figure 1 only the components of the electric vehicle related to the technical solution of the present disclosure are shown. The battery 110 is used to provide power for each component in the electric vehicle 100, such as providing power for the electric compressor 120. The processing module 130 is the central control component of the electric vehicle, which provides computing resources and control capabilities for the electric vehicle 100 and controls the operations of each component of the electric vehicle. The electric compressor 120 is used to achieve temperature regulation in the electric vehicle, such as refrigeration and heating. The electric compressor 120 mainly includes a motor controller 1210, a motor 1220, and a compressor 1230. The inverter 1240 is used to convert the direct current provided by the battery 110 into alternating current under the control of the motor controller 1210 to drive the motor 1220. The motor 1220 converts electrical energy into mechanical motion to drive the compressor 1230 to move. The operation of the compressor 1230 realizes the circulation of the coolant, thereby enabling the temperature in the electric vehicle to be regulated. The motor controller 1210, the inverter 1240, and the motor 1220 can be collectively referred to as the electric drive components of the electric compressor 120.

[0027] As Figure 2 shown, the motor controller 1210 includes a speed controller 12110 and a current controller 12120. The speed controller 12110 receives a feedback signal from the motor 1220 and a speed reference signal, and generates a reference current I ref , the current controller 12130 receives a feedback signal from the motor 1220 and the reference current I ref , and generates a reference voltage U ref . The inverter 1240 can be controlled based on this reference voltage U ref to convert the direct current provided by the battery 110 into alternating current and output it to the motor 1220. It can be understood that in a specific implementation, the reference voltage U ref is further processed to control the operation of the inverter 1240. For example, the reference voltage U ref is subjected to coordinate transformation to be converted from a two-dimensional coordinate system to a three-dimensional coordinate system, and then, for example, through pulse width modulation (PWM), the control signal of the inverter is obtained. Operations such as the above-mentioned coordinate transformation and PWM modulation are all well-known technologies in the art, so no further detailed description will be given to them. Those skilled in the art can understand that the electric compressor 120 is a common component of the electric vehicle, and the motor controller 1210 and each component therein can be implemented in various known or future-known ways. The technical solution of the present disclosure can be applied to various electric compressors and the electric drive components therein.

[0028] Sensor 140 includes various sensors on the electric vehicle. For example, a current sensor, a speed sensor, etc. are installed on the electric drive assembly of the electric compressor 120. In one example, the current sensor can measure the q-axis current of the motor, and the speed sensor can measure the rotational speed of the motor. In one example, the current sensor can measure the phase current of the motor, and the speed sensor can measure the rotational speed of the motor. Also, for example, a vibration sensor can be installed on the compressor 1230 to measure the torque pulsation of the compressor. Sensor 140 can also include other sensors that may be assembled on the electric vehicle, such as sensors related to weather conditions, radar sensors, image sensors, ultrasonic sensors, and so on.

[0029] The electric compressor 120 is a major NVH source in the electric vehicle 100. The motor controller 1210 controls the high-frequency inverter 1240 to convert direct current into alternating current, and the operation of the inverter 1240 may bring high-frequency noise and vibration at the switching rate. In the motor 1220, torque pulsation generated by AC excitation, cogging torque of the stator teeth, and radial force acting on the housing may all generate vibration. NVH mainly comes from current harmonics and torque pulsation. In the compressor 1230, NVH mainly comes from vibration generated by load imbalance and the resulting current harmonics. For example, usually the 5th and 7th order current harmonics and the coupled 48th order torque pulsation are the main signal sources leading to NVH. The order and amplitude of the main current harmonics causing NVH can be measured by the current sensor on the electric drive assembly of the electric compressor 120. The order and amplitude of the main torque pulsation harmonics causing NVH can also be measured by the vibration sensor on the compressor 1230. Furthermore, these current harmonics can be eliminated by the active harmonic injection method to suppress NVH. However, due to differences in the manufacturing, assembly, and dynamic characteristics of each component of the electric vehicle, the NVH characteristics of the electric compressor vary, which brings difficulties to the application of the active harmonic injection method. Even for the same electric vehicle, the NVH characteristics of the electric compressor are dynamically changing. For example, the main noise order of the signal causing NVH changes dynamically, which also brings difficulties to the application of the active harmonic injection method. In addition, the NVH at the controller level caused by electromagnetic harmonics generated by the switching of high-frequency components of the inverter 1240 may directly lead to external radiation noise and output harmonic current to the motor body. Although the resonance between the switching frequency and the NVH at the motor and compressor component levels can be avoided by increasing the switching frequency, as the switching frequency increases, the loss also increases, and the peak power output is also affected. Therefore, simply increasing the switching frequency of the switching element is not an ideal solution. Therefore, in order to reduce the NVH of the electric vehicle, an improved method is needed that can dynamically track the NVH characteristics of the electric compressor and dynamically eliminate or suppress the NVH of the electric compressor according to the tracked NVH characteristics.

