A sound masking control method, device, air conditioner, and storage medium

Abstract

This invention discloses a sound masking control method, device, air conditioner, and storage medium. The invention relates to the field of energy-saving air conditioning technology. The method includes: acquiring the current operating mode, speed change rate, air guide vane motion state determination result, and resonance pass-through determination result; inputting the speed change rate, resonance pass-through determination result, air guide vane motion state determination result, and current operating mode into a pre-constructed auditory abruptness evaluation model to obtain an auditory abruptness level; determining a masking sound frequency band configuration scheme based on the auditory abruptness level, generating a masking sound according to the frequency band configuration scheme, and playing the masking sound before the fan speed switching command is executed. This invention, by pre-playing an appropriate masking sound before the actual fan speed switch, actively implements sound masking before noise abrupt changes occur, achieving a smooth auditory transition during dynamic switching, effectively suppressing auditory contrast at the moment of switching, and improving the user's auditory experience.

Description

Technical Field

[0001] This invention relates to the field of energy-saving air conditioning technology, and in particular to a sound masking control method, device, air conditioner, and storage medium. Background Technology

[0002] As household air conditioners continue to upgrade towards quieter and smarter operation, users are increasingly demanding higher levels of comfort from the noise levels of indoor units. Existing noise reduction technologies for indoor air conditioners mainly rely on passive noise reduction methods such as optimized fan blade design, improved duct structure, and vibration damping design of the casing. While these solutions can reduce the noise level of the indoor unit during steady-state operation to some extent, they cannot solve the problem of deteriorated auditory comfort caused by sudden changes in noise sound pressure level and drastic changes in spectral characteristics due to sudden changes in fan speed during dynamic processes such as fan speed switching and air guide plate angle adjustment.

[0003] In the prior art, patent CN116608585A discloses a novel white noise air conditioner, which utilizes the spectral coverage characteristics of white sound to improve the noise problem of the indoor unit of the air conditioner by spreading the sound frequency spectrum. It can convert the operating noise of the air conditioner itself into white sound to achieve a noise reduction effect. However, it only masks the noise generated under stable operating conditions of the indoor unit and cannot improve the noise abrupt changes during dynamic changes in indoor unit parameters such as fan speed switching and air guide plate adjustment.

[0004] Patent CN118980176A discloses an indoor air conditioner unit with regional active noise cancellation function, an air conditioner, and its working method. This solution reduces noise in the indoor unit through active noise cancellation technology. The indoor air conditioner unit includes a main controller, a reference sound acquisition device for collecting indoor unit noise, an error sound acquisition device for acquiring noise in the target noise reduction area, and a speaker for generating secondary noise. After acquiring the indoor unit noise and the noise in the target noise reduction area, the main controller outputs secondary noise to the speaker to achieve relative cancellation of sound waves. This method relies on an active noise cancellation technology system and requires the configuration of an error microphone, a secondary sound source, and an adaptive filtering algorithm. It has inherent drawbacks such as high hardware costs, difficulty in arranging acoustic devices in the limited space of the indoor unit, complex control logic, and susceptibility to secondary noise. It is difficult to adapt to the needs of low-cost, easily deployable household air conditioner products, and it also fails to effectively solve the problem of abrupt noise generated during changes in indoor unit operating parameters. Summary of the Invention

[0005] The purpose of this invention is to provide a sound masking control method, device, air conditioner, and storage medium, which aims to solve the problem of poor listening comfort caused by sudden noise changes during the switching of the fan speed in an air conditioner.

[0006] In a first aspect, embodiments of the present invention provide a sound masking control method, comprising: The system acquires the current operating parameters and fan speed switching command of the indoor unit, and determines the target fan speed and target air guide plate angle based on the fan speed switching command; the current operating parameters include the current fan speed, the current air guide plate angle, and the current operating mode; The rate of change of rotation speed is determined based on the current fan speed and the target fan speed. The change in the angle of the air guide plate is determined based on the current angle of the air guide plate and the angle of the target air guide plate, and the change in the angle of the air guide plate is compared with a preset angle threshold to obtain the result of the motion state determination of the air guide plate. The resonant point speed of the indoor unit's fan blades is obtained, and it is determined whether the current fan speed changes to the target fan speed and passes through the resonant point speed of the fan blades, so as to obtain the resonance crossing determination result. The rotational speed change rate, the resonance crossing determination result, the wind deflector motion state determination result, and the current operating mode are input into the pre-constructed auditory abruptness assessment model to obtain the auditory abruptness level; The frequency band configuration scheme of the masking sound is determined according to the auditory abruptness level, and the masking sound is generated according to the frequency band configuration scheme. The masking sound is played before the fan speed switching command is executed.

[0007] Secondly, embodiments of the present invention provide a sound masking control device, comprising: The acquisition module is used to acquire the current operating parameters and fan speed switching command of the indoor unit, and determine the target fan speed and target air guide plate angle according to the fan speed switching command; the current operating parameters include the current fan speed, the current air guide plate angle and the current operating mode; The determining module is used to determine the rate of change of rotational speed based on the current fan speed and the target fan speed; The comparison module is used to determine the change in the air guide plate angle based on the current air guide plate angle and the target air guide plate angle, and compare the change in the air guide plate angle with a preset angle threshold to obtain the air guide plate motion state determination result. The judgment module is used to obtain the resonant point speed of the fan blades of the indoor unit, determine whether the current fan speed changes to the target fan speed and passes through the resonant point speed of the fan blades, and obtain the resonance crossing judgment result. The input module is used to input the rotational speed change rate, the resonance crossing determination result, the wind deflector motion state determination result, and the current operating mode into the pre-constructed auditory abruptness assessment model to obtain the auditory abruptness level; The generation module is used to determine the frequency band configuration scheme of the masking sound according to the auditory abruptness level, generate the masking sound according to the frequency band configuration scheme, and play the masking sound before the fan speed switching command is executed.

[0008] Thirdly, embodiments of the present invention provide an air conditioner, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the sound masking control method described in the first aspect above.

[0009] Fourthly, embodiments of the present invention also provide a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program that, when executed by a processor, implements the sound masking control method described in the first aspect.

[0010] This invention discloses a sound masking control method, device, air conditioner, and storage medium. The method includes: acquiring the current operating parameters of the indoor unit and a fan speed switching command, and determining a target fan speed and a target air guide plate angle based on the fan speed switching command; the current operating parameters include the current fan speed, the current air guide plate angle, and the current operating mode; determining the speed change rate based on the current fan speed and the target fan speed; determining the air guide plate angle change amount based on the current air guide plate angle and the target air guide plate angle, and comparing the air guide plate angle change amount with a preset angle threshold to obtain the air guide plate movement. The system determines the state of the fan blades and obtains the resonant point speed of the indoor unit. It then determines whether the fan speed changes from the current speed to the target speed, passing through the resonant point speed, thus obtaining a resonance crossing determination result. The system inputs the speed change rate, the resonance crossing determination result, the air guide plate movement state determination result, and the current operating mode into a pre-built auditory abruptness assessment model to obtain an auditory abruptness level. Based on the auditory abruptness level, a masking sound frequency band configuration scheme is determined, and a masking sound is generated according to the frequency band configuration scheme. This masking sound is played before the fan speed switching command is executed. This invention collects current operating parameters and fan speed switching commands, determines the speed change rate, air guide plate angle change, and resonance crossing determination result, and inputs these parameters into a pre-built auditory abruptness assessment model to achieve a quantitative prediction of auditory abruptness during wind speed switching. Based on the prediction level, a multi-frequency masking sound configuration scheme is dynamically determined, and an appropriate masking sound is played before the actual fan speed switch, thereby proactively masking the sound before a sudden noise change occurs. This invention achieves a smooth auditory transition during dynamic switching, effectively suppressing auditory contrast at the moment of switching; simultaneously, it eliminates the need for expensive hardware such as error microphones and secondary sound sources required for active noise cancellation, thus improving the user's auditory experience at a low cost. The embodiments of this invention also provide a sound masking control device, a computer-readable storage medium, and an air conditioner, all possessing the aforementioned beneficial effects, which will not be elaborated upon further here. Attached Figure Description

[0011] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 A flowchart illustrating the sound masking control method; Figure 2 Another flowchart illustrating the sound masking control method; Figure 3This is a schematic diagram of the sub-processes of the sound masking control method; Figure 4 A flowchart illustrating the process of building an auditory abruptness assessment model; Figure 5 A schematic diagram of the sub-processes for constructing an auditory abruptness assessment model; Figure 6 This is a schematic diagram of another sub-process of the sound masking control method; Figure 7 A schematic block diagram of a sound masking control device. Detailed Implementation

[0013] The technical solutions of the embodiments of the present 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 the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0014] It should be understood that, when used in this specification and the appended claims, the terms “comprising” and “including” indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more of its features, integrals, steps, operations, elements, components and / or collections thereof.

[0015] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0016] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0017] Please see Figure 1 and Figure 2 This embodiment provides a sound masking control method, including: S101: Obtain the current operating parameters and fan speed switching command of the indoor unit, and determine the target fan speed and target air guide plate angle according to the fan speed switching command; the current operating parameters include the current fan speed, the current air guide plate angle and the current operating mode; After the indoor unit of the air conditioner is powered on, the mainboard controller collects the current operating parameters in real time. Specifically, the controller obtains the current fan speed value n1 through the Hall sensor of the drive motor or the speed feedback circuit, obtains the current air guide vane angle value S1 through the drive pulse count of the stepper motor of the air guide vane or the angle sensor, and reads the current operating mode M stored in the system register. The operating modes include sleep mode, silent mode, normal mode and powerful mode, etc., and different modes correspond to different fan speed ranges and air volume output characteristics.

