Sound wave synthesis method and device, electronic equipment and storage medium

Abstract

The application provides a sound wave synthesis method and device, electronic equipment and a storage medium, wherein the method comprises: obtaining sample audio data of a sample vehicle under different throttle modes; performing particle extraction on the sample audio data under each throttle mode according to multi-order main harmonic data corresponding to each sampling point in the sample audio data under each throttle mode, to obtain a particle set under each throttle mode; performing particle matching in the particle set under each throttle mode according to throttle information of a target vehicle, to obtain target particles under each throttle mode, wherein the target particles are particles matched with the throttle information; and performing interpolation fusion on the target particles under a plurality of throttle modes according to the throttle information, to obtain a sound wave signal of the target vehicle. The application realizes sound wave synthesis by particle splicing of real sound wave characteristics under different throttle modes, effectively improving the dynamics and quality of sound wave synthesis.

Description

Technical Field

[0001] This invention relates to the field of signal processing technology, and in particular to a method, apparatus, electronic device, and storage medium for synthesizing sound waves. Background Technology

[0002] New energy vehicles are gradually penetrating the existing internal combustion engine vehicle market. However, compared to internal combustion engine vehicles, new energy vehicles lack the sound signals generated by internal combustion engines, resulting in poorer accuracy of drivers' perception of vehicle speed, which in turn affects driving safety. Therefore, the synthesis of internal combustion engine sound signals for new energy vehicles has become an important research topic that urgently needs to be studied in this field.

[0003] In related technologies, modeling and simulation are commonly used to synthesize the sound waves of internal combustion engines in new energy vehicles. However, this method is an approximate simulation of real samples and has certain errors, which limits the quality of the synthesized sound wave signal. Summary of the Invention

[0004] This invention provides a sound wave synthesis method, apparatus, electronic device, and storage medium to address the shortcomings of existing technologies that use modeling and simulation to synthesize the sound waves of internal combustion engines in new energy vehicles, which results in limited quality of the synthesized sound wave signals. This invention aims to improve the dynamics and quality of sound wave synthesis.

[0005] This invention provides a method for synthesizing sound waves, comprising:

[0006] Acquire sample audio data of the sample vehicle under different throttle modes;

[0007] Based on the multi-order main harmonic data corresponding to each sampling point in the sample audio data of each throttle mode, particle extraction is performed on the sample audio data of each throttle mode to obtain the particle set of each throttle mode.

[0008] Based on the throttle information of the target vehicle, particle matching is performed in the particle set of each throttle mode to obtain the target particle of each throttle mode, wherein the target particle is the particle that matches the throttle information.

[0009] Based on the throttle information, interpolation and fusion are performed on multiple target particles under the throttle modes to obtain the sound signal of the target vehicle.

[0010] According to a sound wave synthesis method provided by the present invention, the step of extracting particles from the sample audio data of each throttle mode based on the multi-order principal harmonic data corresponding to each sampling point in the sample audio data of each throttle mode to obtain a particle set of each throttle mode includes:

[0011] For any sample audio data in any of the throttle modes, the multi-order main harmonic data corresponding to each sampling point in the sample audio data are filtered out from the sample audio data to obtain the target signal;

[0012] Based on the order of the main harmonic data in the sample audio data, determine the segmentation code value corresponding to each signal segment in the target signal;

[0013] Based on the segmentation encoding value, particle extraction is performed on each of the signal segments to obtain the particle set in the throttle mode corresponding to the sample audio data;

[0014] The sample audio data under each of the throttle modes are cyclically processed to obtain the particle set under each of the throttle modes.

[0015] According to a sound wave synthesis method provided by the present invention, the step of extracting particles from each signal segment based on the segmentation encoding value to obtain a particle set in the throttle mode corresponding to the sample audio data includes:

[0016] In each of the signal segments, a target zero point whose position coding value matches the segmentation coding value corresponding to each of the signal segments is obtained;

[0017] Using each target zero point as a dividing point, particle extraction is performed on each signal segment to obtain the particle set in the throttle mode corresponding to the sample audio data.

[0018] According to a sound wave synthesis method provided by the present invention, determining the segmentation coding value corresponding to each signal segment based on the order of the main harmonic data in the sample audio data includes:

[0019] The segmentation code value corresponding to the first signal segment is determined based on the order of the main harmonic data in the sample audio data;

[0020] Based on the order of the main harmonic data in the sample audio data and the segmentation coding value corresponding to the first signal segment, determine the segmentation coding value corresponding to each second signal segment;

[0021] Wherein, the first signal segment is the first signal segment among the plurality of signal segments, and the second signal segment is the signal segment other than the first signal segment among the plurality of signal segments.

[0022] According to a sound wave synthesis method provided by the present invention, the step of determining the segmentation code value corresponding to each second signal segment based on the order of the main harmonic data in the sample audio data and the segmentation code value corresponding to the first signal segment includes:

[0023] For any of the second signal segments, multiple candidate segmentation coding values ​​are determined based on the order of the main harmonic data in the sample audio data;

[0024] Obtain the first particle corresponding to each of the candidate segmentation coding values; the first particle is the first particle formed by particle extraction of the second signal segment based on each of the candidate segmentation coding values;

[0025] Among the multiple candidate segmentation coding values, the target segmentation coding value corresponding to the first particle with the highest correlation to the second particle is obtained; the second particle is the last particle in the previous signal segment of the second signal segment;

[0026] The segmentation code value corresponding to the second signal segment is calculated based on the target segmentation code value and the segmentation code value corresponding to the first signal segment.

[0027] By iterating through each of the second signal segments, the segmentation code value corresponding to each second signal segment is obtained.

[0028] According to a sound wave synthesis method provided by the present invention, the step of performing particle matching in particle sets under each throttle mode based on the throttle information of the target vehicle to obtain the target particles under each throttle mode includes:

[0029] Based on the fundamental frequency data corresponding to each sampling point in the sample audio data of each throttle mode, obtain the first rotational speed corresponding to each particle in the particle set of each throttle mode;

[0030] Based on the throttle information, the second speed corresponding to the throttle information is simulated in the transmission simulation model;

[0031] In the particle set of each throttle mode, the particle corresponding to the first rotational speed with the highest matching degree to the second rotational speed is determined as the target particle of each throttle mode.

[0032] According to a sound wave synthesis method provided by the present invention, the step of interpolating and fusing multiple target particles under the throttle mode based on the throttle information to obtain the sound wave signal of the target vehicle includes:

[0033] Based on the throttle information, the torque parameters corresponding to the throttle information are simulated in the transmission simulation model;

[0034] Based on the torque parameter, interpolation processing is performed on multiple target particles under the throttle mode to obtain a composite particle with the rotational speed corresponding to the torque parameter;

[0035] The acoustic signal of the target vehicle is synthesized based on the synthesized particles.

[0036] According to a sound wave synthesis method provided by the present invention, the steps for obtaining the multi-order principal harmonic data corresponding to each sampling point in the sample audio data of each of the throttle modes include:

[0037] Each sampling point in the sample audio data of each of the aforementioned throttle modes is divided into a first sampling point or a second sampling point according to a preset interval;

[0038] For any first sampling point, based on the position information of the multi-order main harmonic data of the previous sampling point, the multi-order main harmonic data corresponding to the first sampling point is searched in the frequency domain data sequence corresponding to the first sampling point to obtain the multi-order main harmonic data of the first sampling point.