[0030] Figure 3 The figure shows a schematic diagram of a system for suppressing the NVH of an electric compressor in a vehicle according to an embodiment. Figure 3 In Figure 1 and 2 The same numbers as those in [the figure] represent the same or corresponding components.

[0031] The NVH suppression module 150 includes an NVH feature prediction module 1510 and an NVH control module 1520. The NVH feature prediction module 1510 and the NVH control module 1520 can be implemented by a neural network model and thus can also be referred to as the NVH feature prediction model 1510 and the NVH control model 1520. The NVH feature prediction model 1510 generates the NVH features of the electric compressor 120 based on the data related to the operating state of the vehicle 100, and the NVH control model 1520 generates the NVH control parameters based on the NVH features generated by the NVH feature prediction model 1510. In one embodiment, the NVH suppression module 150 can be implemented in the processing module 130. For example, the processor in the processing module 130 executes the machine-executable instructions corresponding to the NVH suppression module 150 to implement the functions of the NVH suppression module 150.

[0032] In one embodiment, the data related to the operating state of the vehicle 100 received by the NVH feature prediction model 1510 includes the weather condition in which the vehicle 100 is located (e.g., the ambient temperature, weather conditions such as rain, snow, wind, fog, etc.), the charging state of the vehicle (e.g., whether it is charging, the battery power, etc.), the communication state of the vehicle, the driving mode of the vehicle (e.g., comfort mode, power-saving mode, sport mode, etc., and the corresponding vehicle control strategies), and the operating conditions of the electric compressor (e.g., rotational speed, acceleration, deceleration, output torque, load increase, load decrease, current, supply voltage, etc.). It can be understood that the data related to the operating state of the vehicle 100 received by the NVH feature prediction model 1510 can include a part of the data types exemplified above or can include more data types. By using various data related to the operating state of the vehicle 100 as the input of the NVH feature prediction model 1510, the NVH features of the electric compressor 120 can be predicted more sensitively and dynamically.

[0033] In one embodiment, the NVH characteristics predicted by the NVH characteristic prediction model 1510 for the electric compressor include the main noise order causing NVH and the amplitude of the signal component corresponding to the main noise order. The main noise order can be a single noise order, such as the order of a current harmonic. The main noise order can also be multiple noise orders, such as the orders of two current harmonics, or the orders of three or more current harmonics. In one embodiment, the NVH characteristic prediction model 1510 outputs the main noise order causing NVH and the corresponding amplitude of the current signal as the NVH characteristic. In one embodiment, the NVH characteristic prediction model 1510 outputs the main noise order causing NVH and the corresponding amplitude of the torque pulsation signal as the NVH characteristic. In one embodiment, the NVH characteristic prediction model 1510 outputs the main noise order causing NVH and the corresponding amplitude of the current signal and the main noise order causing NVH and the corresponding amplitude of the torque pulsation signal as the NVH characteristic. Hereinafter, taking the main noise order causing NVH and the corresponding amplitude of the current signal as the NVH characteristic as an example for illustration, the same principle also applies to the case where the torque pulsation harmonic signal is used as the NVH characteristic or both the current harmonic signal and the torque pulsation harmonic signal are used as the NVH characteristic.

