Engine sound generation method and engine sound generation device

Abstract

It generates engine sounds that sound natural even in vehicles that do not have an engine. [Solution] The controller 20 reads the sound source file for the target engine speed from the storage device 30, which stores sound source files containing simulated sound pressure data of the engine sound, at a predetermined control cycle for each of the multiple engine speeds. The controller 20 then crossfades the simulated sound pressure data of the sound source file read in the current control cycle with the data read in the previous control cycle, thereby continuously playing the simulated sound pressure data of the periodically read sound source files.

Description

Technical Field

[0001] The present invention relates to an engine sound generation method and an engine sound generation device.

Background Art

[0002] Patent Document 1 discloses an engine sound processing device that generates a synthesized engine sound signal according to an output signal corresponding to the operation of a vehicle engine. The engine sound processing device includes a plurality of noise suppression units that convert the output signal of a sensor into a signal in the frequency domain, perform a process of suppressing a noise component on the conversion result, and then convert it into a signal in the time domain and output each of them. The plurality of noise suppression units each have a different frame size when converting into a signal in the frequency domain. The engine sound processing device includes a noise-suppressed signal generation unit that mixes and outputs the output signals of each of the plurality of noise suppression units, and generates a synthesized engine sound signal by processing the output signal of the noise-suppressed signal generation unit.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, since the method disclosed in Patent Document 1 requires real-time acquisition and calculation of a signal corresponding to the operation of the engine, it is impossible to generate an engine sound in a vehicle not equipped with an engine.

[0005] The present invention has been made in view of the above problems, and an object thereof is to provide an engine sound generation method and an engine sound generation device that can generate an engine sound without a sense of incongruity even in a vehicle not equipped with an engine.

Means for Solving the Problems

[0006] An engine sound generation method according to one aspect of the present invention reads out the sound source file for the target engine speed from a storage device that stores sound source files for each of several engine speeds at a predetermined control cycle. The engine sound generation method continuously plays out the simulated sound pressure data of the periodically read sound source files by crossfading the simulated sound pressure data of the sound source file read in the current control cycle with the simulated sound pressure data of the sound source file read in the previous control cycle. [Effects of the Invention]

[0007] According to the present invention, it is possible to generate an engine sound that does not sound unnatural even in vehicles that do not have an engine. [Brief explanation of the drawing]

[0008] [Figure 1] Figure 1 is a block diagram showing the engine sound generation device according to this embodiment. [Figure 2] Figure 2 is a block of the estimated model. [Figure 3] Figure 3 is a diagram illustrating the audio source file. [Figure 4] Figure 4 shows the method for generating the audio source file. [Figure 5] Figure 5 is a flowchart showing a method for generating engine sounds using sound source files. [Figure 6] Figure 6 illustrates a method for continuously playing back sound pressure data. [Modes for carrying out the invention]

[0009] The engine sound generation method and engine sound generation apparatus according to this embodiment will be described below with reference to the drawings.

[0010] The engine sound generator 1 shown in Figure 1 is a device that generates the engine sound of an engine-powered vehicle in a target vehicle. An engine-powered vehicle is a vehicle equipped with an engine and that runs using the engine. Engine-powered vehicles include not only vehicles that run using only the engine, but also hybrid vehicles that run using both an engine and a drive motor. A target vehicle is a vehicle that does not have an engine, such as a battery electric vehicle or a fuel cell vehicle.

[0011] The engine sound generation device 1 includes a microphone 10, a speaker 12, an input interface (input IF) 14, a controller 20, a storage device 30, and a display device 35. The engine sound generation device 1 has a first use for generating engine sound source files and a second use for generating engine sounds using the sound source files.

[0012] When used for the first purpose, the engine sound generator 1 is mounted on an engine vehicle. In this case, the engine sound generator 1 only needs to be equipped with a microphone 10, a controller 20, and a storage device 30. On the other hand, when used for the second purpose, the engine sound generator 1 is mounted on a target vehicle. In this case, the engine sound generator 1 only needs to be equipped with a speaker 12, an input IF 14, a controller 20, and a storage device 30.

[0013] Microphone 10 is a sound collection device located inside the vehicle cabin to collect engine noise from the engine vehicle. For example, microphone 10 is located near the headrest of the driver's seat. The analog signal output from microphone 10 is converted into a digital signal by an AD converter (not shown) and input to controller 20.