[0039] By traversing each of the first sampling points, the multi-order main harmonic data corresponding to each of the first sampling points are obtained;

[0040] Based on the multi-order principal harmonic data corresponding to each of the first sampling points, interpolation estimation is performed on each of the second sampling points to obtain the multi-order principal harmonic data corresponding to each of the second sampling points.

[0041] The present invention also provides a sound wave synthesis device, comprising:

[0042] The data acquisition unit is used to acquire sample audio data of the sample vehicle under different throttle modes;

[0043] The particle extraction unit is used to extract particles from the sample audio data of each throttle mode based on the multi-order main harmonic data corresponding to each sampling point in the sample audio data of each throttle mode, so as to obtain the particle set of each throttle mode.

[0044] The particle matching unit is used to perform particle matching in the particle set of each throttle mode according to the throttle information of the target vehicle, so as to obtain the target particle of each throttle mode, wherein the target particle is the particle that matches the throttle information.

[0045] The sound wave synthesis unit is used to interpolate and fuse multiple target particles under the throttle mode according to the throttle information to obtain the sound wave signal of the target vehicle.

[0046] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the sound wave synthesis method as described above.

[0047] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the sound wave synthesis method as described above.

[0048] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the sound wave synthesis method as described above.

[0049] The sound wave synthesis method, apparatus, electronic device, and storage medium provided by this invention acquire real sample audio data of a sample vehicle under different throttle modes, extract particles from the sample audio data under each throttle mode to obtain more realistic and detailed sound wave characteristics under different throttle modes, and perform particle matching in the particle set of each throttle mode according to the throttle information of the target vehicle. Based on the throttle information, interpolation and fusion are performed on the target particles of multiple throttle modes to seamlessly connect the particles of different throttle modes, thereby generating a continuous and natural sound wave signal. This achieves sound wave synthesis based on real sound wave characteristics under different throttle modes, so that no matter how the vehicle state changes between idling, acceleration, deceleration, and free coasting, the synthesized sound wave signal can reflect the torque and speed state of the internal combustion engine in real time and seamlessly, thus effectively improving the dynamics and quality of sound wave synthesis. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0051] Figure 1 This is one of the flowcharts of the sound wave synthesis method provided by the present invention.

[0052] Figure 2 This is a schematic diagram of the distribution of harmonic signals in the sample audio data provided by the present invention.

[0053] Figure 3 The second schematic diagram of the sound wave synthesis method provided by this invention.

[0054] Figure 4 This is a schematic diagram illustrating the relative change between engine speed and vehicle speed at 20% throttle, as provided by the present invention.

[0055] Figure 5 This is a schematic diagram illustrating the relative change between engine speed and vehicle speed at 80% throttle, as provided by the present invention.

[0056] Figure 6 This is a schematic diagram of the sound wave synthesis device provided by the present invention.

[0057] Figure 7This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0058] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0059] New energy vehicles are gradually penetrating the existing internal combustion engine vehicle market, and many users are prioritizing their choice of new energy vehicles. Compared to internal combustion engine vehicles, new energy vehicles use electric motors to avoid the noise introduced by the internal combustion engine. However, internal combustion engine noise is also necessary. For example, the lack of engine sound signals in new energy vehicles leads to poorer speed perception for drivers, making it easier for them to exceed the speed limit and thus affecting driving safety. On the other hand, for professional race car drivers and drivers who want to experience the ultimate acceleration and speed, the roar of the internal combustion engine can greatly enhance their immersive driving experience. Therefore, the synthesis of internal combustion engine sound for new energy vehicles has become an important research topic in this field.

[0060] In common internal combustion engine sound synthesis technologies, modeling and simulation are typically used to synthesize the sound waves of internal combustion engines in new energy vehicles. For example, some solutions use physical model simulation to synthesize the sound waves of internal combustion engines in new energy vehicles. This approach requires modeling various components of the vehicle, including but not limited to the internal combustion engine, exhaust pipe, and transmission, and then optimizing them based on the vehicle model to ultimately synthesize the sound wave signal. Another approach uses order synthesis methods to synthesize the sound waves of internal combustion engines in new energy vehicles. Specifically, this involves modeling the engine sound based on simplified equivalent harmonic component superimposed noise, and then designing multiple filters to enhance or weaken harmonics in different frequency bands to ultimately synthesize the sound wave signal.

[0061] However, this modeling and simulation method is essentially an approximate simulation of real samples (such as physical structures and engine signals), which has certain errors and limits the quality of the synthesized sound wave signal.

[0062] In response, some solutions offer splicing-based synthesis methods for synthesizing the sound waves of internal combustion engines in new energy vehicles. These splicing-based methods include wavetable synthesis and particle synthesis. Both methods segment and splice the sound wave signal of the target vehicle model to achieve sound wave synthesis. The distinction between wavetable synthesis and particle synthesis lies in the segment length and the splicing method. Wavetable synthesis typically uses segments of 1-2 seconds, and the splicing method employs techniques such as simple vectorized interpolation or precise arithmetic interpolation. Particle synthesis, on the other hand, fully considers the periodicity of the internal combustion engine noise signal itself, dividing it into particles according to its period. Each particle is typically 10-100 ms long, and sound wave synthesis uses windowed, fundamental frequency-synchronized overlapping and addition. While splicing-based synthesis methods offer high real-time performance, they rely on various types of particles or wavetables to achieve high dynamic range. Acquiring multiple types of particles or wavetables is difficult, making it challenging to synthesize dynamic and high-quality sound wave signals.

[0063] Therefore, to address the aforementioned issues, this embodiment provides a sound wave synthesis method. It obtains internal combustion engine control parameters through gear shift simulation, including but not limited to engine speed, accelerator pedal opening, and vehicle speed. By integrating the simulation of internal combustion engine control parameters with audio particle extraction and splicing under different throttle modes, it achieves real-time synthesis of the internal combustion engine sound wave signal. This not only improves synthesis efficiency but also ensures that the output sound wave signal responds quickly to changes in throttle and braking speed, exhibits smooth gear shift transitions, and smoothly switches gear shift logic under different throttle states. This results in a more dynamic and higher-quality synthesized sound wave signal.

[0064] Figure 1 This is one of the flowcharts of the sound wave synthesis method provided by the present invention; the main body executing the method is a sound wave synthesis device, which can be a vehicle controller built into the target vehicle for which sound wave synthesis is required, or an external control device of the target vehicle for which sound wave synthesis is required, etc. This embodiment does not specifically limit it.

[0065] like Figure 1 As shown, the method includes steps 110, 120, 130 and 140.

[0066] Step 110: Obtain sample audio data of the sample vehicle under different throttle modes.

[0067] Optionally, during the sound wave synthesis process, the real audio data emitted by the engine of the sample vehicle when it is running in different throttle modes can be recorded to obtain sample audio data in different throttle modes. This allows for the automatic re-synthesis of a very realistic engine sound based on the throttle and braking information of the real vehicle to be synthesized, reducing the complexity of sound wave synthesis while improving the dynamics and quality of sound wave synthesis.