[0034] The NVH control module 1520 receives the NVH characteristics output by the NVH characteristic prediction model 1510 and generates NVH control parameters. For example, the NVH characteristics received by the NVH control module 1520 are the order and the corresponding amplitude of the current harmonic. For example, the order is 7 and the unit of the amplitude is, for example, amperes. For another example, the NVH characteristics received by the NVH control module 1520 are the order and the corresponding amplitude of the torque pulsation signal. For example, the order is 48 and the unit of the amplitude is, for example, newtons. For another example, the NVH characteristics received by the NVH control module 1520 are the order and the corresponding amplitude of the current harmonic and the order and the corresponding amplitude of the torque pulsation signal. In one embodiment, the NVH control parameter generated by the NVH control module 1520 based on the received NVH characteristics is the current compensation amount I e . As Figure 3 shown, the adjustment unit 12130 generates an adjusted reference current I' ref based on the reference current I e output by the speed controller 12110 and the current compensation I ref output by the NVH control module 1520. The adjustment unit 12130 can be implemented as an adder, which generates the adjusted reference current I' ref by adding the reference current I e and the current compensation I ref . The current controller 12120 generates a reference voltage U based on the adjusted reference current I' ref ​ref The motor controller 1210 then, based on the reference voltage U ref controls the operation of the inverter 1240 to convert direct current into alternating current. By compensating the reference current I' after the current adjustment ref , the phase current of the motor 1220 of the electric compressor 120 is adjusted, ultimately reducing the amplitude of the harmonic components of the main noise orders that cause NVH in the current driving the motor 1220, and correspondingly reducing the amplitude of the harmonic components of the coupled torque ripple, thereby effectively eliminating or suppressing NVH.

[0035] In one embodiment, the NVH feature prediction model 1510 predicts the NVH features of the electric compressor at a specific frequency or time interval, and the NVH control module 1520 generates NVH control parameters based on the predicted NVH features at the specific frequency or time interval. In another embodiment, the NVH feature prediction model 1510 predicts the NVH features of the electric compressor at a specific frequency or time interval. When the predicted NVH features reach a threshold, for example, when the amplitude corresponding to the main noise order causing NVH reaches the threshold, the NVH control module 1520 generates NVH control parameters based on the predicted NVH features. In another embodiment, the NVH feature prediction model 1510 not only predicts the NVH features of the electric compressor, but also predicts an indication signal indicating whether noise reduction is required. For example, the indication signal can be a binary classification value, where one value indicates that the NVH control module 1520 needs to be activated to reduce NVH, and the other value indicates that the NVH control module 1520 does not need to be activated. When the predicted indication signal indicates that the NVH control module 1520 needs to be activated, the NVH control module 1520 generates NVH control parameters based on the predicted NVH features.

[0036] Figure 4 FIG. shows a schematic diagram of a system for suppressing NVH of an electric compressor in a vehicle according to one embodiment. Figure 4 in the same as Figures 1 to 3 The same numbers represent the same or corresponding components, and the repeated parts will not be described in detail again.

[0037] As Figure 4 shown, the speed controller 12110 includes a reference torque generation unit 12110-1 and a reference current generation unit 12110-2. In the absence of the adjustment unit 12140, the reference torque generation unit 12110-1 generates a reference torque T ref , and the reference current generation unit 12110-2 converts the reference torque into a reference current I ref . The reference torque generation unit 12110-1 and the reference current generation unit 12110-2 are components of a known motor controller, and their details will not be described in detail.

[0038] In Figure 4 the illustrated embodiment, the NVH control parameters generated by the NVH control module 1520 based on the received NVH characteristics include the current compensation amount I e and the torque compensation amount T e . The adjustment unit 12140 generates an adjusted reference torque T' ref based on the reference torque T e output by the reference torque generation unit 12110-1 and the torque compensation amount T ref output by the NVH control module 1520 ref . The reference current generation unit 12110-2 generates a reference current I ref based on the adjusted reference torque T' ref . The adjustment unit 12130 generates an adjusted reference current I' e based on the reference current I ref output by the reference current generation unit 12110-2 and the current compensation I ref output by the NVH control module 1520 e . The adjustment units 12130 and 12140 can be implemented as adders, which generate the adjusted reference torque T' ref by adding the reference torque T ref and the torque compensation T e , and generate the adjusted reference current I' ref by adding the reference current I ref and the current compensation I ref . The current controller 12120 generates a reference voltage U ref based on the adjusted reference current I' ref . The motor controller 1210 then controls the operation of the inverter 1240 to convert direct current into alternating current based on the reference voltage U

[0039] Figure 5 FIG. shows a schematic diagram of a system for suppressing NVH of an electric compressor in a vehicle according to an embodiment Figure 5 in which the same numbers as Figures 1 to 4 represent the same or corresponding components, and the repeated parts will not be described again