[0014] Speaker 12 is located inside the vehicle and outputs sound corresponding to the sound signal output from the controller 20. For example, speaker 12 is mounted on the headrest of the driver's seat and positioned near the driver's left and right ears. However, speaker 12 may be the in-vehicle speaker installed in the vehicle.

[0015] Input IF14 is connected to an in-vehicle network such as CAN (Controller Area Network) installed in the vehicle. The controller 20 can acquire information about the vehicle's status through input IF14.

[0016] The controller 20 performs various processes for generating a sound source file containing simulated engine sound pressure data and for outputting engine sound using the sound source file. The controller 20 consists of a hardware processor, various types of memory, input / output interfaces, etc. The hardware processor may be, for example, a CPU (Central Processing Unit) or an MPU (Micro-Processing Unit). Various programs are installed on the controller 20, and various processes can be realized by executing these programs.

[0017] As shown in Figure 2, the controller 20 is equipped with an estimation model. This estimation model is used when generating engine sound in the target vehicle. Specifically, the estimation model performs estimation processing to estimate the engine speed and engine torque of the engine vehicle based on the vehicle state of the target vehicle. In this estimation processing, it is assumed that the engine vehicle is in the same vehicle state as the target vehicle, and the engine speed and engine torque of the engine vehicle are estimated from the vehicle state of the target vehicle. The estimation model is a machine learning model that has been pre-trained using the vehicle state of the engine vehicle and the engine speed and engine torque of the engine vehicle in that vehicle state as training data. In this embodiment, the vehicle state is the vehicle speed VSO, acceleration Gx, and accelerator opening APO. That is, the estimation model takes the vehicle speed VSO, acceleration Gx, and accelerator opening APO detected in the target vehicle as input and outputs the assumed engine speed R(t) and engine torque T(t) for the engine vehicle.

[0018] Note that the estimation model is not limited to a machine - learned learning model, and may be composed of a pre - designed arithmetic circuit or the like, or may be a configuration combining a machine - learning model and an arithmetic circuit that performs a predetermined operation on the output of the machine - learning model. Further, the controller 20 may independently include an estimation model for estimating the engine speed and an estimation model for estimating the engine torque. In this case, the estimation model for estimating the engine torque may be configured to output a first engine torque if the accelerator opening APO is greater than or equal to a threshold value, and output a second engine torque smaller than the first engine torque if the accelerator opening APO is smaller than the threshold value.

[0019] The storage device 30 records base sound pressure data when the engine sound is collected in the engine vehicle. Also, when simulated sound pressure data simulating the engine sound of the engine vehicle is generated from the base sound pressure data, the storage device 30 stores a sound source file recording the simulated sound pressure data. As shown in FIG. 3, the storage device 30 stores sound source files for each of a plurality of engine speeds. For example, the storage device 30 stores 381 sound source files corresponding to each engine speed divided at 10 - rpm intervals from 1200 rpm to 5000 rpm. Also, the storage device 30 stores 381 - pattern sound source files for each of a plurality of engine torques. In the present embodiment, 381 - pattern sound source files corresponding to 40 Nm and 381 - pattern sound source files corresponding to 60 Nm are stored in the storage device 30. The sound source file is, for example, a file in the WAV (Waveform Audio File Format) format, but is not limited thereto. In the present embodiment, 10 rpm is used as the resolution of the engine speed at which the sound source file is prepared, but is not limited thereto.

[0020] The display device 35 is a display that displays various types of information. The display device 35 is controlled by the controller 20 and displays the base sound pressure data of the engine sound collected in the engine vehicle, or displays information for generating the simulated sound pressure data of the engine sound.

[0021] Hereinafter, referring to FIG. 4, a method for creating a sound source file according to this embodiment will be described. The processing shown in the flowchart of FIG. 4 is executed by the controller 20. When creating a sound source file, the engine sound generation device 1 is mounted on the engine vehicle. The microphone 10 is disposed near the left and right ears of the driver, for example, on the headrest of the driver's seat.

[0022] In step S10, the controller 20 turns on the microphone 10 and records the engine sound of the engine vehicle. The recording of the engine sound is performed for a certain period of time (for example, for 10 seconds) while keeping the engine speed constant. The analog signal output from the microphone 10 is converted into a digital signal by the AD converter and acquired by the controller 20. The controller 20 processes the digital signal acquired from the microphone 10 and stores it in the storage device 30. In the storage device 30, base sound pressure data corresponding to the digital signal for a certain period of time is stored (recorded). The base sound pressure data is data of the sound pressure in the time domain and is represented by a waveform with the time axis as the horizontal axis and the sound pressure as the vertical axis.