[0068] The sample vehicles here are various types of vehicles equipped with internal combustion engines (hereinafter referred to as internal combustion engine vehicles); the throttle modes here include, but are not limited to, the throttle mode from idle speed to high throttle acceleration (also known as high throttle mode) and the throttle mode for low throttle acceleration (also known as low throttle mode), which can be configured according to actual needs. This embodiment does not make specific limitations on this.

[0069] Step 120: Based on the multi-order main harmonic data corresponding to each sampling point in the sample audio data of each throttle mode, perform particle extraction on the sample audio data of each throttle mode to obtain the particle set of each throttle mode.

[0070] Optionally, after acquiring the sample audio data for each throttle mode, the main harmonics can be extracted from each sampling point in the sample audio data for each throttle mode to obtain the multi-order main harmonic data corresponding to each sampling point. The order of the main harmonic data here is the same as the number of cylinders in the sample vehicle. For example, if the sample vehicle includes a 12-cylinder engine, the corresponding order of the main harmonic data is also 12. The main harmonic data here includes, but is not limited to, frequency values ​​and location information; this embodiment does not specifically limit these.

[0071] The main harmonic extraction methods here include: searching for the maximum energy value of each frequency domain data sequence corresponding to all sampling points, and determining the multi-order main harmonic data corresponding to all sampling points based on the harmonic data corresponding to the maximum energy value found; or, searching for the maximum energy value of each frequency domain data sequence corresponding to a portion of sampling points, determining the multi-order main harmonic data corresponding to that portion of sampling points based on the harmonic data corresponding to the maximum energy value found, and then interpolating the multi-order main harmonic data corresponding to other portions of sampling points based on the multi-order main harmonic data corresponding to that portion of sampling points. This embodiment does not specifically limit this method.

[0072] After obtaining the multi-order principal harmonic data corresponding to each sampling point in the sample audio data of each throttle mode, particle extraction can be performed on the sample audio data of each throttle mode based on the multi-order principal harmonic data corresponding to each sampling point in the sample audio data of each throttle mode, so as to construct a particle set for each throttle mode based on the multiple particles extracted in each throttle mode.

[0073] It's important to note that if a particle contains more than two work cycles, the particle switching response will be too slow and the particle variability will be insufficient during later synthesis. If a particle has less than one work cycle, switching particles during later synthesis and splicing will cause a significant change in the audio fundamental frequency, severely damaging the audio's listening experience. Therefore, to ensure the quality of sound wave synthesis, particle extraction here involves splitting the sample audio data from the sample vehicle into multiple smaller audio segments, which are the particles. During the splitting process, it's necessary to ensure that the fundamental period of each particle is the same as that of the audio, meaning that the duration of each particle corresponds to one work cycle of the sample vehicle's engine. For example, during particle extraction, a dynamic filter can be used to filter out multi-order main harmonic data from the sample audio data, and particle extraction can be performed using a zero-point finding method to achieve phase alignment between particles.

[0074] After obtaining the particle sets for each throttle mode, they can be stored in the particle storage module so that the corresponding particles can be loaded in real time for sound wave synthesis.

[0075] Step 130: Based on the throttle information of the target vehicle, perform particle matching in the particle set of each throttle mode to obtain the target particle of each throttle mode. The target particle is the particle that matches the throttle information.

[0076] The throttle information here can be generated in one cycle or in multiple cycles. For simplicity, the following description uses throttle information generated in a single cycle as an example to illustrate the method provided in this embodiment.

[0077] Optionally, after the particle set in each throttle mode, the throttle information of the target vehicle can be obtained. The target vehicle here is the vehicle for which internal combustion engine sound synthesis is required, such as a new energy vehicle without an internal combustion engine. The throttle information can be obtained through real-time sensing of the target vehicle or transmitted via other control methods; this embodiment does not specifically limit this. The throttle information includes, but is not limited to, the pedal's action parameters, etc., which are also not specifically limited in this embodiment.

[0078] After obtaining the throttle information of the target vehicle, the particle storage module can be used to load and retrieve particle sets for each throttle mode. Particles matching the throttle information of the target vehicle can then be selected from these particle sets for each throttle mode and used as the target particles for that mode. For example, the simulated speed value corresponding to the throttle information and the speed value corresponding to each particle can be calculated and matched to determine the particles that match the throttle information of the target vehicle.

[0079] It should be noted that because particles of different speeds have varying lengths, the digital signal processor (DSP) platform experiences a longer reading time when retrieving particles from a particle storage module (such as flash memory). Therefore, a single particle can be read in multiple steps, but each particle must be read completely regardless of the number of readings; switching to other particles midway is not allowed.

[0080] Step 140: Based on the throttle information, interpolate and fuse multiple target particles under the throttle modes to obtain the sound signal of the target vehicle.

[0081] Optionally, after acquiring the target particles for each throttle mode, the target particles from multiple throttle modes can be interpolated and fused according to the throttle information to synthesize particles corresponding to the throttle information. Sound wave synthesis is then performed based on the synthesized particles to obtain a sound wave signal dynamically matched to the throttle information of the target vehicle. This sound wave signal can be a synthesized internal combustion engine sound; the sound of the exhaust pipe can also be synthesized in the same way, but this embodiment will not provide specific details.

[0082] It should be noted that, if it is necessary to synthesize sound waves from throttle information generated in multiple cycles, the synthesized particles corresponding to the throttle information generated in each cycle can be obtained as described above. These synthesized particles are then added to a particle queue, and the particles in the queue are spliced ​​together in temporal order to obtain the complete synthesized sound wave signal corresponding to the target vehicle running in different cycles. It should also be noted that after the sound wave signal playback is interrupted, a fixed length of audio output can be retrieved from the particle queue to achieve effective playback of the sound wave signal.

[0083] The method provided in this embodiment obtains real sample audio data of a sample vehicle under different throttle modes, extracts particles from the sample audio data under each throttle mode to obtain more realistic and detailed sound wave characteristics under different throttle modes, and performs particle matching in the particle set of each throttle mode according to the throttle information of the target vehicle. Based on the throttle information, the target particles of multiple throttle modes are interpolated and fused to seamlessly connect the particles of different throttle modes, thereby generating a continuous and natural sound wave signal. This achieves sound wave synthesis by using real sound wave characteristics under different throttle modes, so that no matter how the vehicle state changes between idling, acceleration, deceleration, and free coasting, the synthesized sound wave signal can reflect the torque state and speed state of the internal combustion engine in real time and seamlessly, thereby effectively improving the dynamics and quality of sound wave synthesis.

[0084] In some embodiments, step 120 specifically includes:

[0085] For any sample audio data in any of the throttle modes, the multi-order main harmonic data corresponding to each sampling point in the sample audio data are filtered out from the sample audio data to obtain the target signal;

[0086] Based on the order of the main harmonic data in the sample audio data, determine the segmentation code value corresponding to each signal segment in the target signal;

[0087] Based on the segmentation encoding value, particle extraction is performed on each of the signal segments to obtain the particle set in the throttle mode corresponding to the sample audio data;

[0088] The sample audio data under each of the throttle modes are cyclically processed to obtain the particle set under each of the throttle modes.