[0018] The controller then continuously monitors control signals from the remote control, wired controller, or wireless communication module. When the user presses the fan speed adjustment button or the system automatically triggers fan speed adjustment based on the temperature difference, the controller captures the fan speed switching command. It analyzes the target fan speed setting information contained in the command and, combined with the preset speed value corresponding to that setting in the current operating mode, determines the target fan speed n2. Simultaneously, it analyzes any guide vane angle adjustment information that may be included in the command. If the command does not contain guide vane angle adjustment information, the current guide vane angle is used as the target guide vane angle S2.

[0019] For example, when a user switches the fan speed from low to high using a remote control, the controller reads the switching command and, based on the current operating mode being normal mode, looks up the table to find that the corresponding speed for low fan speed is 400 revolutions per minute and the corresponding speed for high fan speed is 800 revolutions per minute. At the same time, it reads that the current air guide plate angle is 30 degrees. Since the command does not include an air guide plate adjustment requirement, the target air guide plate angle remains 30 degrees.

[0020] For example, during cooling operation, when the difference between the indoor temperature and the set temperature exceeds three degrees Celsius, the system automatically triggers a switch from silent mode to high-power mode. The controller determines the target fan speed and the current air guide vane angle based on the automatic switching command. After obtaining the above parameters, the controller stores the current fan speed, target fan speed, current air guide vane angle, target air guide vane angle, and current operating mode in the cache for subsequent steps.

[0021] S102: Determine the rate of change of rotation speed based on the current fan speed and the target fan speed; Specifically, after acquiring the current fan speed and the target fan speed, the controller calculates the speed change rate based on the difference between the two and the speed switching execution duration. The speed switching execution duration is determined by the drive parameters of the indoor unit controller, which are pre-stored in the controller. The specific value of these parameters is related to the type of motor used in the indoor unit, the response characteristics of the drive chip, and the fan's moment of inertia. The controller reads the current fan speed and the target fan speed from the buffer, calculates the difference between the target fan speed and the current fan speed. This difference can be positive or negative; a positive value indicates an acceleration process, and a negative value indicates a deceleration process. The controller then reads the speed switching execution duration. This duration represents the time span required from when the controller sends a drive signal to when the fan actually reaches the target speed. The controller divides the speed difference by the speed switching execution duration to obtain the speed change rate. For example, if the current fan speed is 400 revolutions per minute and the target fan speed is 800 revolutions per minute, the difference is +400 revolutions per minute. If the speed switching execution time is one second, then the speed change rate is 400 revolutions per second. As another example, if the current fan speed is 900 revolutions per minute and the target fan speed is 500 revolutions per minute, the difference is -400 revolutions per minute. If the switching execution time is 1.2 seconds, then the speed change rate is -333 revolutions per second.

[0022] During the calculation, the controller takes the absolute value of the rotational speed change rate for subsequent auditory abruptness assessment, because abruptness is related to the rate of rotational speed change but not to the direction of acceleration or deceleration. After the calculation is complete, the controller stores the rotational speed change rate in floating-point format at a designated address in the cache, which can be used as an input parameter when calling the auditory abruptness assessment model later. The controller also records the rotational speed switching execution time used during the calculation. This time is also used to determine the pre-start time of the masking sound, that is, to determine how far in advance the masking sound should be started before the wind turbine actually begins to change speed.

[0023] S103: Determine the change in the air guide plate angle based on the current air guide plate angle and the target air guide plate angle, and compare the change in the air guide plate angle with a preset angle threshold to obtain the air guide plate motion state determination result; Specifically, after obtaining the current and target air guide vane angles, the controller reads these angles from the buffer, calculates the difference between the target and current air guide vane angles, and takes the absolute value of this difference to obtain the change in air guide vane angle. The change in the angle of the air guide vane represents the total amplitude of rotation required by the vane during wind speed switching. The magnitude of this value directly affects the degree of change in the cross-section of the air duct outlet, and thus affects the spectral characteristics of airflow noise. For example, if the current air guide vane angle is 20 degrees and the target air guide vane angle is 45 degrees, then the angle change is 25 degrees.

[0024] The controller reads a preset angle threshold from its internal memory. This threshold is determined through testing and calibration before the air conditioner leaves the factory. The calibration is based on the following criteria: when the air guide plate swings at a small angle, its impact on airflow distribution and noise characteristics is minimal, and the resulting noise change is barely perceptible to the human ear; when the swing amplitude of the air guide plate exceeds a certain critical value, the duct outlet area changes significantly, and the vortex noise and turbulence noise generated by the airflow impacting the surface of the air guide plate are significantly enhanced. This critical value is used as the preset angle threshold. In this embodiment, the preset angle threshold ranges from 2 to 5 degrees.

[0025] The controller compares the calculated change in the air guide vane angle with a preset angle threshold. When the change in angle is greater than the preset threshold, the air guide vane is determined to be rotating, and a motion state determination result of "rotating" is generated. When the change in angle is less than or equal to the preset threshold, the air guide vane is determined to be stationary, and a motion state determination result of "stationary" is generated. For example, if the preset angle threshold is 5 degrees and the change in angle is 25 degrees, which is greater than 5 degrees, it is determined to be rotating; conversely, if the change in angle is 2 degrees, which is less than 5 degrees, it is determined to be stationary.

[0026] The controller stores the determination result of the air deflector's motion state in a buffer as a digital identifier, specifically represented by a single binary bit. For example, 1 represents a rotating state, and 0 represents a stationary state. This determination result serves as one of the input parameters for the subsequent auditory abruptness assessment model, and is used together with the rotational speed change rate, resonance crossing determination result, and current operating mode to calculate the auditory abruptness level. The controller also retains the angle change itself in the buffer for reference when configuring the frequency band for generating the masking sound. This is because the larger the angle change, the more prominent the low-frequency noise components caused by airflow disturbance become, requiring corresponding adjustments to the output intensity of the low-frequency masking sound.

[0027] S104: Obtain the resonant speed of the indoor unit's fan blades, determine whether the current fan speed changes to the target fan speed and passes through the resonant speed of the fan blades, and obtain the resonance crossing determination result; Specifically, after acquiring the current fan speed and the target fan speed, the controller performs a resonance pass-through determination. The indoor unit's fan blades exhibit mechanical resonance at specific speeds; this speed is called the fan blade resonance point speed. The resonance point speed is determined by the fan blades' natural frequency and is measured during the air conditioning product development phase through frequency sweep testing. The specific testing method is as follows: the indoor unit is installed in a standard test fixture within an anechoic chamber, and the fan is driven to slowly accelerate from a standstill to its maximum speed. Simultaneously, an acceleration sensor is placed on the surface of the indoor unit's casing to measure the vibration amplitude. The speed corresponding to the peak vibration amplitude is recorded; this speed is the resonance point speed. For some indoor units, the fan blades may have multiple resonance points; the controller pre-stores all resonance point speed values ​​in an array in its memory.

[0028] The controller reads the resonance point speed list from the memory and determines whether the change from the current fan speed to the target fan speed will pass through any resonance point. Specifically, the logic is as follows: when the target fan speed is greater than the current fan speed, the controller checks whether the range formed by the current and target fan speeds contains any resonance point speed value; when the target fan speed is less than the current fan speed, it similarly checks whether the decreasing range contains any resonance point speed. For example, if the current fan speed is 400 revolutions per minute (RPM), the target fan speed is 800 RPM, and the resonance point speed is 650 RPM, since 650 RPM falls within the acceleration range of 400 to 800 RPM, it is determined that the speed change will cross a resonance point. Similarly, if the current fan speed is 700 RPM, the target fan speed is 500 RPM, and the resonance point speed is 650 RPM, since 650 RPM falls within the deceleration range of 500 to 700 RPM, it is also determined that the speed change will cross a resonance point. If the interval between the target fan speed and the current fan speed does not contain any resonance point speed value, it is determined that the resonance point is not crossed. In addition, if the speed change interval exactly contains the resonance point speed and the starting speed or ending speed of the speed change is equal to the resonance point speed, the controller also determines that the resonance point is crossed, because the fan will also experience a resonance state when starting from or stopping at the resonance point.

[0029] The controller stores the resonance crossing determination result in a buffer as a digital identifier, with 1 indicating crossing the resonance point and 0 indicating not crossing it. This determination result serves as one of the key input parameters for the auditory abruptness assessment model. This is because when the wind turbine crosses the resonance point, the amplitude of the casing vibration increases sharply, and the radiated noise sound pressure level rises significantly in a short period of time. This sudden increase in vibration noise can greatly aggravate the auditory abruptness perceived by the human ear, and the model must fully consider the influence of this factor when calculating the abruptness level.

[0030] S105: Input the rotational speed change rate, the resonance crossing determination result, the wind deflector motion state determination result, and the current operating mode into the pre-constructed auditory abruptness assessment model to obtain the auditory abruptness level; In this embodiment, please refer to 3 and Figure 4 The methods for constructing an auditory abruptness assessment model include: S301: Set multiple indoor unit test conditions with different combinations of operating parameters, collect indoor unit operating noise samples under each set of indoor unit test conditions, and form a noise sample set; S302: Obtain subjective perception scores from multiple groups of subjects for each indoor unit operating noise sample in the noise sample set, and form a subjective score set that corresponds one-to-one with each indoor unit operating noise sample. S303: Extract the operating parameter features corresponding to the test conditions of each group of indoor units to form an objective feature set that corresponds one-to-one with the operating noise samples of each indoor unit; S304: Match and associate the subjective rating set with the objective feature set, analyze the correlation between each operating parameter feature and the subjective perception rating, and filter out feature items; S305: Based on the aforementioned feature terms, a multivariate linear regression combined with a nonlinear correction term is used for fitting to obtain the auditory abruptness assessment model.