[0040] Compared with Figure 4 in Figure 5In the embodiment of the present invention, the adjustment unit 12130 is reduced, and the NVH control parameters generated by the NVH control module 1520 based on the received NVH characteristics include the torque compensation amount T e , but does not include the current compensation I e The adjustment unit 12140 adjusts the reference torque T based on the reference torque output by the reference torque generation unit 12110-1. ref and the torque compensation T output by the NVH control module 1520 e Generates the adjusted reference torque T' ref The reference current generating unit 12110-2 generates a reference current based on the adjusted reference torque T' ref Generate reference current I ref The current controller 12120 is based on the reference current I ref Generate reference voltage U ref The motor controller 1210 further generates a voltage based on the reference voltage U ref The operation of the inverter 1240 is controlled to convert the DC power into the AC power. The reference current I adjusted by the compensation torque ref , ultimately reducing the amplitude of the harmonic component of the main noise order that causes NVH in the current of the drive motor 1220, and correspondingly reducing the amplitude of the harmonic component of the torque pulsation coupled thereto, thereby effectively eliminating or suppressing NVH.

[0041] Figure 6 A schematic diagram of a process for training an NVH feature prediction model according to one embodiment is shown.

[0042] In one embodiment, the NVH feature prediction model 1510 can be implemented by a fully connected neural network model. The training of the NVH feature prediction model 1510 can be performed offline using historical big data as a training data set. The model input data in the training data set may include data related to the working state of the vehicle, for example, the weather conditions of the vehicle (for example, ambient temperature, weather conditions such as rain, snow, wind, fog, etc.), the charging state of the vehicle (for example, whether it is charging, battery power, etc.), the communication state of the vehicle, the driving mode of the vehicle (for example, comfort mode, power saving mode, sports mode, etc., and corresponding vehicle control strategies), and the working state of the electric compressor (for example, speed, acceleration, deceleration, output torque, load increase, load reduction, current, supply voltage, etc.). It can be understood that the data related to the working state of the vehicle 100 as input data of the NVH feature prediction model may include part of the data types exemplified above, and may also include more data types. The label data in the training data set may include the NVH characteristics of the electric compressor obtained by measurement, such as the main noise order of the electric compressor that causes NVH and the amplitude of the signal component corresponding to the main noise order. As shown in FIG.Figure 6 As shown, the NVH feature prediction model 1510 receives input data in the training dataset and outputs predicted NVH features. The loss determination module 610 determines a loss value L1 based on the predicted NVH features and the measured NVH features in the training dataset. Further, the NVH feature prediction model 1510 can be updated based on the loss value L1. Any suitable neural network model update method can be employed to update the parameters of the NVH feature prediction model 1510 based on the loss value L1. For example, the known AdamW optimizer can be used to update the NVH feature prediction model 1510 based on the loss.

[0043] Figure 7 A schematic diagram for training an NVH control model according to an embodiment is shown.

[0044] In one embodiment, the NVH control model 1520 can be implemented by a fully connected neural network model. The training of the NVH control model 1520 can be performed in an online training manner. For example, the NVH control model 1520 can be trained online during the driving of the vehicle 100. Another example is that the NVH control model 1520 can be trained online by simulating the driving process of the vehicle on a vehicle test bench.

[0045] The NVH control model 1520 receives the NVH features of the electric compressor, such as the main noise orders of the electric compressor that cause NVH and the amplitudes of the signal components corresponding to the main noise orders. In one embodiment, the NVH features received by the NVH control model 1520 can be measured values. In another embodiment, the NVH features received by the NVH control model 1520 can be predicted values generated by the trained NVH feature prediction model 1510. The NVH control model 1520 generates NVH control parameters based on the received NVH features. For example, Figure 3 the current compensation amount I shown e , Figure 5 the torque compensation amount T shown e , Figure 4 the torque compensation amount T shown e and the current compensation amount I e . The motor controller 1210 adjusts the operating parameters of the electric compressor based on the NVH control parameters. For example, referring to Figures 3 to 5 , the motor controller 1210 adjusts the reference torque T e and / or the current compensation amount I e to adjust the reference torque T ref and / or the reference current I ref, thereby adjusting the current supplied to the motor 1220. The NVH detection module 710 detects the NVH characteristics of the electric compressor. For example, the NVH detection module 710 detects the harmonic signal amplitude corresponding to the order of the input signal of the model 1520. For example, when the input signal of the NVH control model 1520 is the order n of the noise source signal and the corresponding harmonic amplitude, the NVH detection module 710 detects the amplitude of the nth harmonic of the noise source signal. As described above, the noise source signal as the model input signal can be current and / or torque ripple. Taking the current signal as an example, the NVH detection module 710 detects the amplitude of the nth current harmonic of the electric compressor. For example, it can detect the amplitude of the nth harmonic of the q-axis current. The loss determination module 720 determines the loss value L2 based on the detected NVH characteristic value and the target NVH characteristic value. In one embodiment, the target NVH characteristic value can be set to 0. In another embodiment, the target NVH characteristic value can be set to the reference current value Iref. It can be understood that the target NVH characteristic value can be set to other appropriate values. Furthermore, the NVH control model 1520 can be updated based on the loss value L2.