[0089] Optionally, for sample audio data in any throttle mode, the following steps are performed to extract the particle set in that throttle mode:

[0090] To ensure smooth particle splicing and a seamless transition, it is necessary to ensure that the start and end of all particles are within a fixed phase of the working cycle of one of the cylinders. Since the fundamental frequency of a single cylinder's vibration has been canceled out by multiple cylinders, assuming the number of cylinders is N, after acquiring the Nth-order principal harmonic data corresponding to each sampling point in the sample audio data of the throttle mode, a dynamic bandpass filter can be applied to filter out the Nth-order principal harmonic data from the sample audio data of this throttle mode based on the frequency in the Nth-order principal harmonic data, in order to obtain the target signal. This target signal can be an approximate sine wave signal.

[0091] After acquiring the target signal, the amplitude value of the target signal can be detected. Based on the amplitude value of the target signal, the target signal can be divided into multiple signal segments. For example, weak parts with amplitude values ​​less than a set value can be discarded, while stronger parts with amplitude values ​​greater than or equal to the set value can be acquired to form multiple signal segments.

[0092] After acquiring multiple signal segments, the segmentation coding value corresponding to each signal segment can be determined based on the order of the main harmonic data in the sample audio data; the segmentation coding value here is used to identify the location information or characteristic information of each signal segment that needs to be segmented.

[0093] The method for determining the segmentation coding value corresponding to each signal segment can be the same or different. For example, the order N of the main harmonic data in the sample audio data is multiplied by a preset variable value to obtain the segmentation coding value corresponding to all signal segments. Alternatively, the order N of the main harmonic data in the sample audio data is multiplied by a preset variable value to obtain the segmentation coding value corresponding to some signal segments. Based on the segmentation coding value corresponding to some signal segments and the order N of the main harmonic data in the sample audio data, the segmentation coding value corresponding to another part of the signal segments is obtained again. This embodiment does not specifically limit this.

[0094] The default variable values ​​here are positive integers, and they change adaptively within the set range of values.

[0095] After obtaining the segmentation code value corresponding to each signal segment, the particle segmentation point in each signal segment can be determined according to the segmentation code value corresponding to each signal segment. Then, the particle segment in each signal segment can be divided into particles according to the particle segmentation point in each signal segment to obtain the particle set in this throttle mode.

[0096] For example, in some embodiments, after obtaining the segmentation coding values ​​corresponding to each signal segment, particle extraction can be performed according to the following steps:

[0097] In each of the signal segments, a target zero point whose position coding value matches the segmentation coding value corresponding to each of the signal segments is obtained;

[0098] Using each target zero point as a dividing point, particle extraction is performed on each signal segment to obtain the particle set in the throttle mode corresponding to the sample audio data.

[0099] Optionally, since the segmentation coding values ​​corresponding to each signal segment in the sample audio data under the throttle mode are different, when extracting particles, the target zero point that matches the position coding value with the segmentation coding value of each signal segment can be matched in each signal segment. The target zero point is used as the segmentation point to extract particles from each signal segment, thereby obtaining a particle subset under different signal segments. The particle subsets under different signal segments are combined to obtain the particle set under the throttle mode.

[0100] For example, for the first signal segment in the sample audio data under this throttle mode, its corresponding segmentation code value is 2. N i, where Then take the 2nd one. N Given i zero points (where i is a positive integer), revert these zero point positions to the original audio. The small segments divided by these zero points are the multiple particles obtained from the first signal segment. For the j-th signal segment other than the first signal segment, its corresponding segmentation code value is 2. N i , To adaptively obtain the target segmentation encoding value for the j-th signal segment, the second value is taken. N i There are zero points. These zero point positions are reverted to the original audio. The small segments divided by the zero points are the multiple particles obtained in the j-th signal segment.

[0101] By iterating through the sample audio data in each throttle mode according to the above particle extraction steps, the particle set for each throttle mode can be obtained.

[0102] The method provided in this embodiment extracts particles by filtering out multiple main harmonics, determining the segmentation encoding values ​​of different signal segments according to the order of the main harmonic data in different sample audio data, and then taking the zero point. This method can accurately segment signal segments and extract multiple phase-aligned particles, thereby efficiently obtaining high-quality particle sets in various throttle modes to ensure the smoothness of particle splicing and improve the dynamics of sound wave synthesis.

[0103] In some embodiments, the step of determining the segmentation coding value corresponding to each signal segment includes:

[0104] The segmentation code value corresponding to the first signal segment is determined based on the order of the main harmonic data in the sample audio data;

[0105] Based on the order of the main harmonic data in the sample audio data and the segmentation coding value corresponding to the first signal segment, determine the segmentation coding value corresponding to each second signal segment;

[0106] Wherein, the first signal segment is the first signal segment among the plurality of signal segments, and the second signal segment is the signal segment other than the first signal segment among the plurality of signal segments.

[0107] Optionally, the segmentation coding values ​​corresponding to different signal segments can be determined through the following steps:

[0108] For the first signal segment, the segmentation code value corresponding to the first signal segment can be calculated based on the order of the main harmonic data in the sample audio data. For example, the segmentation code value C1 corresponding to the first signal segment is obtained by multiplying the order N of the main harmonic data in the sample audio data with a preset variable value. Here, the preset variable value is a positive integer and varies adaptively within a set range. Accordingly, the segmentation code value C1 can be specifically expressed as C1=2. N i, where .

[0109] For the j-th signal segment following the first signal segment, i.e., the second signal segment, the segmentation code value can be determined by combining the order of the main harmonic data in the sample audio data and the segmentation code value C1 corresponding to the first signal segment. For example, based on the order of the main harmonic data in the sample audio data, the corresponding target segmentation code value is adaptively determined for the j-th signal segment, and the segmentation code value Cj corresponding to the j-th signal segment is obtained by combining the target segmentation code value with the segmentation code value C1 corresponding to the first signal segment. Accordingly, the segmentation code value Cj can be specifically expressed as Cj=2. N i , This is the target segmentation encoding value adaptively obtained for the j-th signal segment.

[0110] For example, in some embodiments, the specific acquisition of the target segmentation coding value corresponding to the second signal segment includes:

[0111] For any of the second signal segments, multiple candidate segmentation coding values ​​are determined based on the order of the main harmonic data in the sample audio data;

[0112] Obtain the first particle corresponding to each of the candidate segmentation coding values; the first particle is the first particle formed by particle extraction of the second signal segment based on each of the candidate segmentation coding values;

[0113] Among the multiple candidate segmentation coding values, the target segmentation coding value corresponding to the first particle with the highest correlation to the second particle is obtained; the second particle is the last particle in the previous signal segment of the second signal segment;

[0114] The segmentation code value corresponding to the second signal segment is calculated based on the target segmentation code value and the segmentation code value corresponding to the first signal segment.

[0115] By iterating through each of the second signal segments, the segmentation code value corresponding to each second signal segment is obtained.

[0116] Optionally, for any second signal segment, the following steps are performed to obtain the segmentation code value:

[0117] Multiple candidate segmentation coding values ​​are determined based on the order of the main harmonic data in the sample audio data, such as [0, (2 Positive integers contained in the range [N-1)] are used as candidate segmentation code values.

[0118] The segmentation coding value is calculated according to each candidate segmentation coding value. Particle extraction is performed on the second signal segment, and the first particle extracted from the second signal segment under each candidate segmentation coding value is obtained as the first particle corresponding to each candidate segmentation coding value.