[0031] This embodiment collects noise samples under multiple test conditions with different combinations of operating parameters, comprehensively covering various speed adjustment ranges, air guide vane operation states, and operating modes of the air conditioner indoor unit, ensuring the scenario coverage and data representativeness of the sample set. By introducing subjective perception scores from multiple subjects to construct a corresponding score set, using real human auditory perception as the evaluation benchmark, it avoids the problem of relying solely on physical acoustic parameters and the disconnect between them and actual user hearing, making the model output results more consistent with human perception characteristics. By matching and associating objective operating parameter features with subjective score data, highly correlated feature items are selected, eliminating redundant parameters, simplifying the model input dimensions, and reducing the computational load on the subsequent whole-unit controller. Finally, a fitting method combining multiple linear regression with nonlinear correction terms is adopted. This retains the advantages of linear models—simple computation and easy implementation on the air conditioner main control chip—while adapting to the human ear's perception of sound changes through nonlinear correction, effectively improving the model's evaluation accuracy in different speed change ranges, providing accurate and reliable quantitative judgment basis for subsequent graded scheduling of masking sounds.

[0032] Specifically, an air conditioner indoor unit noise testing platform was constructed in a semi-anechoic chamber. The floor of the semi-anechoic chamber was a reflective surface, while the other five surfaces were covered with sound-absorbing wedges to simulate a near-free-field acoustic environment, eliminating interference from external environmental noise and reflected sound on the recording. The testing platform included one indoor unit of the air conditioner under test, a programmable fan drive control device, a multi-channel data acquisition and analysis instrument, and two free-field microphones. The indoor unit under test was fixed to a mounting bracket in the center of the semi-anechoic chamber according to product installation standards, at a height of 1.5 meters above the ground, simulating an actual wall-mounted installation. The two microphones were positioned 0.3 meters directly in front of the air outlet of the indoor unit and 0.5 meters to the side, respectively, to pick up radiated noise signals from different directions.

[0033] Multiple test conditions with different combinations of operating parameters were set up to cover various states that may occur during wind speed switching. Specific variables included the rate of change of fan speed, whether the fan crossed the resonance point, whether the air guide plate rotated, and the current operating mode. For the rate of change of fan speed, five gradient values ​​(e.g., 2%, 5%, 10%, 15%, 20%) were set, corresponding to the fan speed change per unit time. This was achieved by controlling the motor to change from different initial speeds to different target speeds using a programmable drive control device with a set acceleration. For the resonance crossing dimension, the resonance point speed of the indoor unit's fan blades was selected as the criterion. This resonance point speed value was obtained beforehand through frequency sweep testing. In the test design, the speed change range of some test conditions covered the resonance point speed, while the speed change range of others did not pass through the resonance point, thus forming two states: crossing and not crossing. For the air guide plate, two states were set: the air guide plate remained stationary during speed changes and oscillated synchronously during speed changes. The oscillation angle range was set to multiple oscillation amplitudes greater than five degrees. In terms of operating modes, there are four states: sleep mode, silent mode, normal mode and powerful mode. The basic speed range and air volume output characteristics of the fan are different in different modes.

[0034] The different values ​​of the above four dimensions are combined using all factors to form a complete test condition matrix. Each combination represents a test condition, resulting in multiple test condition sets. For each test condition set, a preset speed control sequence is executed via a programmable drive control device, simultaneously activating the data acquisition and analysis instrument to begin recording. During recording, two microphones synchronously acquire the sound pressure signal radiated by the indoor unit throughout the entire fan speed switching process. The sampling frequency is set to 48 kHz, and the quantization bit depth is set to 24 bits to ensure the frequency domain integrity and dynamic range of the recorded noise samples. The duration of each recording covers the entire time period from two seconds before speed change starts to three seconds after speed change completes, ensuring that all acoustic information is included, including steady-state noise before speed change, transient noise during speed change, and steady-state noise after speed change.

[0035] Each test condition was recorded multiple times to eliminate the influence of random factors on the noise samples. After recording all test conditions, the sound pressure signal of each recording was segmented along the time axis to remove any unstable trigger levels at the beginning and end, retaining complete acoustic data of the speed change process. The segmented sound pressure signals were saved in waveform audio format, and corresponding operating condition labeling information was created for each noise sample. This labeling information included the speed change rate gradient of the sample, whether it crossed a resonance point, whether the air guide plate rotated, and the operating mode. All saved noise samples and their corresponding operating condition labeling information together constitute a noise sample set. Each sample in this set has a complete operating condition label, which can be used for subsequent subjective evaluation tests and objective feature extraction.

[0036] After recording and organizing the noise sample set, a subjective hearing evaluation experiment was conducted to obtain subjective perception scores for each sample. Subjects were selected from candidates with normal hearing in a soundproof chamber. Candidates were required to pass a pure-tone audiometry test covering frequencies from 250 Hz to 8 kHz, with all frequency thresholds not exceeding 25 dB. A total of fifteen subjects were selected, with a balanced male-to-female ratio and an age distribution ranging from 22 to 45 years old, to simulate the auditory perception characteristics of different age groups.

[0037] The subjective evaluation test was conducted in a dedicated listening room. The background noise level in the listening room was below 20 decibels, and the reverberation time was controlled within 0.3 seconds, meeting the acoustic environment requirements for subjective evaluation tests of the human ear. Before the test, the evaluation criteria and scoring rules were explained to all participants. The scoring used a 10-point scale, with one point representing a completely smooth and natural sound change, and ten points representing an extremely harsh and unbearable noise abrupt change. Several sets of typical noise samples were played as examples to help participants develop an intuitive understanding of each level of the scoring scale. For example, a steady-state operating noise was played as a low-score reference, and a variable-speed noise that rapidly passed through the resonance point was played as a high-score reference.

[0038] During the formal experiment, each noise sample in the set was presented to each subject in a random order. Each playback used high-fidelity monitoring headphones or full-range monitoring speakers to reproduce a single noise sample. The playback sound pressure level was calibrated according to the actual operating sound level of the indoor unit to simulate the sound intensity in a real-world usage scenario. The playback duration of each sample was a complete speed-changing process, lasting from two seconds before the speed change to three seconds after it ended. After each sample was played, the subject independently rated its abruptness on an electronic scoring terminal. There was no time limit for the scoring process, but subjects were required not to communicate with other subjects before making their judgments. After scoring every ten samples, subjects were given a five-minute break to alleviate the impact of auditory fatigue on scoring accuracy. Furthermore, several duplicate samples were interspersed in the random sequence to test the consistency of the subjects' ratings. When a subject's rating of a duplicate sample deviated by more than two points, all of that subject's rating data was marked and reviewed.

[0039] After all samples were played and scored, the scores given by each subject for each noise sample were aggregated. For each noise sample, the arithmetic mean of the scores from fifteen subjects was calculated as the final subjective perception score for that sample. The standard deviation of the score for that sample was also calculated to assess the degree of consistency among subjects. If the standard deviation of a sample's score was too large, it indicated significant disagreement among subjects regarding their perception of that sample; this sample was then marked and a supplementary evaluation was arranged. Ultimately, each noise sample corresponded to a specific subjective perception score, which was linked to the operational labeling information of that noise sample. All noise samples and their corresponding subjective scores constituted a subjective score set, with each score in this set maintaining a one-to-one correspondence with a corresponding sample in the noise sample set, serving as the regression target variable for subsequently establishing an auditory abruptness assessment model.

[0040] After completing the noise sample recording and subjective scoring experiments, the operating parameter characteristics corresponding to each test condition were extracted and quantified. Each test condition had its complete parameter configuration recorded during the experimental design phase, including the specific value of the speed change rate, whether the speed range included the fan blade resonance point speed, whether the guide vane rotated during speed changes and its rotation amplitude, and the operating mode to which the condition belonged. These parameters were then extracted from the experimental record table and converted into numerical features according to a unified coding rule. For the speed change rate, the experimentally set change rate value was directly used; for resonance crossing, the crossing state was coded as 1, and the non-crossing state as 0; for the guide vane rotation state, the rotation state was coded as 1, and the stationary state as 0; for the operating mode, different mode coefficient values ​​were assigned according to sleep mode, silent mode, normal mode, and powerful mode. The four encoded feature values ​​were combined in a fixed order to form an objective feature vector. Each noise sample in the noise sample set corresponds to an objective feature vector. The objective feature vectors of all samples together constitute the objective feature set. Each feature vector in this set maintains a strict one-to-one correspondence with the corresponding sample and its subjective score in the noise sample set.

[0041] After constructing the objective feature set, the subjective rating set is matched and associated with the objective feature set to form a complete modeling dataset. Each record in this dataset contains a subjective rating value corresponding to a noise sample and four objective feature values ​​for that sample. Preliminary statistical analysis is performed on the dataset, calculating the Pearson correlation coefficient (a statistical indicator that measures the degree of linear correlation between two variables) between each objective feature and the subjective rating to quantify the strength of the linear association between each feature and the perceived auditory abruptness. For the rotational speed change rate, its correlation coefficient with the subjective rating is calculated to observe whether there is a positive correlation, i.e., whether a higher rotational speed change rate corresponds to a higher subjective rating. For the resonance crossing judgment results, the mean subjective ratings for both crossing and non-crossing samples are statistically analyzed to compare whether the difference in ratings between the two types of samples is significant. For the wind deflector rotation judgment results, the difference in the mean subjective ratings between rotating and stationary samples is also compared. For the operating mode coefficient, the analysis examines whether there are systematic differences in the subjective ratings corresponding to the same rotational speed change rate under different modes.