[0046] In one embodiment, the parameters of the NVH control model 1520 can be directly updated based on the loss value L2. For example, a known AdamW optimizer can be used to update the parameters of the NVH control model 1520 based on the loss value L2.

[0047] In another embodiment, the NVH control model 1520 can be further updated based on the physical constraint conditions of the motor controller 1210. The NVH control model 1520 can be expressed as y NN = f θ (x), where θ represents the trainable parameters of the NVH control model 1520, x represents the input of the NVH control model 1520, and y NN represents the output of the NVH control model 1520. The motor controller 1210 can be expressed as the physical constraint condition i = g(y NN ), where i represents the current, such as the q-axis current, and g represents the physical constraint condition of the motor controller 1210. It can be understood that in different implementations, this physical constraint condition can be different, but it can be expressed by physical formulas. Based on the above conditions, when updating the parameters θ of the NVH control model 1520 based on the loss value L2, the gradient of the loss value L2 with respect to θ can be expressed as wherein the and can be calculated through the physical constraint conditions of the motor controller 1210, can be calculated through standard neural network update algorithms (such as the AdamW optimizer). In this embodiment, based on To update the parameters θ of the NVH control model 1520, where η represents a configurable learning coefficient. It can be understood that although two exemplary methods for updating the NVH control model 1520 are described above, any suitable training method known in the art or to be known in the future can be used to update the parameters θ of the NVH control model 1520 based on the loss value L2, and the technical solution of the present disclosure is not limited to a specific model update method.

[0048] Figure 8 A flowchart of a method for controlling the NVH of an electric compressor of a vehicle according to an embodiment is shown.

[0049] In step 810, data related to the operating state of the vehicle is obtained. In one embodiment, the data related to the operating state of the vehicle includes at least a part of the following: the weather condition in which the vehicle is located, the charging state of the vehicle, the communication state of the vehicle, the driving mode of the vehicle, the rotational speed, acceleration, deceleration, output torque, load increase, load decrease, current, and supply voltage of the electric compressor of the vehicle.

[0050] In step 820, the NVH characteristics of the electric compressor are generated by the NVH feature prediction model based on the data related to the operating state of the vehicle. In one embodiment, the NVH characteristics of the electric compressor include the main noise orders causing NVH and the amplitudes of the signal components corresponding to the main noise orders. In one embodiment, the main noise orders and the amplitudes of the signal components corresponding to the main noise orders include at least a part of the following: the main noise orders of the current signal of the electric compressor (e.g., the current signal of the motor of the electric compressor) and the amplitudes of the current harmonic components corresponding to the main noise orders; and the main noise orders of the torque pulsation signal of the electric compressor (e.g., the torque pulsation of the motor of the electric compressor and / or the torque pulsation of the compressor) and the amplitudes of the torque harmonic components corresponding to the main noise orders.

[0051] In step 830, the NVH control parameters are generated by the NVH control model based on the generated NVH characteristics. In one embodiment, the NVH control parameters include the current compensation value and / or the torque compensation value of the electric compressor.

[0052] In step 840, the operation of the electric compressor is adjusted based on the generated NVH control parameters. In one embodiment, in step 840, the current and / or torque of the electric compressor are adjusted based on the current compensation value and / or the torque compensation value. In one embodiment, in step 840, the current reference value and / or the torque reference value of the motor controller of the electric compressor are adjusted based on the current compensation value and / or the torque compensation value.

[0053] In one embodiment, a determination step is included before step 830, in which it is determined whether the NVH characteristics of the electric compressor generated by the NVH characteristic prediction model reach a threshold, and step 830 is executed when the threshold is reached.