[0119] In addition, the last particle extracted from the previous signal segment of the second signal segment is obtained as the second particle.

[0120] Cross-correlation calculation is performed on the first particle corresponding to each candidate segmentation coding value. Among the multiple candidate segmentation coding values, the candidate segmentation coding value corresponding to the first particle with the highest cross-correlation with the second particle is selected as the target segmentation coding value corresponding to the second signal segment.

[0121] The segmentation code value corresponding to the second signal segment is obtained by adding the segmentation code value corresponding to the first signal segment.

[0122] By traversing each second signal segment according to the above steps, the segmentation code value corresponding to each second signal segment can be obtained.

[0123] In summary, the method provided in this embodiment first calculates the segmentation coding value corresponding to the first signal segment based on the order of the main harmonic data in the sample audio data. Then, for each second signal segment after the first signal segment, in addition to considering the order of the main harmonic data, it also combines the segmentation coding value of the first signal segment with the interrelationship between segments to adaptively determine the segmentation coding value corresponding to each second signal segment. This ensures that the selected segmentation coding value is both related to the audio features and maintains continuity, so that the signal segment can be accurately segmented and multiple phase-aligned particles can be extracted based on the segmentation coding value. This allows for the efficient acquisition of high-quality particle sets in each throttle mode, ensuring the smoothness of particle splicing and thus improving the dynamics of sound wave synthesis.

[0124] In some embodiments, step 130 specifically includes:

[0125] Based on the fundamental frequency data corresponding to each sampling point in the sample audio data of each throttle mode, obtain the first rotational speed corresponding to each particle in the particle set of each throttle mode;

[0126] Based on the throttle information, the second speed corresponding to the throttle information is simulated in the transmission simulation model;

[0127] In the particle set of each throttle mode, the particle corresponding to the first rotational speed with the highest matching degree to the second rotational speed is determined as the target particle of each throttle mode.

[0128] Figure 2 This is a schematic diagram of the distribution of harmonic signals in the sample audio data provided by the present invention; Figure 3 This is the second schematic diagram of the sound wave synthesis method provided by the present invention.

[0129] It should be noted that by superimposing the mechanical vibrations of different cylinder blocks, the main harmonic positions of the corresponding audio frequencies for different numbers and arrangements of cylinders can be obtained. For vertically arranged cylinder blocks, the audio frequency of multiple cylinders is equivalent to the single-cylinder audio frequency being continuously shifted to the right according to the ignition phase offset and then added to the resulting audio frequency, similar to a comb filter. For V-shaped cylinder blocks, both vertical and horizontal components must be considered. Therefore, for multi-cylinder engines, regardless of the arrangement or the number of cylinders, the fundamental frequency of the single-cylinder audio frequency will basically cancel each other out by the superposition of the audio frequencies from different cylinder blocks, while the harmonics corresponding to the number of cylinders are basically not canceled out. Figure 2 As shown, the engine of the sample vehicle is a V12 type, which contains a 12-cylinder engine, so the 12th harmonic corresponding to its sample audio data is one of the strongest harmonics.

[0130] Therefore, for a sample vehicle with N cylinders, in order to extract the fundamental frequency data corresponding to each sampling point in the sample audio data of each throttle mode and thus obtain its basic power cycle and speed, the Nth order main harmonic data corresponding to each sampling point in the sample audio data of each throttle mode can be detected first; then, the Nth order main harmonic data corresponding to each sampling point in the sample audio data of each throttle mode can be divided by the number of cylinders N to obtain the fundamental frequency data corresponding to each sampling point in the sample audio data of each throttle mode.

[0131] By acquiring the fundamental frequency data corresponding to each sampling point in the sample audio data of each throttle mode, the engine speed corresponding to each sampling point can be calculated based on the fundamental frequency data. For example, by calculating the fundamental frequency data... 2 The corresponding rotational speed can be obtained by setting the value to 60. Then, based on the rotational speed corresponding to each sampling point, the first rotational speed corresponding to each particle in the particle set under each throttle mode can be obtained.

[0132] After obtaining the first rotation speed corresponding to each particle in the particle set, the particles can be reordered according to the magnitude of the first rotation speed for use in subsequent sound wave information synthesis steps.

[0133] In addition, such as Figure 3 As shown, the throttle information of the target vehicle can also be input into the transmission simulation model, so that the transmission simulation model can simulate the second speed corresponding to the throttle information. The transmission simulation model is a peripheral module that is simulated and run on a computer (PC) for demonstration purposes.

[0134] The specific implementation methods for simulating the second speed using the transmission simulation model include the following:

[0135] Figure 4 This is a schematic diagram illustrating the relative relationship between engine speed and vehicle speed at 20% throttle, as provided by the present invention. Figure 5 This is a schematic diagram illustrating the relative change between engine speed and vehicle speed at 80% throttle, as provided by the present invention; wherein, Figure 4 and Figure 5 In this context, speed refers to vehicle speed, rotatcSpeed ​​refers to engine speed, Gear1, Gear2, Gear3, Gear4, Gear5, and Gear6 represent different gears; nowSpeedRpm, pedalMaxRpm, and minRpm represent the current engine speed, the maximum engine speed limit, and the minimum engine speed limit, respectively.

[0136] like Figure 4 and Figure 5As shown, the transmission simulation model is designed entirely based on a real vehicle transmission. The vehicle speed and engine speed are directly proportional to each gear, differing only by a factor. Therefore, the core of the transmission simulation model lies in the shifting strategy and shift transitions. The principle of designing the shifting strategy is that the greater the throttle, the more inclined to use a lower gear to obtain higher acceleration. The smaller the throttle, the more inclined to use a higher gear to keep the engine running in a more economical and comfortable speed range. When the gear ratios of each gear in the transmission are known, knowing the shift point is equivalent to knowing the shift speed, allowing for real-time calculation of the speed-to-gear mapping at different throttle positions.

[0137] Therefore, as Figure 3 As shown, the first implementation method for simulating the second speed using the transmission simulation model includes: the transmission simulation model has a built-in speed simulation module and a gear logic simulation module; wherein, the speed simulation module is used to simulate the vehicle speed information of the target vehicle based on the throttle information; the gear logic simulation module is used to simulate the gear information of the target vehicle based on the throttle information. Specifically, it works as follows: first, a mapping method is set from the throttle position to the maximum speed under this throttle state, such as linear mapping. After obtaining the vehicle speed, the corresponding speed values ​​are calculated for each of the six gears. Then, gears that do not meet the maximum speed limit under the throttle are removed, and gears that do not meet the minimum speed limit are also removed. The smallest gear is selected from the remaining gears. Furthermore, the speed simulation module is also used to simulate and generate the internal combustion engine speed, i.e., the second speed, based on the vehicle speed information and gear information.

[0138] Furthermore, a second method for simulating the second speed using the transmission simulation model includes: the transmission simulation model can directly use the throttle position. To determine the internal combustion engine speed when the transmission shifts up In other words, the engine speed is simulated based on the mapping relationship between throttle information and speed. The specific calculation formula is as follows:

[0139] ;

[0140] Among them, throttle size This represents data normalized to between 0 and 1. 'b' represents the base speed of the internal combustion engine. This speed satisfies the condition that after upshifting at this speed in all gears, the engine speed in the next gear at the same speed is greater than the engine's idle speed, and the minimum value is taken if this condition is met. This is the maximum speed of the internal combustion engine.