[0023] The base sound pressure data is acquired for each of a plurality of predetermined base engine speeds. The plurality of base engine speeds are, for example, 14 patterns of engine speeds obtained by dividing from 1200 rpm to 5000 rpm in increments of 200 rpm or 500 rpm. Also, the 14 patterns of base sound pressure data are acquired for each of a plurality of engine torques. In this embodiment, 14 patterns of base sound pressure data corresponding to 40 Nm and 14 patterns of base sound pressure data corresponding to 60 Nm are acquired respectively.

[0024] Note that the number of base engine speed patterns is not limited to 14; it should be less than the number of engine speed patterns related to the final generated simulated sound pressure data. Also, the following explanation describes the processing of 14 base sound pressure data patterns corresponding to 40 Nm, but the same processing is performed for 14 base sound pressure data patterns corresponding to 60 Nm.

[0025] In step S11, the controller 20 performs frequency analysis on the base sound pressure data using the Fourier transform. This frequency analysis converts the time-domain data into frequency-domain data, generating sound pressure level data. The sound pressure level data is frequency-domain sound pressure level data and is represented by a waveform with frequency on the horizontal axis and sound pressure level on the vertical axis. The processing in step S11 generates 14 patterns of sound pressure level data.

[0026] In step S12, the controller 20 separates the sound pressure level data for each of the 14 patterns into an engine rotation order component and a base component other than the engine rotation order component. The separation of the engine rotation order component and the base component is performed by filtering or the like. For example, the controller 20 separates the engine rotation order components up to the 15th order from the other base components.

[0027] In step S13, the controller 20 performs interpolation calculations using 14 patterns of engine rotation order components. Specifically, for each of the 14 patterns of engine rotation order components, the controller 20 generates engine rotation order components by shifting the sound pressure level and frequency of the engine rotation order component, thereby interpolating between each pattern in increments of 10 rpm. Through this interpolation calculation, 367 patterns of engine rotation order components are generated, corresponding to engine rotation speeds other than the base engine rotation speed, out of the 381 patterns of engine rotation speeds that are to be generated as sound source files. As a result of the processing in step S13, engine rotation order components are generated for each of the 367 patterns of engine rotation speeds that are not the target of engine sound recording.

[0028] In step S14, the controller 20 generates 381 patterns of sound pressure level data by combining 381 patterns of engine rotation order components with a base component.

[0029] In step S15, the controller 20 converts the sound pressure level data into simulated sound pressure data. The simulated sound pressure data is time-domain sound pressure data, for example, 10 seconds of sound pressure data. In step S16, the controller 20 generates a sound source file by writing the sound pressure data to a WAV file. The processes in steps S15 and S16 are performed for each of the 381 data patterns.

[0030] The 381 sound source files corresponding to 40 Nm and the 381 sound source files corresponding to 60 Nm, generated through this series of processes, are stored in the storage device 30.

[0031] Next, with reference to Figures 5 and 6, the method for generating engine sound according to this embodiment will be described. The process shown in the flowchart of Figure 5 is executed by the controller 20 at a predetermined control cycle. When generating engine sound, the engine sound generation device 1 is mounted on the target vehicle. The speaker 12 is positioned near the driver's left and right ears, for example, on the headrest of the driver's seat.

[0032] First, in order to perform this process, the controller 20 starts timing a counter that counts up in sync with the control cycle. When the counter reaches the length of the sound source file, for example 10 seconds, it is reset to zero and starts counting up again from zero.

[0033] In step S20, the controller 20 acquires the vehicle status of the target vehicle. The vehicle status consists of the vehicle speed VSO, acceleration Gx, and accelerator opening APO.

[0034] In step S21, the controller 20 estimates the engine speed and engine torque of the engine vehicle based on the vehicle speed VSO, acceleration Gx, and accelerator opening APO. This estimation process is performed using an estimation model.

[0035] In step S22, the controller 20 refers to the storage device 30 and searches for the estimated engine speed and engine torque and the corresponding sound source file.

[0036] In step S23, the controller 20 refers to the counter's count value. The simulated sound pressure data in the sound source file is indexed in control cycle increments from 0 to 10 seconds. The controller 20 refers to the simulated sound pressure data of the searched sound source file and identifies the index corresponding to the current count value. The controller 20 reads simulated sound pressure data equivalent to one buffer, starting from the index corresponding to the current count value. The length of one buffer corresponds to the length of two control cycles.