[0042] Based on the above correlation analysis results, stepwise regression was further used to screen the four feature items. The basic strategy of stepwise regression is to start with a blank model containing only constant terms, add each feature to the model one by one, calculate the increment of the model's goodness of fit after each addition, and select the feature that most significantly improves the goodness of fit to be included in the model first. At the same time, after each new feature is added, the significance of the existing features in the model is re-examined. When a feature becomes insignificant due to the introduction of other features, it is removed from the model. Through this iterative screening process, feature items that have an independent and significant contribution to the subjective score are retained, while feature items that are highly collinear with other features or have no significant contribution are removed. In this embodiment, after stepwise regression screening, all four feature items have significant explanatory power for the subjective score and are all retained. Among them, the rotation speed change rate has the largest contribution to the subjective score, and its proportion of score variation that can be explained alone is the highest; the resonance crossing judgment result is the second largest; the wind deflector rotation judgment result and the operating mode coefficient also show statistical significance. After screening, the above four feature items are determined as the set of independent variables for constructing the auditory abruptness assessment model for subsequent regression fitting analysis.

[0043] In this embodiment, please refer to Figure 5 Based on the feature terms, a multiple linear regression model combined with a nonlinear correction term is used for fitting, resulting in an auditory abruptness assessment model including: S501: Using the rotational speed change rate, resonance crossing judgment result, wind deflector motion state judgment result and current operating mode as linear independent variables, and the subjective perception score as the dependent variable, an initial regression model is established. S502: Calculate the prediction residual of the initial regression model, and determine the systematic bias of the initial regression model in the low-speed and high-speed change ranges based on the trend of the prediction residual with the rate of change of the rotational speed. S503: Apply a nonlinear transformation to the speed change rate based on the systematic deviation, and use the resonance crossing judgment result, the wind deflector motion state judgment result, the current operating mode, and the speed change rate after nonlinear transformation as updated independent variables, and perform regression fitting again to obtain an updated regression model; S504: Based on the fitting results of the updated regression model, determine the coefficient values ​​of each independent variable in the updated regression model and the level threshold range of the hearing abruptness level to obtain the hearing abruptness assessment model.

[0044] This embodiment uses various feature terms as linear independent variables and subjective perception scores as dependent variables to build an initial regression model. This quickly clarifies the basic correlation between various operating parameters and the perceived abruptness of hearing. The model structure is simple and clear, and the computational logic is convenient, facilitating rapid basic verification and parameter influence ranking. By calculating the prediction residuals of the initial model and analyzing their changing trend with the rate of change of rotational speed, the systematic bias of the linear model in the low-speed and high-speed change ranges can be accurately located, providing clear data support for subsequent model correction and avoiding unfounded structural adjustments. Applying a corresponding nonlinear transformation to the rate of change of rotational speed before re-implementing regression fitting effectively adapts to the perceptual saturation characteristics of human ears regarding sound changes, corrects the prediction bias of the linear model under different rotational speed change amplitudes, and significantly improves the fit between the model output results and the actual human hearing. The fitting process retains the linear form of parameters such as resonance crossing, air guide plate motion state, and operating mode, and only performs nonlinear adjustments on core influencing parameters. This effectively controls the overall computational load of the model while ensuring evaluation accuracy, making it easy to implement and run on embedded hardware platforms such as the main control chip of the air conditioner indoor unit. Finally, based on the fitting results, the coefficients of each independent variable and the threshold range of abruptness level are determined. This can transform the continuous abruptness calculation results into clear classification criteria, which can be directly connected to the control logic of subsequent masking sound scheduling and has good engineering application value.

[0045] Specifically, after completing the feature selection, the regression modeling stage begins. The rotational speed change rate, resonance crossing determination result, wind deflector motion state determination result, and current operating mode are used as linear independent variables, and the subjective perception score corresponding to each noise sample is used as the dependent variable to establish an initial regression model.

[0046] In practice, the four independent variable values ​​and corresponding subjective rating values ​​for each sample record are extracted from the dataset to form a modeling sample set. This sample set contains multiple complete sets of data, each consisting of four independent variable values ​​and one dependent variable value. The data is input into regression analysis software, and the regression equation is set to the form where the dependent variable equals a constant term plus the sum of the following: rotational speed change rate multiplied by the first coefficient, resonance crossing judgment result multiplied by the second coefficient, wind deflector motion state judgment result multiplied by the third coefficient, and current operating mode multiplied by the fourth coefficient. This equation assumes a simple additive relationship between the subjective abruptness and the four physical quantities; that is, the contributions of each factor to the abruptness are independent, and the total abruptness equals the linear sum of the contributions of each factor. The least squares method is used to estimate the parameters of the regression equation. The basic principle of the least squares method is to find a set of coefficient values ​​that minimizes the sum of squares of the differences between the predicted scores and the actual subjective scores of all samples. The regression analysis software solves the normal equations through matrix operations to obtain the estimated values ​​of each coefficient that minimize the sum of squared residuals.

[0047] After coefficient estimation, an initial regression model expression containing specific coefficient values ​​is obtained. This model can calculate the corresponding abruptness prediction score for any given set of independent variable values. Further calculations are made of the residuals between the predicted scores and subjective scores of this initial regression model across all samples, as well as the goodness-of-fit index, to assess the model's overall explanatory power for the data. Once the initial regression model is established, its predicted residual data and goodness-of-fit index will serve as inputs for subsequent identification of systematic biases and nonlinear corrections.

[0048] After estimating the parameters of the initial regression model using the least squares method and obtaining the specific values ​​of each coefficient, the modeling process enters the model diagnosis and residual analysis stage. The values ​​of the four independent variables for each sample record in the modeling sample set are substituted into the initial regression model to calculate the predicted score for that sample. Then, the predicted score is subtracted from the actual subjective perception score of the sample to obtain the predicted residual. A positive residual indicates that the model underestimates the abruptness of the sample, and the actual abruptness is higher than the model estimate; a negative residual indicates that the model overestimates the abruptness of the sample, and the actual abruptness is lower than the model estimate. All samples are sorted in ascending order of rotational speed change rate, and a residual distribution graph is plotted with the rotational speed change rate on the x-axis and the predicted residual on the y-axis.

[0049] Observe the overall shape and changing pattern of the residual distribution plot. When the rotational speed change rate is in the lower range, a large number of samples show negative residuals, meaning the model's predicted value is higher than the actual subjective score, indicating that the abruptness of the low-speed change range is overestimated within the framework of the linear model. When the rotational speed change rate is in the higher range, a large number of samples show positive residuals, meaning the model's predicted value is lower than the actual subjective score, indicating that the abruptness of the high-speed change range is underestimated. This phenomenon shows that the initial regression model has a systematic prediction bias in different rotational speed change rate ranges, and the direction of the bias changes systematically from negative to positive as the rotational speed change rate increases.

[0050] Further quantitative analysis was conducted on the correlation between residuals and the rate of change of rotational speed. The correlation coefficient between residuals and the rate of change of rotational speed was calculated to observe whether a significant positive correlation existed. If the correlation coefficient was significantly positive, it statistically confirmed that the residuals increased with the rate of change of rotational speed. Simultaneously, the rate of change of rotational speed was divided into three intervals: low, medium, and high. The average residuals of the samples in each interval were calculated, and the differences in the average residuals between the intervals were compared to see if they were significant. The average residuals in the low-speed interval were significantly negative, while those in the high-speed interval were significantly positive, further confirming a systematic bias in the initial regression model across different rate of change intervals. The root of this bias lies in the linear model's assumption that the perceived abruptness has a simple linear relationship with the rate of change of rotational speed. However, actual human auditory perception follows a non-linear law. That is, at low speeds, the human ear is extremely sensitive to differences in rotational speed; even a small increment in the rate of change can cause a significant increase in perceived abruptness. At high speeds, this sensitivity gradually decreases, requiring a larger increment in the rate of change to cause the same increase in perceived abruptness. The linear model cannot characterize this saturation characteristic, leading to a systematic bias of overestimation at low speeds and underestimation at high speeds. Based on the above residual analysis conclusions, the direction and characteristics of this systematic bias will be used as the basis for subsequently introducing nonlinear corrections to optimize model accuracy.

[0051] After completing the residual analysis and identifying systematic biases in the initial regression model across low-speed and high-speed ranges, the modeling process enters the nonlinear correction phase. The rotational speed change rate for each sample record in the modeling sample set is extracted, and a nonlinear transformation is applied to this value. Based on the patterns revealed by the residual analysis, the initial regression model overestimates the low-speed range and underestimates the high-speed range, indicating that the human ear's perception of the rotational speed change rate follows a saturation characteristic: sensitivity is high at low rates of change and gradually decreases at high rates of change. A nonlinear function with saturation characteristics is selected to transform the rotational speed change rate. This nonlinear function is characterized by a larger derivative in the low-speed range and a gradual decrease in the high-speed range, resulting in a higher growth rate of the transformed value in the low-speed range compared to the high-speed range. During the transformation, the rotational speed change rate of each sample is substituted into the nonlinear function to calculate the transformed value, which is then used to replace the linear term of the rotational speed change rate in the original linear model. For the resonance crossing judgment results, the wind deflector motion state judgment results, and the current operating mode, keep their original coding forms unchanged and do not apply any changes, because these variables are categorical variables, and their relationship with the abruptness can be directly represented by the regression coefficients, without the need for nonlinear processing.