[0141] After obtaining the second rotation speed corresponding to the throttle information and the first rotation speed corresponding to each particle, the particle splicing module is called to load the particle set of each throttle mode. In the particle set of each throttle mode, the particle corresponding to the first rotation speed with the highest matching degree with the second rotation speed corresponding to the throttle information is selected and read out as the target particle matching the throttle information in each throttle mode. Based on the specific throttle information, the target particles matching the throttle information in each throttle mode are interpolated and fused to synthesize the corresponding sound wave signal.

[0142] The method provided in this embodiment obtains the first rotational speed corresponding to each particle by substituting the fundamental frequency data of the sampling points in the sample audio data of each throttle mode. Then, based on the throttle information, the corresponding second rotational speed is simulated in the transmission simulation model. Finally, the particle corresponding to the first rotational speed with the highest matching degree with the second rotational speed is determined from the particle set of each throttle mode as the target particle. In order to ensure, through physical simulation and particle matching, when the throttle changes in the later splicing stage, the throttle change not only indirectly affects the rotational speed through speed and thus affects the audio, but also allows the timbre change after the engine power change to be immediately experienced, thereby improving the dynamics and quality of the sound synthesis.

[0143] In some embodiments, step 140 specifically includes:

[0144] Based on the throttle information, the torque parameters corresponding to the throttle information are simulated in the transmission simulation model;

[0145] Based on the torque parameter, interpolation processing is performed on multiple target particles under the throttle mode to obtain a composite particle with the rotational speed corresponding to the torque parameter;

[0146] The acoustic signal of the target vehicle is synthesized based on the synthesized particles.

[0147] Optionally, the throttle information can be input into the transmission simulation model so that the torque parameters corresponding to the throttle information can be simulated by the transmission simulation model.

[0148] For example, the transmission simulation model can first simulate the throttle position. The relationship between throttle position and power P is established to obtain the power corresponding to the throttle information. Then, the power corresponding to the throttle information is divided by the speed corresponding to the throttle information to obtain the torque parameter corresponding to the throttle information. Here, throttle position... The relationship between the power P and the power P can be expressed by the following formula:

[0149] ;

[0150] in, The value is between 0 and 1. Its design principle is mainly to satisfy the principle that throttle and power are proportional, and that power increases faster in the initial stage of throttle increase. That is, when n is small, power increases faster in the initial stage of throttle increase; when n is large, the effect of throttle increase on power increase is more stable.

[0151] In the sound wave synthesis process, since the number of target particles is the same under different throttle modes, and the segment length of the target particles is the same under the same speed, during the dynamic changes of the vehicle, the target particles for each throttle mode can be obtained from the particle set of each throttle mode based on the speed. Then, based on the torque parameter, the synthesized particles of the speed corresponding to the torque parameter are obtained by interpolation of the target particles under multiple throttle modes. Based on these synthesized particles, the sound wave signal of the target vehicle is synthesized, so that no matter how the vehicle state changes between idling, acceleration, deceleration, free coasting, etc., the synthesized sound wave signal can reflect the torque state and speed state of the internal combustion engine in real time and seamlessly, thereby improving the dynamics and quality of the sound wave synthesis.

[0152] In some embodiments, the step of obtaining the multi-order principal harmonic data corresponding to each sampling point in the sample audio data of each of the throttle modes includes:

[0153] Each sampling point in the sample audio data of each of the aforementioned throttle modes is divided into a first sampling point or a second sampling point according to a preset interval;

[0154] For any first sampling point, based on the position information of the multi-order main harmonic data of the previous sampling point, the multi-order main harmonic data corresponding to the first sampling point is searched in the frequency domain data sequence corresponding to the first sampling point to obtain the multi-order main harmonic data of the first sampling point.

[0155] By traversing each of the first sampling points, the multi-order main harmonic data corresponding to each of the first sampling points are obtained;

[0156] Based on the multi-order principal harmonic data corresponding to each of the first sampling points, interpolation estimation is performed on each of the second sampling points to obtain the multi-order principal harmonic data corresponding to each of the second sampling points.

[0157] It should be noted that, in the process of harmonic data extraction, to simplify the computational workload, each sampling point in the sample audio data of each throttle mode can be divided into a first sampling point or a second sampling point according to a preset interval. The preset interval can be set according to actual needs, such as using every 4800 sampling points as an interval, so that a first sampling point is determined every 4800 sampling points, and the main harmonic data is extracted once for the first sampling point. For other sampling points, i.e., the second sampling points, their corresponding main harmonic data are estimated through interpolation calculations.

[0158] Optionally, for any first sampling point, the following steps are performed to extract harmonic data:

[0159] In the sample audio data, the time-domain data sequence corresponding to the sampling point is obtained; a frequency-domain transformation algorithm, such as the short-time Fourier transform algorithm, is used to transform the time-domain data sequence corresponding to the first sampling point to obtain the frequency-domain data sequence corresponding to the first sampling point. The parameters in the frequency-domain transformation algorithm can be set according to actual needs, such as a basic audio sampling rate of 48K and a window length of 65536.

[0160] Subsequently, using historical primary harmonic results as prior information, data corresponding to the new maximum energy value is obtained near the maximum energy value of the historical primary harmonic results in the frequency domain data sequence corresponding to the first sampling point, and this data is used as the multi-order primary harmonic data corresponding to the first sampling point. Specifically, based on the location information of the multi-order primary harmonic data of the previous sampling point, the target search range is determined. Within the target search range in the frequency domain data sequence corresponding to the first sampling point, the data corresponding to the maximum energy value is searched and used as the multi-order primary harmonic data corresponding to the first sampling point.

[0161] By iterating through each first sampling point, the multi-order main harmonic data corresponding to each first sampling point can be obtained.

[0162] Furthermore, since the changes in the main harmonic data between adjacent sampling points have a certain similarity, in order to simplify the calculation, interpolation estimation can be performed on each second sampling point based on the multi-order main harmonic data corresponding to each first sampling point to obtain the multi-order main harmonic data corresponding to the second sampling point.

[0163] In summary, by following the steps described above, we can efficiently and accurately acquire the multi-order main harmonic data corresponding to all sampling points while reducing the amount of computation.

[0164] The sound wave synthesis device provided by the present invention is described below. The sound wave synthesis device described below can be referred to in correspondence with the sound wave synthesis method described above.

[0165] Figure 6 This is a schematic diagram of the sound wave synthesis device provided by the present invention; as shown. Figure 6 As shown, the device includes:

[0166] The data acquisition unit 610 is used to acquire sample audio data of the sample vehicle under different throttle modes;

[0167] The particle extraction unit 620 is used to extract particles from the sample audio data of each throttle mode based on the multi-order main harmonic data corresponding to each sampling point in the sample audio data of each throttle mode, so as to obtain the particle set of each throttle mode.

[0168] The particle matching unit 630 is used to perform particle matching in the particle set of each throttle mode according to the throttle information of the target vehicle, so as to obtain the target particle of each throttle mode, wherein the target particle is the particle that matches the throttle information.