[0037] In step S24, the controller 20 plays back the read simulated sound pressure data. Specifically, as shown in Figure 6, the controller 20 crossfades the simulated sound pressure data read in the current control cycle with the simulated sound pressure data read in the previous control cycle. Crossfading is a method of playing back sound smoothly without separating the two sound pressure data by fading out the previous simulated sound pressure data while fading in the current simulated sound pressure data. The controller 20 outputs a signal corresponding to the playback of the simulated sound pressure data to the speaker 12. As a result, the speaker 12 outputs an engine sound corresponding to the played simulated sound pressure data, generating engine sound inside the target vehicle.

[0038] This series of processes is performed periodically, allowing the simulated sound pressure data of the periodically read audio files to be played back continuously.

[0039] In this embodiment of the engine sound generation method, the controller 20 reads out the sound source file for the target engine speed from the storage device 30, which stores sound source files containing simulated sound pressure data of the engine sound, at predetermined control cycles for each of the multiple engine speeds. The controller 20 then crossfades the simulated sound pressure data of the sound source file read in the current control cycle with the data read in the previous control cycle, thereby continuously playing back the simulated sound pressure data of the periodically read sound source files.

[0040] This method allows for the generation of engine sounds within the program by searching for and loading the sound source file corresponding to the engine speed, since sound source files are prepared in advance. This eliminates the need for complex calculations to generate sound data, making it easy to generate engine sounds. In addition, because sound source files can be prepared with fine resolution, smooth and varied engine sounds can be generated. In this way, it is possible to generate engine sounds that sound natural even in vehicles that do not have an engine.

[0041] The engine sound generation method of this embodiment further includes acquiring the vehicle state of the target vehicle at a control cycle and performing an estimation process at the same control cycle to estimate the engine speed of the engine vehicle when the engine vehicle is in the same vehicle state as the target vehicle. The process of reading the sound source file is performed based on the engine speed estimated in the estimation process.

[0042] This method allows for the generation of sound files with engine RPMs that match the vehicle's condition. This enables the creation of more natural-sounding engine noises that are in tune with the vehicle's state.

[0043] In the engine sound generation method of this embodiment, the storage device 30 stores multiple sound source files for each of the multiple engine torques, corresponding to the number of engine rotations. In this case, the estimation process includes estimating the engine torque in addition to the engine rotation speed. The process of reading the sound source files is performed based on the engine rotation speed and engine torque estimated in the estimation process.

[0044] This method allows you to read sound files that match both engine speed and engine torque. This enables the generation of a wide variety of engine sounds, as differences in engine torque are reflected in addition to engine speed.

[0045] The engine sound generation method of this embodiment estimates the engine speed based on the vehicle state of the target vehicle, using the vehicle state of the target vehicle as input, and is based on an estimation model that has been pre-trained using the vehicle state of the target vehicle and the engine speed of the engine vehicle as training data.

[0046] This method allows for the accurate identification of the engine speed of a corresponding engine vehicle based on the vehicle's condition. This enables the generation of a more natural engine sound that matches the vehicle's condition.

[0047] In the engine sound generation method of this embodiment, the multiple sound source files stored in the storage device 30 are generated by the file generation process described below. The first process involves recording the engine sound when the engine is operated at multiple base engine speeds to obtain base sound pressure data for each of the multiple base engine speeds. The second process involves performing frequency analysis on the base sound pressure data for each of the multiple base engine speeds to generate sound pressure level data, which is the sound pressure level in the frequency domain, for each of the multiple base engine speeds. The third process separates the sound pressure level data for each of the multiple base engine speeds into an engine rotation order component and a base component other than the engine rotation order component. The fourth process generates an engine rotation order component for each of the multiple engine speeds by performing interpolation calculations using the engine rotation order component at each of the multiple base engine speeds. The fifth process generates sound pressure level data for each of the multiple engine speeds by adding the engine rotation order component at each of the multiple engine speeds to the base component. The sixth process generates a sound source file for each of the multiple engine speeds by converting the sound pressure level data for each of the multiple engine speeds into sound pressure data in the time domain.

[0048] This method eliminates the need to record all the sound pressure data corresponding to a vast number of sound source files from the engine vehicle. Therefore, a huge number of sound source files can be created using a simple method.