[0054] In one embodiment, at step 820, the NVH characteristic prediction model further generates a trigger signal indicating whether to activate the NVH control model. A determination step is included before step 830, in which it is determined whether the trigger signal generated by the NVH characteristic prediction model indicates activation of the NVH control model, and step 830 is executed when the trigger signal indicates activation of the NVH control model.

[0055] Figure 9 A flowchart showing a method for training a neural network model for controlling the NVH of an electric compressor of a vehicle according to one embodiment is shown.

[0056] At step 910, the NVH control model in the neural network model generates NVH control parameters based on the NVH characteristics of the electric compressor. In one embodiment, the NVH control parameters include a current compensation value and / or a torque compensation value of the motor of the electric compressor.

[0057] At step 920, the operation of the electric compressor is adjusted based on the generated NVH control parameters. At step 920, the current and / or torque of the electric compressor is adjusted based on the current compensation value and / or the torque compensation value. In one embodiment, at step 920, the current reference value and / or the torque reference value of the motor controller of the electric compressor is adjusted based on the current compensation value and / or the torque compensation value.

[0058] At step 930, the NVH characteristics of the adjusted electric compressor are detected.

[0059] At step 940, the NVH control model is updated based on the detected NVH characteristics.

[0060] In one embodiment, method 900 further includes at Figure 9Steps 950 and 960 not shown in the figure. In step 950, the NVH feature prediction model in the neural network model generates the NVH features of the electric compressor based on data related to the operating state of the vehicle. In step 960, the NVH feature prediction model is updated based on the generated NVH features of the electric compressor and the measured NVH features of the electric compressor to obtain a trained NVH feature prediction model. In one embodiment, in step 910, the NVH control model generates NVH control parameters based on the NVH features of the electric compressor generated by the trained NVH feature prediction model based on data related to the operating state of the vehicle. In one embodiment, in step 910, the NVH control model generates NVH control parameters based on the measured NVH features of the electric compressor.

[0061] In one embodiment, the data related to the operating state of the vehicle includes at least a part of the following: the weather condition where the vehicle is located, the charging state of the vehicle, the communication state of the vehicle, the driving mode of the vehicle, the rotational speed, acceleration, deceleration, output torque, load increase, load decrease, current, and supply voltage of the electric compressor of the vehicle. In one embodiment, the NVH features of the electric compressor include the main noise orders causing NVH and the amplitudes of the signal components corresponding to the main noise orders. In one embodiment, the main noise orders and the amplitudes of the signal components corresponding to the main noise orders include at least a part of the following: the main noise orders of the current signal of the electric compressor (e.g., the current signal of the motor of the electric compressor) and the amplitudes of the current harmonic components corresponding to the main noise orders; and the main noise orders of the torque pulsation signal of the electric compressor (e.g., the torque pulsation of the motor of the electric compressor and / or the torque pulsation of the compressor) and the amplitudes of the torque pulsation harmonic components corresponding to the main noise orders.

[0062] Figure 10 The block diagram of a device for controlling the NVH of an electric compressor of a vehicle according to one embodiment is shown.

[0063] The device 100 includes a data acquisition module 1010, an NVH feature prediction module 1020, and an NVH control module 1030. The data acquisition module 1010 acquires data related to the operating state of the vehicle. The NVH feature prediction module 1020 generates the NVH features of the electric compressor based on the data related to the operating state of the vehicle. The NVH control module 1030 generates NVH control parameters based on the generated NVH features to adjust the operation of the electric compressor based on the generated NVH control parameters.

[0064] In one embodiment, the data related to the operating state of the vehicle includes at least a part of the following: the weather condition in which the vehicle is located, the charging state of the vehicle, the communication state of the vehicle, the driving mode of the vehicle, the rotational speed of the electric compressor of the vehicle, acceleration, deceleration, output torque, load increase, load decrease, current, and supply voltage.

[0065] In one embodiment, the NVH characteristics of the electric compressor include the main noise orders causing NVH and the amplitudes of the signal components corresponding to the main noise orders. In one embodiment, the main noise orders and the amplitudes of the signal components corresponding to the main noise orders include at least a part of the following: the main noise orders of the current signal of the electric compressor (for example, the current signal of the motor of the electric compressor) and the amplitudes of the current harmonic components corresponding to the main noise orders; and the main noise orders of the torque pulsation signal of the electric compressor (for example, the torque pulsation of the motor of the electric compressor and / or the torque pulsation of the compressor) and the amplitudes of the torque harmonic components corresponding to the main noise orders.