[0169] The sound wave synthesis unit 640 is used to interpolate and fuse multiple target particles under the throttle mode according to the throttle information to obtain the sound wave signal of the target vehicle.

[0170] The device provided in this embodiment acquires real sample audio data of a sample vehicle under different throttle modes, extracts particles from the sample audio data under each throttle mode to obtain more realistic and detailed sound wave characteristics under different throttle modes, and performs particle matching in the particle set of each throttle mode according to the throttle information of the target vehicle. Based on the throttle information, it interpolates and fuses the target particles of multiple throttle modes to seamlessly connect the particles of different throttle modes, thereby generating a continuous and natural sound wave signal. This enables the synthesis of sound waves through real sound wave characteristics under different throttle modes, so that no matter how the vehicle state changes between idling, acceleration, deceleration, and free coasting, the synthesized sound wave signal can reflect the torque and speed state of the internal combustion engine in real time and seamlessly, thereby effectively improving the dynamics and quality of sound wave synthesis.

[0171] In some embodiments, the particle extraction unit is specifically used for:

[0172] For any sample audio data in any of the throttle modes, the multi-order main harmonic data corresponding to each sampling point in the sample audio data are filtered out from the sample audio data to obtain the target signal;

[0173] Based on the order of the main harmonic data in the sample audio data, determine the segmentation code value corresponding to each signal segment in the target signal;

[0174] Based on the segmentation encoding value, particle extraction is performed on each of the signal segments to obtain the particle set in the throttle mode corresponding to the sample audio data;

[0175] The sample audio data under each of the throttle modes are cyclically processed to obtain the particle set under each of the throttle modes.

[0176] In some embodiments, the particle extraction unit is further configured to:

[0177] In each of the signal segments, a target zero point whose position coding value matches the segmentation coding value corresponding to each of the signal segments is obtained;

[0178] Using each target zero point as a dividing point, particle extraction is performed on each signal segment to obtain the particle set in the throttle mode corresponding to the sample audio data.

[0179] In some embodiments, the particle extraction unit is further configured to:

[0180] The segmentation code value corresponding to the first signal segment is determined based on the order of the main harmonic data in the sample audio data;

[0181] Based on the order of the main harmonic data in the sample audio data and the segmentation coding value corresponding to the first signal segment, determine the segmentation coding value corresponding to each second signal segment;

[0182] Wherein, the first signal segment is the first signal segment among the plurality of signal segments, and the second signal segment is the signal segment other than the first signal segment among the plurality of signal segments.

[0183] In some embodiments, the particle extraction unit is further configured to:

[0184] For any of the second signal segments, multiple candidate segmentation coding values ​​are determined based on the order of the main harmonic data in the sample audio data;

[0185] Obtain the first particle corresponding to each of the candidate segmentation coding values; the first particle is the first particle formed by particle extraction of the second signal segment based on each of the candidate segmentation coding values;

[0186] Among the multiple candidate segmentation coding values, the target segmentation coding value corresponding to the first particle with the highest correlation to the second particle is obtained; the second particle is the last particle in the previous signal segment of the second signal segment;

[0187] The segmentation code value corresponding to the second signal segment is calculated based on the target segmentation code value and the segmentation code value corresponding to the first signal segment.

[0188] By iterating through each of the second signal segments, the segmentation code value corresponding to each second signal segment is obtained.

[0189] In some embodiments, the particle matching unit is specifically used for:

[0190] Based on the fundamental frequency data corresponding to each sampling point in the sample audio data of each throttle mode, obtain the first rotational speed corresponding to each particle in the particle set of each throttle mode;

[0191] Based on the throttle information, the second speed corresponding to the throttle information is simulated in the transmission simulation model;

[0192] In the particle set of each throttle mode, the particle corresponding to the first rotational speed with the highest matching degree to the second rotational speed is determined as the target particle of each throttle mode.

[0193] In some embodiments, the sound wave synthesis unit is specifically used for:

[0194] Based on the throttle information, the torque parameters corresponding to the throttle information are simulated in the transmission simulation model;

[0195] Based on the torque parameter, interpolation processing is performed on multiple target particles under the throttle mode to obtain a composite particle with the rotational speed corresponding to the torque parameter;

[0196] The acoustic signal of the target vehicle is synthesized based on the synthesized particles.

[0197] In some embodiments, the data acquisition unit is specifically used for:

[0198] Each sampling point in the sample audio data of each of the aforementioned throttle modes is divided into a first sampling point or a second sampling point according to a preset interval;

[0199] For any first sampling point, based on the position information of the multi-order main harmonic data of the previous sampling point, the multi-order main harmonic data corresponding to the first sampling point is searched in the frequency domain data sequence corresponding to the first sampling point to obtain the multi-order main harmonic data of the first sampling point.

[0200] By traversing each of the first sampling points, the multi-order main harmonic data corresponding to each of the first sampling points are obtained;

[0201] Based on the multi-order principal harmonic data corresponding to each of the first sampling points, interpolation estimation is performed on each of the second sampling points to obtain the multi-order principal harmonic data corresponding to each of the second sampling points.

[0202] The apparatus provided by the present invention is used to execute the above-described method embodiments. For specific processes and details, please refer to the above embodiments, which will not be repeated here.

[0203] Figure 7 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 7As shown, the electronic device may include a processor 710, a communications interface 720, a memory 730, and a communication bus 740, wherein the processor 710, the communications interface 720, and the memory 730 communicate with each other via the communication bus 740. The processor 710 can call logical instructions in the memory 730 to execute a sound wave synthesis method, which includes: acquiring sample audio data of a sample vehicle under different throttle modes; extracting particles from the sample audio data under each throttle mode based on the multi-order main harmonic data corresponding to each sampling point in the sample audio data under each throttle mode, obtaining a particle set for each throttle mode; matching particles in the particle sets under each throttle mode based on the throttle information of the target vehicle, obtaining target particles for each throttle mode, wherein the target particles are particles that match the throttle information; and interpolating and fusing the target particles under multiple throttle modes based on the throttle information to obtain the sound wave signal of the target vehicle.

[0204] Furthermore, the logical instructions in the aforementioned memory 730 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0205] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the sound wave synthesis method provided by the above methods. The method includes: acquiring sample audio data of a sample vehicle under different throttle modes; extracting particles from the sample audio data under each throttle mode based on the multi-order main harmonic data corresponding to each sampling point in the sample audio data under each throttle mode to obtain a particle set under each throttle mode; matching particles in the particle set under each throttle mode based on the throttle information of the target vehicle to obtain target particles under each throttle mode, wherein the target particles are particles that match the throttle information; and interpolating and fusing the target particles under multiple throttle modes based on the throttle information to obtain the sound wave signal of the target vehicle.

[0206] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the sound wave synthesis method provided by the above methods. The method includes: acquiring sample audio data of a sample vehicle under different throttle modes; extracting particles from the sample audio data under each throttle mode based on the multi-order principal harmonic data corresponding to each sampling point in the sample audio data under each throttle mode, to obtain a particle set under each throttle mode; matching particles in the particle sets under each throttle mode based on the throttle information of a target vehicle, to obtain target particles under each throttle mode, wherein the target particles are particles matched with the throttle information; and interpolating and fusing the target particles under multiple throttle modes based on the throttle information to obtain the sound wave signal of the target vehicle.