[0052] After the nonlinear transformation, the transformed rotational speed change rate, along with the original resonance crossing judgment result, the wind deflector motion state judgment result, and the current operating mode, are used as the updated set of independent variables. The subjective perception score corresponding to each sample is used as the dependent variable to re-establish the regression equation. The updated regression equation is in the form of a constant term plus the sum of the nonlinearly transformed rotational speed change rate multiplied by the first update coefficient, the resonance crossing judgment result multiplied by the second update coefficient, the wind deflector motion state judgment result multiplied by the third update coefficient, and the current operating mode multiplied by the fourth update coefficient. The least squares method is used to estimate the parameters of this updated regression equation again, finding a new set of coefficient values ​​that minimizes the sum of squares of the differences between the predicted scores and the actual subjective scores for all samples. The regression analysis software uses matrix operations to obtain the estimated values ​​of each coefficient that minimize the sum of squared residuals.

[0053] After parameter estimation, the updated regression model expression containing each updated coefficient is obtained. The residuals between the predicted scores and subjective scores of the updated regression model on all samples are calculated, and the goodness-of-fit index of the updated model is calculated. This goodness-of-fit index is compared with the goodness-of-fit index of the initial regression model to verify whether the nonlinear correction effectively improves the model's interpretability. When the goodness-of-fit index of the updated model is significantly higher than that of the initial model, it confirms that the introduction of nonlinear transformation significantly improves the model's prediction accuracy for subjective scores. Simultaneously, the residual distribution plot of the updated model is plotted to observe whether the residuals exhibit a systematic bias trend with increasing rotational speed. If the residuals are uniformly distributed near the zero line and no longer show a regular pattern of negative bias in the low-speed range and positive bias in the high-speed range, it indicates that the introduced nonlinear transformation has successfully eliminated the systematic bias in the initial model. After the updated regression model is established, its fitting results and residual analysis conclusions will serve as the input basis for subsequently determining the specific values ​​of each coefficient and classifying the threshold intervals for hearing abruptness levels.

[0054] After estimating the parameters of the updated regression model and obtaining the specific values ​​of each updated coefficient, the modeling process enters the coefficient determination and threshold division stage. The parameter estimation results output by the regression analysis software include the coefficient values ​​of each independent variable in the updated regression model and their corresponding statistical significance test results. The constant term and the coefficient values ​​of each of the four independent variables are extracted separately. The nonlinearly transformed rotational speed change rate corresponds to the first coefficient, the resonance crossing judgment result corresponds to the second coefficient, the wind deflector motion state judgment result corresponds to the third coefficient, and the current operating mode corresponds to the fourth coefficient. At the same time, the standard error and confidence interval of each coefficient are recorded to assess the reliability of the coefficient estimates. When the statistical significance test value of each coefficient is less than the preset significance level, it is confirmed that all independent variables have significant independent contributions to the dependent variable, and the estimated coefficient values ​​are statistically significant. Substituting the above coefficient values ​​into the regression equation, a formally defined auditory abruptness assessment model expression is obtained. This expression takes the nonlinearly transformed rotational speed change rate, the resonance crossing judgment result, the wind deflector motion state judgment result, and the current operating mode as inputs and outputs the corresponding abruptness prediction score.

[0055] After the model form is determined, the threshold intervals for different levels are defined. The predicted abruptness scores of all samples in the modeling sample set are statistically analyzed, and the range, mean, and standard deviation of the predicted scores are calculated. Based on the overall distribution of the predicted scores and the actual perceptual differences of subjects to different scores in the subjective rating experiment, the predicted abruptness scores are divided into multiple level intervals. The specific division principle is as follows: when the predicted score is in a low interval, the auditory abruptness caused by the corresponding combination of physical parameter changes is slight and almost imperceptible to the human ear, and is classified as low abruptness; when the predicted score is in a medium interval, the corresponding combination of physical parameter changes will cause a certain degree of auditory discomfort, and is classified as medium abruptness; when the predicted score is in a high interval, the corresponding combination of physical parameter changes will cause obvious auditory abruptness and irritability, and is classified as high abruptness. The boundary value between the low-level interval and the medium-level interval is defined as the first preset threshold, and the boundary value between the medium-level interval and the high-level interval is defined as the second preset threshold. The first preset threshold is less than the second preset threshold. The threshold values ​​for each level range and the coefficient values ​​in the model expression are recorded together to form a complete auditory abruptness assessment model.

[0056] In this embodiment, the calculation formula for the auditory abruptness assessment model is as follows: Where J represents the abruptness level, R represents the rate of change of rotational speed; R represents whether the resonance point is crossed, with R being 1 when it is crossed and 0 when it is not; S represents whether the air guide plate rotates, with S being 1 when it rotates and 0 when it does not; M represents the operating mode, where M is 0.5 for sleep mode, 0.8 for silent mode, 1.0 for normal mode, and 1.2 for powerful mode.

[0057] In this embodiment, please refer to Figure 6 The rotational speed change rate, resonance crossing determination result, wind deflector motion state determination result, and current operating mode are input into a pre-built auditory abruptness assessment model to obtain auditory abruptness levels, including: S601: Input the rotational speed change rate, the resonance crossing determination result, the wind deflector motion state determination result, and the current operating mode into the pre-constructed auditory abruptness evaluation model to obtain the auditory abruptness; S602: Determine whether the auditory abruptness is less than a first preset threshold; S603: If the auditory abruptness is less than the first preset threshold, it is determined to be a low abruptness level; S604: If the auditory abruptness is greater than or equal to the first preset threshold, determine whether the auditory abruptness is less than or equal to the second preset threshold; Wherein, the second preset threshold is greater than the first preset threshold; S605: If the auditory abruptness is less than or equal to the second preset threshold, it is determined to be of medium abruptness level; S606: If the auditory abruptness is greater than the second preset threshold, it is determined to be a high abruptness level.

[0058] This embodiment outputs continuous quantitative values ​​of abruptness through a pre-built evaluation model, and then completes a progressive three-level classification based on two preset thresholds. The overall judgment logic is clear and well-organized, and the calculation steps are simple and controllable. It can adapt to the limited computing resources of the air conditioner indoor unit's main control chip, ensuring real-time response efficiency in wind speed switching scenarios. The progressive threshold comparison process first completes the rapid judgment of low abruptness level with the first preset threshold, and then distinguishes medium and high abruptness levels with a higher second preset threshold. The judgment path is clear and orderly, and the judgment result of each level can directly correspond to the subsequent differentiated masking sound scheduling scheme, which facilitates the modular implementation of control logic and subsequent parameter debugging and optimization. The granularity of the three-level classification can accurately match the masking requirements of different abruptness levels. In low abruptness scenarios, only the minimum masking resources are used to avoid additional sound output from disrupting the quietness of the environment. In high abruptness scenarios, full-dimensional masking configuration is used to ensure a smooth auditory transition, achieving a reasonable balance between masking effect and sound environment comfort. Meanwhile, both preset thresholds can be flexibly adjusted according to the air duct structure and overall noise characteristics of different models, and can also adapt to the differences in hearing sensitivity of different user groups, with strong product adaptability and scenario expansion.

[0059] Specifically, after the current operating parameters are input into the pre-built auditory abruptness assessment model, the model calculates and outputs a specific auditory abruptness value. This value reflects the quantified intensity of the abruptness that the human ear may perceive during the current wind speed change. The controller reads this value from the model's output register and records it as the auditory abruptness value. The dimensions of this value are consistent with the scale of the subjective scores during model building, and the value range corresponds to the distribution range of the subjective scores during modeling.

[0060] The controller reads a pre-stored first preset threshold from its internal memory. This first preset threshold is determined during the model building phase by analyzing the predicted score distribution of low-abruptness samples and medium-abruptness samples in the modeling sample set, representing the boundary value between low-abruptness level and medium-high level. The controller compares the auditory abruptness value output by the model with the first preset threshold. When the auditory abruptness value is less than the first preset threshold, the controller determines that the auditory abruptness of the current speed change process is slight, almost imperceptible to the human ear, or only produces a very slight perceptual change, and marks this state as low-abruptness level.

[0061] When the auditory abruptness value is greater than or equal to the first preset threshold, the controller proceeds to the second level of judgment. The controller reads the second preset threshold from memory. This second preset threshold is greater than the first preset threshold and is determined during the model building phase by analyzing the predicted score distribution of medium-abruptness samples and high-abruptness samples, representing the boundary value between medium and high levels. The controller compares the auditory abruptness value with the second preset threshold. When the auditory abruptness value is less than or equal to the second preset threshold, the controller determines that the abruptness of the current speed change is at a medium level; the human ear can clearly perceive the noise change but it has not yet reached an unbearable level, and this state is marked as medium abruptness. When the auditory abruptness value is greater than the second preset threshold, the controller determines that the abruptness of the current speed change is strong, the noise change is significant and easily causes the user's irritability, and this state is marked as high abruptness.

[0062] In one specific embodiment of this application, based on the correspondence between the numerical range distribution of the model output and the subjective rating, a first preset threshold is set to forty, and a second preset threshold is set to sixty-five. The controller records the auditory abruptness value of the model output as J. When the controller determines that the value of J is less than forty, the abruptness level of the current gear shift process is determined to be low abruptness. When the controller determines that the value of J is greater than or equal to forty, it continues to compare J with sixty-five. If J is less than or equal to sixty-five, it is determined to be medium abruptness; if J is greater than sixty-five, it is determined to be high abruptness.