[0066] In one embodiment, the NVH control parameters include the current compensation value and / or the torque compensation value of the drive motor of the electric compressor. In one embodiment, adjusting the operating parameters of the electric compressor based on the generated NVH control parameters includes: adjusting the current and / or torque of the motor based on the current compensation value and / or the torque compensation value. In one embodiment, adjusting the current and / or torque of the motor based on the current compensation value and / or the torque compensation value includes: adjusting the current reference value and / or the torque reference value of the motor controller of the electric compressor based on the current compensation value and / or the torque compensation value.

[0067] In one embodiment, device 1000 may include modules 1010 to 1030 and possibly other modules for performing the various operations and functions described above in conjunction with Figures 1 - 9 description.

[0068] Figure 11 FIG. shows a block diagram of a processing device according to one embodiment. According to one embodiment, device 1100 may include one or more control units or processing units 1110 that execute one or more machine-readable instructions stored or encoded in a machine-readable storage medium (i.e., memory 1120). Although not shown in Figure 11 FIG., those skilled in the art can understand that device 1100 may include various other components, such as various communication modules, bus modules, etc. In one embodiment, device 1100 may be used to implement Figure 1The processing module 130 shown. In another embodiment, the apparatus 1100 can be used to implement the training process of the neural network model described above. The processing unit 1110 is configured to perform the various operations and functions described above in connection with Figures 1 - 10 the description.

[0069] According to one embodiment, a program product such as a non-transitory machine-readable medium is provided. The non-transitory machine-readable medium may have instructions that, when executed by the processing unit 1110, cause the apparatus to perform the various operations and functions described above in the various embodiments of the present application in connection with Figures 1 - 10 the description.

[0070] The specific embodiments described above in connection with the accompanying drawings describe exemplary embodiments, but do not represent all embodiments that can be implemented or fall within the scope of the claims. The term "example" or "exemplary" used throughout this specification means "serving as an example, instance, or illustration" and does not mean "preferred" or "advantageous" over other embodiments. For the purpose of providing an understanding of the described technology, the specific embodiments include specific details. However, the technology can be practiced without these specific details. In some instances, well-known structures and devices are shown in block diagram form to avoid obscuring the concepts of the described embodiments.

[0071] The foregoing description of the subject matter of this application is provided to enable any person of ordinary skill in the art to make or use the subject matter of this application. Various modifications to the subject matter of this application will be apparent to persons of ordinary skill in the art, and the general principles defined herein can be applied to other variations without departing from the scope of the subject matter of this application. Thus, the subject matter of this application is not limited to the examples and designs described herein, but is consistent with the broadest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for controlling the noise, vibration, and harshness (NVH) of an electric compressor of a vehicle, comprising: obtaining data related to the operating state of the vehicle; generating, by an NVH feature prediction model, the NVH features of the electric compressor based on the data related to the operating state of the vehicle; generating, by an NVH control model, NVH control parameters based on the generated NVH features; adjusting the operation of the electric compressor based on the generated NVH control parameters.

2. The method according to claim 1, wherein, the NVH features of the electric compressor include the main noise orders causing NVH and the amplitudes of the signal components corresponding to the main noise orders.

3. The method according to claim 2, wherein, the main noise orders and the amplitudes of the signal components corresponding to the main noise orders include at least a part of the following: the main noise orders of the current signal of the electric compressor and the amplitudes of the current harmonic components corresponding to the main noise orders; and the main noise orders of the torque pulsation signal of the electric compressor and the amplitudes of the torque harmonic components corresponding to the main noise orders.

4. The method according to claim 3, wherein, the NVH control parameters include the current compensation value and / or the torque compensation value of the electric compressor.

5. The method according to claim 4, wherein, the adjusting the operation of the electric compressor based on the generated NVH control parameters includes: adjusting the current and / or torque of the electric compressor based on the current compensation value and / or the torque compensation value.

6. The method according to claim 5, wherein, the adjusting the current and / or torque of the electric compressor based on the current compensation value and / or the torque compensation value includes: adjusting the current reference value and / or the torque reference value of the motor controller of the electric compressor based on the current compensation value and / or the torque compensation value.