[0207] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0208] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0209] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for synthesizing sound waves, characterized in that, include: Based on the multi-order main harmonic data corresponding to each sampling point in the sample audio data of the sample vehicle under different throttle modes, particle extraction is performed on the sample audio data under each throttle mode to obtain the particle set under each throttle mode. Based on the throttle information of the target vehicle, particle matching is performed in the particle set of each throttle mode to obtain the target particle of each throttle mode, wherein the target particle is the particle that matches the throttle information. Based on the throttle information, interpolation and fusion are performed on multiple target particles under the throttle modes to obtain the sound signal of the target vehicle; The steps for obtaining the particle set include: For any sample audio data in any of the throttle modes, the multi-order main harmonic data corresponding to each sampling point in the sample audio data are filtered out from the sample audio data to obtain the target signal; Based on the order of the main harmonic data in the sample audio data, the segmentation coding value corresponding to the first signal segment is determined, and based on the order of the main harmonic data and the segmentation coding value corresponding to the first signal segment, the segmentation coding value corresponding to each second signal segment is determined. Based on the segmentation coding value, particle extraction is performed on each signal segment in the target signal to obtain the particle set in the throttle mode corresponding to the sample audio data; The sample audio data under each of the throttle modes are cyclically processed to obtain the particle set under each of the throttle modes; The first signal segment is the first of a plurality of signal segments, and the second signal segment is a signal segment other than the first signal segment among the plurality of signal segments; The steps for determining the segmentation code value corresponding to each of the second signal segments include: For any of the second signal segments, multiple candidate segmentation coding values ​​are determined based on the order of the main harmonic data in the sample audio data; Obtain the first particle corresponding to each of the candidate segmentation coding values; the first particle is the first particle formed by particle extraction of the second signal segment based on each of the candidate segmentation coding values; Among the multiple candidate segmentation coding values, the target segmentation coding value corresponding to the first particle with the highest correlation to the second particle is obtained; the second particle is the last particle in the previous signal segment of the second signal segment; The segmentation code value corresponding to the second signal segment is calculated based on the target segmentation code value and the segmentation code value corresponding to the first signal segment. By iterating through each of the second signal segments, the segmentation code value corresponding to each second signal segment is obtained.

2. The sound wave synthesis method according to claim 1, characterized in that, The step of extracting particles from each signal segment in the target signal based on the segmentation encoding value to obtain the particle set in the throttle mode corresponding to the sample audio data includes: In each of the signal segments, a target zero point whose position coding value matches the segmentation coding value corresponding to each of the signal segments is obtained; Using each target zero point as a dividing point, particle extraction is performed on each signal segment to obtain the particle set in the throttle mode corresponding to the sample audio data.

3. The sound wave synthesis method according to any one of claims 1-2, characterized in that, The step of performing particle matching in the particle sets of each throttle mode based on the throttle information of the target vehicle to obtain the target particles of each throttle mode includes: Based on the fundamental frequency data corresponding to each sampling point in the sample audio data of each throttle mode, obtain the first rotational speed corresponding to each particle in the particle set of each throttle mode; Based on the throttle information, the second speed corresponding to the throttle information is simulated in the transmission simulation model; In the particle set of each throttle mode, the particle corresponding to the first rotational speed with the highest matching degree to the second rotational speed is determined as the target particle of each throttle mode.

4. The sound wave synthesis method according to any one of claims 1-2, characterized in that, The step of interpolating and fusing multiple target particles under the throttle modes based on the throttle information to obtain the sound signal of the target vehicle includes: Based on the throttle information, the torque parameters corresponding to the throttle information are simulated in the transmission simulation model; Based on the torque parameter, interpolation processing is performed on multiple target particles under the throttle mode to obtain a composite particle with the rotational speed corresponding to the torque parameter; The acoustic signal of the target vehicle is synthesized based on the synthesized particles.

5. The sound wave synthesis method according to any one of claims 1-2, characterized in that, The steps for obtaining the multi-order principal harmonic data corresponding to each sampling point in the sample audio data of each of the throttle modes include: Each sampling point in the sample audio data of each of the aforementioned throttle modes is divided into a first sampling point or a second sampling point according to a preset interval; For any first sampling point, based on the position information of the multi-order main harmonic data of the previous sampling point, the multi-order main harmonic data corresponding to the first sampling point is searched in the frequency domain data sequence corresponding to the first sampling point to obtain the multi-order main harmonic data of the first sampling point. By traversing each of the first sampling points, the multi-order main harmonic data corresponding to each of the first sampling points are obtained; Based on the multi-order principal harmonic data corresponding to each of the first sampling points, interpolation estimation is performed on each of the second sampling points to obtain the multi-order principal harmonic data corresponding to each of the second sampling points.

6. A sound wave synthesis device, characterized in that, include: The data acquisition unit is used to acquire sample audio data of the sample vehicle under different throttle modes; The particle extraction unit is used to extract particles from the sample audio data of each throttle mode based on the multi-order main harmonic data corresponding to each sampling point in the sample audio data of each throttle mode, so as to obtain the particle set of each throttle mode. The particle matching unit is used to perform particle matching in the particle set of each throttle mode according to the throttle information of the target vehicle, so as to obtain the target particle of each throttle mode, wherein the target particle is the particle that matches the throttle information. The sound wave synthesis unit is used to interpolate and fuse multiple target particles under the throttle mode according to the throttle information to obtain the sound wave signal of the target vehicle. The steps for obtaining the particle set include: For any sample audio data in any of the throttle modes, the multi-order main harmonic data corresponding to each sampling point in the sample audio data are filtered out from the sample audio data to obtain the target signal; Based on the order of the main harmonic data in the sample audio data, the segmentation coding value corresponding to the first signal segment is determined, and based on the order of the main harmonic data and the segmentation coding value corresponding to the first signal segment, the segmentation coding value corresponding to each second signal segment is determined. Based on the segmentation coding value, particle extraction is performed on each signal segment in the target signal to obtain the particle set in the throttle mode corresponding to the sample audio data; The sample audio data under each of the throttle modes are cyclically processed to obtain the particle set under each of the throttle modes; The first signal segment is the first of a plurality of signal segments, and the second signal segment is a signal segment other than the first signal segment among the plurality of signal segments; The steps for determining the segmentation code value corresponding to each of the second signal segments include: For any of the second signal segments, multiple candidate segmentation coding values ​​are determined based on the order of the main harmonic data in the sample audio data; Obtain the first particle corresponding to each of the candidate segmentation coding values; the first particle is the first particle formed by particle extraction of the second signal segment based on each of the candidate segmentation coding values; Among the multiple candidate segmentation coding values, the target segmentation coding value corresponding to the first particle with the highest correlation to the second particle is obtained; the second particle is the last particle in the previous signal segment of the second signal segment; The segmentation code value corresponding to the second signal segment is calculated based on the target segmentation code value and the segmentation code value corresponding to the first signal segment. By iterating through each of the second signal segments, the segmentation code value corresponding to each second signal segment is obtained.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the sound wave synthesis method as described in any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the sound wave synthesis method as described in any one of claims 1 to 5.

Images