[0063] After completing the level determination, the controller stores the determination result in the form of a level code in a designated address in the buffer. This level code is used to subsequently determine the frequency band configuration scheme for the masking sound: a high-absence level corresponds to three-segment coordinated masking of the high-frequency, mid-frequency, and low-frequency bands; a mid-absence level corresponds to two-segment coordinated masking of the high-frequency and mid-frequency bands; and a low-absence level corresponds to a single-segment masking of the low-frequency band. The controller's subsequent processes directly read this level code from the buffer and call the corresponding masking sound generation parameters accordingly, without repeating the above comparison and judgment logic. In addition, the first and second preset thresholds used by the controller in the determination process are stored in non-volatile memory and can be configured differently according to different product models and user scenarios. For example, for different indoor unit models with wind noise characteristics and blade resonance characteristics, the thresholds can be adjusted to adapt to the personalized auditory perception characteristics of each model.

[0064] S106: Determine the frequency band configuration scheme of the masking sound according to the auditory abruptness level, generate the masking sound according to the frequency band configuration scheme, and play the masking sound before the fan speed switching command is executed.

[0065] In this embodiment, determining the frequency band configuration scheme of the masking sound based on the auditory abruptness level includes: When the hearing abruptness level is high abruptness, the frequency band configuration scheme is determined to be to simultaneously output high-frequency masking sound, mid-frequency masking sound and low-frequency masking sound; When the hearing abruptness level is medium abruptness, the frequency band configuration scheme is determined to be to output high-frequency masking sound and mid-frequency masking sound simultaneously; When the auditory abruptness level is low, the frequency band configuration scheme is determined to output high-frequency masking sound.

[0066] This embodiment directly links the auditory abruptness level with the activation of multi-band masking sounds in a three-tiered manner, achieving precise matching and efficient utilization of masking resources. When the level is high abrupt, high, mid, and low frequency bands are activated simultaneously, ensuring a sufficiently wide spectral coverage of the masking sound in the most dramatic noise change scenarios, providing comprehensive auditory masking for broadband noise changes that may occur during speed changes. When the level is medium abrupt, high and mid frequency bands are activated. At this point, the intensity of the noise change is reduced, and discarding the low-frequency masking sound effectively covers the mid-to-high frequency wind noise components that the human ear is most sensitive to, while avoiding unnecessary redundant acoustic energy output. When the level is low abrupt, only the high-frequency band is activated. Because the noise change is slight, only moderate masking of the most prominent and sensitive high-frequency components of wind noise is needed to meet the requirements for a smooth auditory transition, avoiding secondary auditory discomfort that may be introduced by excessive masking. This also helps reduce speaker power consumption and auditory fatigue caused by prolonged playback of masking sounds. This three-level progressive strategy ensures that the bandwidth and energy distribution of the masking sound always match the actual auditory abruptness, ensuring sufficient masking in necessary scenarios and minimizing masking in unnecessary scenarios, thus achieving a dynamic balance between masking effect and auditory comfort.

[0067] Specifically, after determining the auditory abruptness level and storing the level code in the buffer, the controller determines the frequency band configuration scheme for the masking sound based on the level code. The controller reads the level code from the buffer. When the code indicates a high abruptness level, the controller determines the frequency band configuration scheme to simultaneously output high-frequency, mid-frequency, and low-frequency masking sounds. These three frequency bands work together to cope with the most severe noise abrupt changes. When the code indicates a medium abruptness level, the controller determines the frequency band configuration scheme to simultaneously output high-frequency and mid-frequency masking sounds. In this case, the severity of the noise abrupt change is reduced, and the low-frequency masking sound does not need to be activated. When the code indicates a low abruptness level, the controller determines the frequency band configuration scheme to output only high-frequency masking sounds. In this case, the noise change is slight, and moderate masking with a single high-frequency band is sufficient to meet the requirements for a smooth auditory transition.

[0068] While determining the frequency band configuration scheme, the controller reads the masking sound signal type and spectral parameters corresponding to each frequency band. The high-frequency masking sound band is 1kHz-2kHz, corresponding to the region where the human ear is most sensitive to the high-frequency components of wind noise. Pink noise is used as the signal source for the high-frequency masking sound. The energy of pink noise gradually decreases as the frequency increases, and its energy distribution remains equal across all octaves, resulting in a uniform and gentle sound. This simulates the acoustic characteristics of rain or rustling leaves in nature, effectively masking the high-frequency turbulence noise generated when airflow passes through a wind tunnel at high speed. Simultaneously, its gentle spectral characteristics avoid causing harshness to the human ear. The mid-frequency masking sound band is 500Hz-1kHz, covering the prominent mid-frequency components of wind noise. White noise is used as the signal source for the mid-frequency masking sound. White noise has a uniform energy distribution across the entire frequency band, effectively masking mid-frequency noise. The low-frequency masking sound ranges from 20Hz to 500Hz. This range corresponds to the low-frequency components generated by wind tunnel resonance and airflow pulsation. The low-frequency masking sound is generated by superimposing a low-frequency modulated signal with white noise. The low-frequency modulated signal provides periodic amplitude fluctuations, while the white noise provides broadband background filling. The superposition of the two creates an acoustic effect that simulates the sound of ocean waves. This composite signal retains the energy concentration characteristics required for low-frequency masking and introduces natural dynamic changes through modulation, avoiding the dullness and oppression that may result from continuous playback of pure low-frequency signals.

[0069] According to the determined frequency band configuration scheme, the controller activates the masking sound signal generation channels for the corresponding frequency bands. For high-frequency channels, the controller reads the digital signal sample of pink noise from memory and sends it to the digital-to-analog converter (DAC); for mid-frequency channels, the controller sends the digital signal sample of white noise to the corresponding DAC; for low-frequency channels, the controller first generates a low-frequency modulation signal, multiplies and superimposes it with the white noise signal in the time domain, and then sends the superimposed digital signal to the DAC. After DAC conversion, the signals from each channel are amplified and played by speakers arranged on the indoor unit casing. When the controller determines to use only a single frequency band, it activates only the corresponding channel and shuts down the other channels; when it determines to use two or three frequency bands, the activated channels work in parallel, and their respective output signals are naturally superimposed in the sound field to form a multi-frequency band collaborative masking sound field. The controller stores the determined frequency band configuration scheme along with the signal generation parameters corresponding to each frequency band in a buffer for subsequent masking sound signal generation and playback processes.

[0070] In some embodiments, the method further includes: when the determined frequency band configuration scheme contains low-frequency masking sound, acquiring the original noise signal of the indoor unit currently in operation, and extracting the real-time phase data corresponding to the low-frequency band from the original noise signal.

[0071] Based on the extracted real-time phase data, the phase parameters of the low-frequency masking sound are adjusted to generate a low-frequency masking sound signal that has a preset phase relationship with the low-frequency component of the original noise.

[0072] The generated low-frequency masking sound signal is combined with the masking sound signals of the other frequency bands in the frequency band configuration scheme to obtain the final masking sound signal used for playback.

[0073] Specifically, when the frequency band configuration scheme determined by the controller based on the abruptness level assessment includes low-frequency masking noise (i.e., a high abruptness level), the controller first executes the original noise acquisition and phase analysis process before initiating the masking noise generation process. The controller uses a microphone mounted on the indoor unit casing to acquire the ambient noise signal during the current operation of the indoor unit. This signal includes all acoustic components, such as aerodynamic noise generated by the fan rotation, turbulence noise within the duct, and any potential resonance noise. The microphone converts the acquired sound pressure signal into an analog electrical signal, which is then converted into a digital audio signal by an analog-to-digital converter and sent to the controller's digital signal processing unit.

[0074] After receiving the raw noise digital signal from the microphone, the digital signal processing unit performs spectral analysis on the signal. First, the processing unit converts the time-domain digital signal to the frequency domain using a Fast Fourier Transform (FFT) to obtain the full-band spectral distribution of the current ambient noise. Then, the processing unit extracts a low-frequency range from 20 Hz to 500 Hz from the full-band spectrum, extracting the phase data corresponding to each frequency component within this range. The specific operation for extracting the phase data is as follows: for each low-frequency point in the FFT result, the ratio of its real part to its imaginary part is calculated, and the arctangent function is taken to obtain the phase angle corresponding to that frequency point. The phase angles of all low-frequency points together constitute the real-time phase data set corresponding to the low-frequency band.

[0075] After phase extraction, the controller acquires a pre-stored low-frequency masking acoustic substrate signal. This substrate signal is a low-frequency modulated white noise signal without any phase adjustment. The controller selects the phase angles of several low-frequency points with the most concentrated energy from the real-time phase data set and calculates their average value as a low-frequency phase reference value. Based on this phase reference value, the controller applies a uniform phase shift to each frequency component of the low-frequency masking acoustic substrate signal, so that the phase of each frequency component in the final output low-frequency masking acoustic signal presents a preset phase relationship with the phase of the corresponding frequency component of the original noise. In this embodiment, the preset phase relationship is an anti-phase relationship, that is, the phase of the low-frequency masking acoustic signal is shifted by 180 degrees relative to the phase of the original noise low-frequency component, so that when the two meet on the propagation path, partial phase cancellation occurs, thereby enhancing the subjective masking effect of low-frequency noise without increasing additional acoustic energy.

[0076] After phase adjustment, the controller sends the phase-adjusted low-frequency masking sound signal and the masking sound signals of other frequency bands determined in the frequency band configuration scheme to the synthesizer for mixing. When the abruptness level is high, the other frequency bands include mid-frequency and high-frequency masking sounds; when the abruptness level is medium, the low-frequency masking sound is not enabled, so there is no need to perform the phase acquisition and adjustment process; when the abruptness level is low, low-frequency masking sounds are also not involved. The synthesizer superimposes the digital audio signals of each frequency band in the time domain according to their respective preset sound pressure level weights, merging them into a multi-band composite digital masking sound signal. During the synthesis process, the time reference of each frequency band signal remains consistent to ensure that the phase-adjusted low-frequency masking sound is precisely synchronized with the mid-frequency and high-frequency masking sounds in time, avoiding the introduction of new phase interference due to time alignment errors. The synthesized masking sound signal is sent to the digital-to-analog converter and power amplifier to drive the speaker for playback. The controller continuously monitors the phase change of the original noise through the microphone throughout the entire speed change process. When the phase change exceeds the preset threshold, it updates the phase offset of the low-frequency masking sound in real time to ensure that the phase relationship between the low-frequency masking sound and the original noise remains in the preset out-of-phase state throughout the entire speed change process.