7. The method according to any one of claims 1 to 6, wherein, the data related to the operating state of the vehicle includes at least a part of the following: the weather conditions where the vehicle is located, the charging state of the vehicle, the communication state of the vehicle, the driving mode of the vehicle, the rotational speed, acceleration, deceleration, output torque, load increase, load decrease, current, and supply voltage of the electric compressor of the vehicle.

8. The method according to claim 1, wherein, the step of generating, by the NVH control model, the NVH control parameters based on the generated NVH features is triggered by at least one of the following conditions: the generated NVH features of the electric compressor reach a threshold; the NVH feature prediction model generates a trigger signal for starting the NVH control model.

9. A method for training a neural network model for controlling the noise, vibration, and harshness (NVH) of an electric compressor of a vehicle, comprising: generating, by the NVH control model in the neural network model, NVH control parameters based on the NVH features of the electric compressor; adjusting the operation of the electric compressor based on the generated NVH control parameters; detecting the NVH features of the adjusted electric compressor; Update the NVH control model based on the detected NVH characteristics.

10. The method according to claim 9, further comprising: generating NVH characteristics of the electric compressor by an NVH characteristics prediction model in the neural network model based on data related to the operating state of the vehicle; updating the NVH characteristics prediction model based on the generated NVH characteristics of the electric compressor and the measured NVH characteristics of the electric compressor to obtain a trained NVH characteristics prediction model; wherein, the generating of the NVH control parameters includes: generating NVH control parameters by the NVH control model based on the NVH characteristics of the electric compressor generated by the trained NVH characteristics prediction model based on data related to the operating state of the vehicle.

11. The method according to claim 9, wherein, the generating of the NVH control parameters includes: generating NVH control parameters by the NVH control model based on the measured NVH characteristics of the electric compressor.

12. The method according to any one of claims 9 to 11, wherein, the NVH characteristics of the electric compressor include the main noise orders causing NVH and the amplitudes of the signal components corresponding to the main noise orders.

13. The method according to claim 12, wherein, the main noise orders and the amplitudes of the signal components corresponding to the main noise orders include at least a part of the following: the main noise orders of the current signal of the electric compressor and the amplitudes of the current harmonic components corresponding to the main noise orders; and the main noise orders of the torque pulsation signal of the electric compressor and the amplitudes of the torque harmonic components corresponding to the main noise orders.

14. The method according to claim 13, wherein, the NVH control parameters include the current compensation value and / or torque compensation value of the motor of the electric compressor.

15. The method according to claim 14, wherein, the adjusting of the operation of the electric compressor based on the generated NVH control parameters includes: adjusting the current reference value and / or torque reference value of the motor controller of the electric compressor based on the current compensation value and / or torque compensation value.

16. A device for controlling the noise, vibration and harshness (NVH) of an electric compressor of a vehicle, comprising: a data acquisition module for acquiring data related to the operating state of the vehicle; an NVH characteristics prediction module for generating NVH characteristics of the electric compressor based on data related to the operating state of the vehicle; an NVH control module for generating NVH control parameters based on the generated NVH characteristics to adjust the control parameters of the electric compressor based on the generated NVH control parameters.

17. The device according to claim 16, wherein, the NVH characteristics of the electric compressor include the main noise orders causing NVH and the amplitudes of the signal components corresponding to the main noise orders.

18. The device according to claim 17, wherein, the main noise orders and the amplitudes of the signal components corresponding to the main noise orders include at least a part of the following: The main noise order of the current signal of the electric compressor and the amplitude of the current harmonic component corresponding to the main noise order; and The main noise order of the torque pulsation signal of the electric compressor and the amplitude of the torque harmonic component corresponding to the main noise order.

19. The apparatus according to claim 18, wherein, The NVH control parameter includes a current compensation value and / or a torque compensation value of the drive motor of the electric compressor.

20. The apparatus according to claim 19, wherein, Adjusting the control parameter of the electric compressor based on the generated NVH control parameter includes: adjusting the current and / or torque of the motor based on the current compensation value and / or torque compensation value.

21. A noise, vibration, and harshness (NVH) processing apparatus for an electric compressor of a vehicle, comprising: One or more processors; and A memory that stores computer-executable instructions that, when executed, cause the one or more processors to perform the method according to any one of claims 1 to 15.

22. A machine-readable storage medium storing executable instructions that, when executed, cause one or more processors to perform the method according to any one of claims 1 to 15.

Images