[0049] In addition to the engine sound generation method described above, an engine sound generation device 1 is also included as one or more embodiments of the present invention. The engine sound generation device 1 includes a controller 20 that generates the engine sound of an engine-equipped vehicle in a target vehicle, a storage device 30 which stores sound source files, each containing simulated sound pressure data of the engine sound for multiple engine speeds, and a speaker 12 that outputs the engine sound. The controller 20 executes the engine sound generation method described above. With this configuration, since sound source files are prepared in advance, the engine sound generation can be executed within the program by searching for and reading the sound source file corresponding to the engine speed. This eliminates the need for complex calculations to generate sound data, and engine sounds can be easily generated even in vehicles that do not have an engine. In addition, since sound source files can be prepared with fine resolution, smooth and varied engine sounds can be generated.

[0050] As described above, embodiments of the present invention have been presented, but the statements and drawings that constitute part of this disclosure should not be understood as limiting the invention. Various alternative embodiments, examples, and operational techniques will become apparent to those skilled in the art from this disclosure. [Explanation of Symbols]

[0051] 1: Engine sound generator, 12: Speaker, 14: Input interface, 20: Controller, 30: Storage device, 35: Display device

Claims

1. An engine sound generation method performed by a controller of an engine sound generation device that generates engine sounds in a target vehicle, wherein the engine sound of an engine-equipped vehicle is generated in the target vehicle, For each of the multiple engine speeds, the sound source file containing simulated sound pressure data of the engine sound is read from the storage device, which stores the sound source file for the engine speed to be played back, at a predetermined control cycle. An engine sound generation method that includes crossfading the simulated sound pressure data of the sound source file read in the current control cycle with the simulated sound pressure data of the sound source file read in the previous control cycle, thereby continuously playing the simulated sound pressure data of the sound source file that is read periodically.

2. The vehicle status of the target vehicle is acquired at the control cycle, The control cycle further includes performing an estimation process to estimate the engine speed of the engine vehicle when the engine vehicle is in the same vehicle state as the target vehicle, The engine sound generation method according to claim 1, wherein the process of reading the sound source file is performed based on the engine speed estimated in the estimation process.

3. The storage device stores, for each of the multiple engine torques, sound source files corresponding to the number of engine rotations. The estimation process includes estimating the engine torque in addition to the engine speed, The engine sound generation method according to claim 2, wherein the process of reading the sound source file is performed based on the engine speed and engine torque estimated in the estimation process.

4. The engine sound generation method according to claim 2, wherein the engine speed is estimated based on the vehicle state of the target vehicle, using the vehicle state of the target vehicle as input, based on an estimation model that has been pre-machine-trained using the vehicle state and engine speed of the engine vehicle as training data.

5. The multiple sound source files stored in the aforementioned storage device are: By recording the engine sound when the engine is operated at multiple base engine speeds, base sound pressure data is obtained for each of the multiple base engine speeds. By performing frequency analysis on the base sound pressure data for each of the multiple base engine rotation speeds, sound pressure level data, which is the sound pressure level in the frequency domain, is generated for each of the multiple base engine rotation speeds. For each of the aforementioned multiple base engine rotation speeds, the sound pressure level data is separated into an engine rotation order component and a base component other than the engine rotation order component. By performing interpolation calculations using the engine rotational order components of the multiple base engine rotational speeds, the engine rotational order components are generated for each of the multiple engine rotational speeds. By adding the engine rotational order component at each of the multiple engine rotational speeds to the base component, sound pressure level data is generated for each of the multiple engine rotational speeds. By converting the sound pressure level data into time-domain sound pressure data for each of the multiple engine speeds, the sound source file is generated for each of the multiple engine speeds. The engine sound generation method according to claim 1, which is generated by a file generation process that includes the following.

6. A controller that generates the engine sound of an engine-equipped vehicle in the target vehicle, A storage device in which sound source files containing simulated sound pressure data of the engine sound are stored for each of several engine speeds, It has a speaker that outputs the aforementioned engine sound, The aforementioned controller, From the storage device, the sound source file of the engine speed to be played back is read out at a predetermined control cycle. An engine sound generation device that continuously plays back the simulated sound pressure data of a sound source file that is read periodically by crossfading the simulated sound pressure data of the sound source file read in the current control cycle and the simulated sound pressure data of the sound source file read in the previous control cycle.

Images