[0077] In this embodiment, after playing the masking sound, the following steps are included: The moment when the current fan speed reaches the target fan speed and the fluctuation range is less than the preset speed fluctuation threshold is recorded as the stable start moment; Obtain the first intensity value of the masking sound corresponding to the stable start time, and generate an intensity attenuation control curve based on the first intensity value and the preset attenuation time; Starting from the stable initial moment, the real-time output intensity at each time point is calculated one by one according to the attenuation coefficient corresponding to each time point in the intensity attenuation control curve. Based on the real-time output intensity at each time point, the masking sound is played at the corresponding time point until the attenuation coefficient decreases to zero and then the masking sound is stopped.

[0078] This embodiment uses the moment when the fan speed reaches the target value and the fluctuation amplitude is less than a preset threshold as the attenuation start point. This ensures that the masking sound completely covers the entire process of speed switching, including the small fluctuation stage after the speed reaches the target value. This avoids the exposure of fluctuation noise due to premature attenuation of the masking sound, ensuring a smooth auditory transition throughout the wind speed switching process. An attenuation control curve is generated based on the actual masking sound intensity at a stable starting moment, rather than using a fixed initial attenuation intensity. This adapts to the differentiated masking output intensity corresponding to different abruptness levels, ensuring a seamless connection between the initial state of the attenuation process and the preceding masking stage, without any intensity jumps. The real-time output intensity is calculated point-by-point according to the attenuation curve, and playback is adjusted synchronously. This achieves continuous and smooth changes in masking sound intensity, avoiding the auditory step-like feeling caused by stepped intensity adjustments. Finally, the attenuation coefficient drops to zero as the criterion for stopping playback, forming a natural fade-out effect and avoiding secondary auditory abruptness caused by sudden cutoff of the masking sound. The entire control logic is clearly hierarchical, and parameters such as decay time and speed fluctuation threshold can be flexibly adjusted according to the characteristics of the model and the needs of the scenario. The amount of computation is controllable, making it easy to implement in the main control unit of the air conditioner indoor unit. While ensuring the effectiveness of shielding, it also takes into account the restoration of the quietness of the steady-state environment.

[0079] Specifically, after the fan speed adjustment process is completed, the controller monitors the fluctuation of the current fan speed in real time. The controller continuously reads the instantaneous speed value from the Hall sensor or speed feedback signal of the motor drive circuit, compares each instantaneous speed value with the target fan speed, and calculates the absolute value of the deviation between the two. At the same time, the controller compares the instantaneous speed values ​​of multiple adjacent samples to calculate the fluctuation range of the speed over time. The controller reads the preset speed fluctuation threshold from the internal memory. This threshold is calibrated and determined before the air conditioner leaves the factory and represents the maximum speed fluctuation range allowed for the fan to reach a stable operating state. The controller determines whether the current fan speed meets two conditions: the absolute value of the deviation between the instantaneous speed value and the target fan speed is less than the preset fluctuation threshold, and the speed fluctuation range remains within the preset fluctuation threshold for multiple consecutive sampling periods. When both conditions are met, the controller records the current moment as the stable start moment and stores the timestamp of this moment in the buffer.

[0080] After determining the stable start time, the controller executes the masking sound intensity attenuation control procedure. The controller reads the masking sound playback intensity value corresponding to the stable start time from the buffer and records this value as the first intensity value. This first intensity value reflects the real-time sound pressure level output by the speaker when the fan speed has just reached stability. Simultaneously, the controller reads the preset attenuation duration from memory. This attenuation duration is preset to five seconds. This duration ensures that the gradual disappearance of the masking sound is not noticeably perceptible to the human ear, while also avoiding excessively prolonged attenuation that could cause auditory interference from residual masking sound. Based on the first intensity value and the preset attenuation duration, the controller... A decay control curve is generated, where A(t) is the instantaneous intensity of the controlled sound at time t, A is the initial maximum intensity at the start of the fade-out, and t is the time elapsed since the start of the fade-out. This decay control curve uses time as the x-axis and the decay coefficient as the y-axis. The decay coefficient is initially one at a stable starting point and gradually decreases to zero over time according to an exponential function with the natural constant e as the base. When the time reaches the end of the preset decay duration, the decay coefficient exactly decreases to zero.

[0081] After the intensity attenuation control curve is generated, the controller enters the point-by-point real-time control phase. Using the system clock as a reference, the controller divides the preset attenuation duration into multiple equally spaced time control points. The interval between adjacent control points is determined by the controller's audio output refresh rate. For each time control point, the controller reads the attenuation coefficient corresponding to that time point from the intensity attenuation control curve, multiplies this coefficient by the first intensity value at the stable starting moment, and obtains the real-time output intensity value for that time point. The controller sequentially calculates the real-time output intensity at each time control point and updates the speaker's output power at the corresponding time point based on the calculation results, causing the masking sound's playback intensity to decrease point by point according to the calculated value. In the initial stage of the attenuation process, the attenuation coefficient is close to one, the real-time output intensity is basically consistent with the first intensity value, and the loudness of the masking sound does not change significantly. As time progresses, the attenuation coefficient gradually decreases, the real-time output intensity decreases synchronously, and the masking sound gradually becomes weaker. When the preset attenuation duration is reached, the attenuation coefficient becomes zero, the real-time output intensity is zero, the controller shuts off the speaker's signal output channel, and stops playing the masking sound. Throughout the attenuation process, the intensity of the masking sound changes continuously along a smooth curve of an exponential function, without any jumps or abrupt changes. The listener can hardly perceive that the masking sound is gradually disappearing, thus achieving a seamless integration of the masking sound exit process with the background acoustic environment.

[0082] Please see Figure 7 This embodiment provides a sound masking control device 700, including: The acquisition module 701 is used to acquire the current operating parameters and fan speed switching command of the indoor unit, and determine the target fan speed and target air guide plate angle according to the fan speed switching command; the current operating parameters include the current fan speed, the current air guide plate angle and the current operating mode; The determining module 702 is used to determine the rate of change of rotational speed based on the current fan speed and the target fan speed; Comparison module 703 is used to determine the change in the angle of the air guide plate based on the current angle of the air guide plate and the angle of the target air guide plate, and compare the change in the angle of the air guide plate with a preset angle threshold to obtain the motion state determination result of the air guide plate; The judgment module 704 is used to obtain the speed of the resonant point of the fan blade of the indoor unit, determine whether the current fan speed changes to the target fan speed and passes through the speed of the fan blade resonant point, and obtain the resonance crossing judgment result. The input module 705 is used to input the rotational speed change rate, the resonance crossing determination result, the wind deflector motion state determination result and the current operating mode into the pre-constructed auditory abruptness assessment model to obtain the auditory abruptness level; The generation module 706 is used to determine the frequency band configuration scheme of the masking sound according to the auditory abruptness level, generate the masking sound according to the frequency band configuration scheme, and play the masking sound before the fan speed switching command is executed.

[0083] Furthermore, the method for constructing the auditory abruptness assessment model includes: Set up multiple indoor unit test conditions with different combinations of operating parameters, collect indoor unit operating noise samples under each set of indoor unit test conditions, and form a noise sample set; Obtain subjective perception scores from multiple groups of subjects for each indoor unit operating noise sample in the noise sample set, forming a subjective score set that corresponds one-to-one with each indoor unit operating noise sample. Extract the operating parameter features corresponding to the test conditions of each group of indoor units to form an objective feature set that corresponds one-to-one with the operating noise sample of each indoor unit. The subjective rating set is matched and associated with the objective feature set to analyze the correlation between each operating parameter feature and the subjective perception rating, and feature items are selected. Based on the aforementioned features, a multivariate linear regression model combined with a nonlinear correction term is used for fitting to obtain the auditory abruptness assessment model.

[0084] Furthermore, the auditory abruptness assessment model, obtained by fitting the model based on the aforementioned feature terms using multiple linear regression combined with a nonlinear correction term, includes: Using the rotational speed change rate, resonance crossing judgment result, wind deflector motion state judgment result, and current operating mode as linear independent variables, and the subjective perception score as the dependent variable, an initial regression model is established. Calculate the prediction residual of the initial regression model, and determine the systematic bias of the initial regression model in the low-speed and high-speed change ranges based on the trend of the prediction residual with the rate of change of the rotational speed. Based on the systematic deviation, a nonlinear transformation is applied to the rotational speed change rate, and the resonance crossing judgment result, the wind deflector motion state judgment result, the current operating mode, and the rotational speed change rate after nonlinear transformation are used as updated independent variables. Regression fitting is then performed again to obtain an updated regression model. Based on the fitting results of the updated regression model, the coefficient values ​​of each independent variable in the updated regression model and the level threshold range of the hearing abruptness level are determined to obtain the hearing abruptness assessment model.

[0085] Furthermore, the input module 705 includes: The abruptness acquisition unit is used to input the rotation speed change rate, the resonance crossing determination result, the wind deflector motion state determination result and the current operating mode into the pre-constructed auditory abruptness evaluation model to obtain the auditory abruptness. The first judgment unit is used to determine whether the auditory abruptness is less than a first preset threshold. The low abruptness determination unit is used to determine the auditory abruptness level as low if the auditory abruptness is less than the first preset threshold. The second determination unit is used to determine whether the auditory abruptness is less than or equal to the second preset threshold if the auditory abruptness is greater than or equal to the first preset threshold; wherein the second preset threshold is greater than the first preset threshold. The abruptness determination unit is used to determine the auditory abruptness level as medium if the auditory abruptness is less than or equal to the second preset threshold. The high abruptness determination unit is used to determine the auditory abruptness level as high if the auditory abruptness is greater than the second preset threshold.

[0086] Furthermore, the generation module 706 includes: The first determining unit is used to determine the frequency band configuration scheme as simultaneously outputting high-frequency masking sound, mid-frequency masking sound and low-frequency masking sound when the auditory abruptness level is high abruptness level. The second determining unit is used to determine that when the auditory abruptness level is medium abruptness level, the frequency band configuration scheme is to simultaneously output high-frequency masking sound and mid-frequency masking sound; The third determining unit is used to determine that the frequency band configuration scheme is to output high-frequency masking sound when the auditory abruptness level is low abruptness level.

[0087] Furthermore, the frequency band of the high-frequency masking sound is 1kHz-2kHz; the frequency band of the mid-frequency masking sound is 500Hz-1kHz; and the frequency band of the low-frequency masking sound is 20Hz-500Hz.

[0088] Furthermore, the generation module 706 also includes: The time acquisition unit is used to acquire the time when the current wind turbine speed reaches the target wind turbine speed and the fluctuation range is less than a preset speed fluctuation threshold, and to record the time as the stable start time. The curve generation unit is used to obtain the first intensity value of the masking sound corresponding to the stable start time, and generate an intensity attenuation control curve based on the first intensity value and a preset attenuation time. The intensity calculation unit is used to calculate the real-time output intensity of each time point according to the attenuation coefficient corresponding to each time point in the intensity attenuation control curve, starting from the stable start time. The playback stop unit is used to control the playback of the masking sound at each of the corresponding time points according to the real-time output intensity at each of the time points, until the attenuation coefficient decreases to zero and the playback of the masking sound stops.

[0089] Specific limitations regarding the sound masking control device can be found in the limitations of the sound masking control method described above, and will not be repeated here. Each unit in the aforementioned sound masking control device can be implemented entirely or partially through software, hardware, or a combination thereof. These units can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each unit.

[0090] In one embodiment, an air conditioner is provided. The air conditioner includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the sound masking control method of the above embodiment.

[0091] In one embodiment, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, causes the processor to perform the sound masking control method as described in the above embodiment.

[0092] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0093] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. 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. Such 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, and should all be included within the protection scope of the present invention.

Claims

1. A sound masking control method, characterized in that, include: Obtain the current operating parameters of the indoor unit and the fan speed switching command, and determine the target fan speed and the target air guide plate angle according to the fan speed switching command; The current operating parameters include the current fan speed, the current air guide plate angle, and the current operating mode; The rate of change of rotation speed is determined based on the current fan speed and the target fan speed. The change in the angle of the air guide plate is determined based on the current angle of the air guide plate and the angle of the target air guide plate, and the change in the angle of the air guide plate is compared with a preset angle threshold to obtain the result of the motion state determination of the air guide plate. The resonant point speed of the indoor unit's fan blades is obtained, and it is determined whether the current fan speed changes to the target fan speed and passes through the resonant point speed of the fan blades, so as to obtain the resonance crossing determination result. The rotational speed change rate, the resonance crossing determination result, the wind deflector motion state determination result, and the current operating mode are input into the pre-constructed auditory abruptness assessment model to obtain the auditory abruptness level; The frequency band configuration scheme of the masking sound is determined according to the auditory abruptness level, and the masking sound is generated according to the frequency band configuration scheme. The masking sound is played before the fan speed switching command is executed.

2. The sound masking control method according to claim 1, characterized in that, The method for constructing the auditory abruptness assessment model includes: Set up multiple indoor unit test conditions with different combinations of operating parameters, collect indoor unit operating noise samples under each set of indoor unit test conditions, and form a noise sample set; Obtain subjective perception scores from multiple groups of subjects for each indoor unit operating noise sample in the noise sample set, forming a subjective score set that corresponds one-to-one with each indoor unit operating noise sample. Extract the operating parameter features corresponding to the test conditions of each group of indoor units to form an objective feature set that corresponds one-to-one with the operating noise sample of each indoor unit. The subjective rating set is matched and associated with the objective feature set to analyze the correlation between each operating parameter feature and the subjective perception rating, and feature items are selected. Based on the aforementioned features, a multivariate linear regression model combined with a nonlinear correction term is used for fitting to obtain the auditory abruptness assessment model.

3. The sound masking control method according to claim 2, characterized in that, The auditory abruptness assessment model, obtained by fitting the model based on the aforementioned feature terms using multiple linear regression combined with a nonlinear correction term, includes: Using the rotational speed change rate, resonance crossing judgment result, wind deflector motion state judgment result, and current operating mode as linear independent variables, and the subjective perception score as the dependent variable, an initial regression model is established. Calculate the prediction residual of the initial regression model, and determine the systematic bias of the initial regression model in the low-speed and high-speed change ranges based on the trend of the prediction residual with the rate of change of the rotational speed. Based on the systematic deviation, a nonlinear transformation is applied to the rotational speed change rate, and the resonance crossing judgment result, the wind deflector motion state judgment result, the current operating mode, and the rotational speed change rate after nonlinear transformation are used as updated independent variables. Regression fitting is then performed again to obtain an updated regression model. Based on the fitting results of the updated regression model, the coefficient values ​​of each independent variable in the updated regression model and the level threshold range of the hearing abruptness level are determined to obtain the hearing abruptness assessment model.

4. The sound masking control method according to claim 1, characterized in that, The process of inputting the rotational speed change rate, the resonance crossing determination result, the wind deflector motion state determination result, and the current operating mode into a pre-built auditory abruptness assessment model yields auditory abruptness levels, including: The rotational speed change rate, the resonance crossing determination result, the wind deflector motion state determination result, and the current operating mode are input into the pre-constructed auditory abruptness evaluation model to obtain the auditory abruptness. Determine whether the auditory abruptness is less than a first preset threshold; If the auditory abruptness is less than the first preset threshold, it is determined to be a low abruptness level; If the auditory abruptness is greater than or equal to the first preset threshold, then it is determined whether the auditory abruptness is less than or equal to the second preset threshold; wherein the second preset threshold is greater than the first preset threshold; If the auditory abruptness is less than or equal to the second preset threshold, it is determined to be of medium abruptness level; If the auditory abruptness is greater than the second preset threshold, it is determined to be a high abruptness level.

5. The sound masking control method according to claim 4, characterized in that, The frequency band configuration scheme for determining the masking sound based on the auditory abruptness level includes: When the auditory abruptness level is high abruptness level, the frequency band configuration scheme is determined to be simultaneously outputting high-frequency masking sound, mid-frequency masking sound and low-frequency masking sound; When the auditory abruptness level is medium abruptness level, the frequency band configuration scheme is determined to be simultaneously outputting high-frequency masking sound and mid-frequency masking sound; When the auditory abruptness level is low abruptness, the frequency band configuration scheme is determined to output high-frequency masking sound.

6. The sound masking control method according to claim 5, characterized in that, The high-frequency masking sound has a frequency band of 1kHz-2kHz; the mid-frequency masking sound has a frequency band of 500Hz-1kHz; and the low-frequency masking sound has a frequency band of 20Hz-500Hz.

7. The sound masking control method according to claim 1, characterized in that, After the masking sound is played, the following is included: The moment when the current fan speed reaches the target fan speed and the fluctuation range is less than a preset speed fluctuation threshold is recorded as the stable start moment. Obtain the first intensity value of the masking sound corresponding to the stable start time, and generate an intensity attenuation control curve based on the first intensity value and a preset attenuation time; Starting from the stable start time, the real-time output intensity at each time point is calculated one by one according to the attenuation coefficient corresponding to each time point in the intensity attenuation control curve. Based on the real-time output intensity at each of the aforementioned time points, the masking sound is controlled to be played at the corresponding time points until the attenuation coefficient decreases to zero, at which point the masking sound is stopped.

8. A sound masking control device, characterized in that, include: The acquisition module is used to acquire the current operating parameters of the indoor unit and the fan speed switching command, and determine the target fan speed and the target air guide plate angle according to the fan speed switching command; The current operating parameters include the current fan speed, the current air guide plate angle, and the current operating mode; The determining module is used to determine the rate of change of rotational speed based on the current fan speed and the target fan speed; The comparison module is used to determine the change in the air guide plate angle based on the current air guide plate angle and the target air guide plate angle, and compare the change in the air guide plate angle with a preset angle threshold to obtain the air guide plate motion state determination result. The judgment module is used to obtain the resonant point speed of the fan blades of the indoor unit, determine whether the current fan speed changes to the target fan speed and passes through the resonant point speed of the fan blades, and obtain the resonance crossing judgment result. The input module is used to input the rotational speed change rate, the resonance crossing determination result, the wind deflector motion state determination result, and the current operating mode into the pre-constructed auditory abruptness assessment model to obtain the auditory abruptness level; The generation module is used to determine the frequency band configuration scheme of the masking sound according to the auditory abruptness level, generate the masking sound according to the frequency band configuration scheme, and play the masking sound before the fan speed switching command is executed.

9. An air conditioner, 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 computer program, it implements the sound masking control method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, causes the processor to perform the sound masking control method as described in any one of claims 1 to 7.

Images